WORKING PAPER No 3

ANALYZING CONDITIONS OF SERVICE FOR AGRICULTURAL

RESEARCHERS AN EXPERIMENT

USING EARNINGS FUNCTIONS

ntemational Service for National Aqicultural Research

The international Service for National Agricultural Research (ISNAR) began operating at its headquarters in The Hague Netherlands on September 1 1980 It was established by the Consultative Group on International Agricultural Research (CGIAR) on the basis of recommendations from an international task force for the purpose of assisting goverrunents of developing countries to strengthen their agricultural research It is a non-profit autonomous agency international in character and non-political in management staffing and operations

Of the thirteen centers in the CGIIR network ISNAP is the only one which focuses primarily on national agricultural research issues it provides advice to governments upon request on organization planning manpower development staff requirements firtancial and infrastucture requirements and related mattervs thus comp]ementing the activities of other ass 1stan-e agencies Additionally lSNAR har an active taining and communications program which cooperates with iatonal agricultural research pr eorains in developing countries

ISHAR also plays an active rol in assisting these national programs to establitsh links with both the intenational agricultural research centers and doriors

ISNIAR is upported by a rumter of the members of CGIAR an informal group of approximately 30 donors it includes countries development banks international organizations and foundations In 1985 funding for ISNARs core program was provided by

Austral ia Belgium Canada

European Economic Coivmunity Federal Republic of Germany

Ford Foundation France Ireland Italy

Netherlands

Philippines Spa in Swedjn

Switzerland

United Kingdom United States Agency for International Development

World Bank

Citation

International Servce for National Agricultural Research Analyzing Conditions of Service for Agricultural Researchers An Experiment Using Earnings Functions 1985 The Hague Netherlands

WORKING PAPER No 3

ANALYZING CONDITIONS OF SERVICE FOR AGRICULTURAL

RESEARCHERS AN EXPERIMENT

USING EARNINGS FUNCTIONS

Howard ELLIOTT and Sandra KANG

July 1985

inlrr International Service for National ricultural Research

SNq -- ING ERS

anlys~n t aL uq r ale Qa a

pbestreearchsyagr c esetrc s mdrs ev S

bra a q~r aeilsw v soQ

forma~~~pout~ flcatiors program The work 1nrgapers0 seres

enacsi program in several impoL-Eant1 was

1 These papers are iitended t o be a rapidmeans of pree n e

Y results of~wor~k and eperiences tha arestill in progres r

~ka~ead~rdudngre esh~tcoldbe of use to other

3Tey are-ntended tobea 7 efcv hI1sfriepn e

is~cussion of continuing workteb increas inthe qual tyof

final products critical comnent s welcomed

3 The series provides anul1et for diffusing materials~adnd If

that because o the6ir 1imited coeag do n6t mee the requ rements-

o eneralI audience publication

The~ sei isitnq mail1y for~ the fuJoomaerial r u

ISNA1R stafE butN t isalso vai1ab1 e r thejpublia on o dc nt

A produced bot n itns boulPtheyw sh to tae advnaeo

oportuniy

ACKNOWLEDGEMENTS

The opinions expressed in this document are the sole responsibility of

the authors

This paper represnts ar attempt to develop i fthoIologanalyzingor

conditions of service usincg administratve data or ato f row surveys

carried out as part of ISNAR rev ews of iration i ag ricultural research

systems

We would like to acknowledqe the constructive cotrnzts received from

colleagues and the assistance of those staff members who first generated

the data used in the pape r

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

The international Service for National Agricultural Research (ISNAR) began operating at its headquarters in The Hague Netherlands on September 1 1980 It was established by the Consultative Group on International Agricultural Research (CGIAR) on the basis of recommendations from an international task force for the purpose of assisting goverrunents of developing countries to strengthen their agricultural research It is a non-profit autonomous agency international in character and non-political in management staffing and operations

Of the thirteen centers in the CGIIR network ISNAP is the only one which focuses primarily on national agricultural research issues it provides advice to governments upon request on organization planning manpower development staff requirements firtancial and infrastucture requirements and related mattervs thus comp]ementing the activities of other ass 1stan-e agencies Additionally lSNAR har an active taining and communications program which cooperates with iatonal agricultural research pr eorains in developing countries

ISHAR also plays an active rol in assisting these national programs to establitsh links with both the intenational agricultural research centers and doriors

ISNIAR is upported by a rumter of the members of CGIAR an informal group of approximately 30 donors it includes countries development banks international organizations and foundations In 1985 funding for ISNARs core program was provided by

Austral ia Belgium Canada

European Economic Coivmunity Federal Republic of Germany

Ford Foundation France Ireland Italy

Netherlands

Philippines Spa in Swedjn

Switzerland

United Kingdom United States Agency for International Development

World Bank

Citation

International Servce for National Agricultural Research Analyzing Conditions of Service for Agricultural Researchers An Experiment Using Earnings Functions 1985 The Hague Netherlands

WORKING PAPER No 3

ANALYZING CONDITIONS OF SERVICE FOR AGRICULTURAL

RESEARCHERS AN EXPERIMENT

USING EARNINGS FUNCTIONS

Howard ELLIOTT and Sandra KANG

July 1985

inlrr International Service for National ricultural Research

SNq -- ING ERS

anlys~n t aL uq r ale Qa a

pbestreearchsyagr c esetrc s mdrs ev S

bra a q~r aeilsw v soQ

forma~~~pout~ flcatiors program The work 1nrgapers0 seres

enacsi program in several impoL-Eant1 was

1 These papers are iitended t o be a rapidmeans of pree n e

Y results of~wor~k and eperiences tha arestill in progres r

~ka~ead~rdudngre esh~tcoldbe of use to other

3Tey are-ntended tobea 7 efcv hI1sfriepn e

is~cussion of continuing workteb increas inthe qual tyof

final products critical comnent s welcomed

3 The series provides anul1et for diffusing materials~adnd If

that because o the6ir 1imited coeag do n6t mee the requ rements-

o eneralI audience publication

The~ sei isitnq mail1y for~ the fuJoomaerial r u

ISNA1R stafE butN t isalso vai1ab1 e r thejpublia on o dc nt

A produced bot n itns boulPtheyw sh to tae advnaeo

oportuniy

ACKNOWLEDGEMENTS

The opinions expressed in this document are the sole responsibility of

the authors

This paper represnts ar attempt to develop i fthoIologanalyzingor

conditions of service usincg administratve data or ato f row surveys

carried out as part of ISNAR rev ews of iration i ag ricultural research

systems

We would like to acknowledqe the constructive cotrnzts received from

colleagues and the assistance of those staff members who first generated

the data used in the pape r

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

WORKING PAPER No 3

ANALYZING CONDITIONS OF SERVICE FOR AGRICULTURAL

RESEARCHERS AN EXPERIMENT

USING EARNINGS FUNCTIONS

Howard ELLIOTT and Sandra KANG

July 1985

inlrr International Service for National ricultural Research

SNq -- ING ERS

anlys~n t aL uq r ale Qa a

pbestreearchsyagr c esetrc s mdrs ev S

bra a q~r aeilsw v soQ

forma~~~pout~ flcatiors program The work 1nrgapers0 seres

enacsi program in several impoL-Eant1 was

1 These papers are iitended t o be a rapidmeans of pree n e

Y results of~wor~k and eperiences tha arestill in progres r

~ka~ead~rdudngre esh~tcoldbe of use to other

3Tey are-ntended tobea 7 efcv hI1sfriepn e

is~cussion of continuing workteb increas inthe qual tyof

final products critical comnent s welcomed

3 The series provides anul1et for diffusing materials~adnd If

that because o the6ir 1imited coeag do n6t mee the requ rements-

o eneralI audience publication

The~ sei isitnq mail1y for~ the fuJoomaerial r u

ISNA1R stafE butN t isalso vai1ab1 e r thejpublia on o dc nt

A produced bot n itns boulPtheyw sh to tae advnaeo

oportuniy

ACKNOWLEDGEMENTS

The opinions expressed in this document are the sole responsibility of

the authors

This paper represnts ar attempt to develop i fthoIologanalyzingor

conditions of service usincg administratve data or ato f row surveys

carried out as part of ISNAR rev ews of iration i ag ricultural research

systems

We would like to acknowledqe the constructive cotrnzts received from

colleagues and the assistance of those staff members who first generated

the data used in the pape r

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

SNq -- ING ERS

anlys~n t aL uq r ale Qa a

pbestreearchsyagr c esetrc s mdrs ev S

bra a q~r aeilsw v soQ

forma~~~pout~ flcatiors program The work 1nrgapers0 seres

enacsi program in several impoL-Eant1 was

1 These papers are iitended t o be a rapidmeans of pree n e

Y results of~wor~k and eperiences tha arestill in progres r

~ka~ead~rdudngre esh~tcoldbe of use to other

3Tey are-ntended tobea 7 efcv hI1sfriepn e

is~cussion of continuing workteb increas inthe qual tyof

final products critical comnent s welcomed

3 The series provides anul1et for diffusing materials~adnd If

that because o the6ir 1imited coeag do n6t mee the requ rements-

o eneralI audience publication

The~ sei isitnq mail1y for~ the fuJoomaerial r u

ISNA1R stafE butN t isalso vai1ab1 e r thejpublia on o dc nt

A produced bot n itns boulPtheyw sh to tae advnaeo

oportuniy

ACKNOWLEDGEMENTS

The opinions expressed in this document are the sole responsibility of

the authors

This paper represnts ar attempt to develop i fthoIologanalyzingor

conditions of service usincg administratve data or ato f row surveys

carried out as part of ISNAR rev ews of iration i ag ricultural research

systems

We would like to acknowledqe the constructive cotrnzts received from

colleagues and the assistance of those staff members who first generated

the data used in the pape r

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

ACKNOWLEDGEMENTS

The opinions expressed in this document are the sole responsibility of

the authors

This paper represnts ar attempt to develop i fthoIologanalyzingor

conditions of service usincg administratve data or ato f row surveys

carried out as part of ISNAR rev ews of iration i ag ricultural research

systems

We would like to acknowledqe the constructive cotrnzts received from

colleagues and the assistance of those staff members who first generated

the data used in the pape r

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

TABLE OF CONTENTS

I INTRODUCTION 1

II THE UMA1N CAPITAL APPROACH AND THE EARNINGS FUNCrION 3

Ill THE USE OF ADMIII TI F DATA FOR POLICY ANALYSIS 9

THREE CASE STUDIES

A Dominican Republic 11

B ZImbabwe 16

C Tha i land 20

IV STAFF DEVELOPMENT AND THE EARNINGS FUNCTION 31

V CONCLUD[NG OSERVATIOHS 32

ANNElt 1 The Earningn Function Approach and its Interpretation 34

AJNEX 2 Descriptive Statistics on Key Variables 37

ANNEX 3 Human Resource Inventory Survey 39

BIBLIOGRAPHY 42

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

4

C- ra res arc

ag~1~ rserchr( o~ searc Tsite~wnarewqia sy

retain and mok t it scientificf laboz orce

Inany f its sstem revew-ISNAR~hsds~ -ta hgvrmn

wa osdrngarvso of the-salaries -or researchiers mo reoof

S than not in the face of loss of its big hampumn re46 rces o othe

sectors~The oft-proposed solution of paying higher salaries o

resoearchers involves howver questlona a 1ioth the Iii el~and the

structure of saaieISNAR also requires away lof Prjetng the cos

implicatofoCan recommended increase whether it be across theboard

or infavor of particular groupsin sc ce suppl 4 ~ ~ A N

present~ pae snvlpntolfrih nis~~~l~

~The~p~n~pr( a first step in e pn ol o the anlsa

of salary structure in agricultural researh ssitem sw Thepr

method can be used as asada4 part of ISNA y em rviewa joro

providinginsights ito theraward structure as a apea al -rojec to 4

A

studycondiutios~fsevqeina particularcut hear

-reogflcze thatto be~tiaeful a tool mustbe to ativelamp e to us

a Xtmuat not ioealargeprobleofdacco

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

3 b 11

athe ser

dqa t 1cnWrw~ f s

~Th~ ~ae s~ ducioin a t ith ec I es~Ms zit -realrpolm~

Th ngbt n~ uc~on terass uct Eerig n dplusmnc

0 ex gene esa a sruesfo e hr C

stuw i is nate o A friame(Zirnl-wh eon cralla

relapeTh rei r giocioharstre asdene S ii eCn

intrdtoe heean s~ 4e alo coe

- tn

isl nonposfbl as~ Se Ietribes te~dae coaroefi Ies-L toaII a sues r par hj a tren _1

analsesandthewayin dhichgethe-datas-worevean as

conr ~ ao flStu~epadise 6 ~oire aa

relnt sutoreasd sttc u oprI a h soheibrer ce

soeveom~etsa n thesefulnss otoe pe arr e tons a twayor

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

CERAPIL~PP AND THEAIUIINGS FUNCTION

a p c trats I m~ nua

a ivetmnn pyic 1-cp a ivdalsor e

~ gf~~a1~liaeud cationi cone rte ary in an

n exer encae acquired-onapart-time b~asis The_aquisiti o

caianvolv~es rpaI costs the- i diviid a or -hi snsocsa Ti ay

binthe form of 1foregone incomie redcued pnrduCtivi y duri g t ra i gr

and diroct uition and training costs Hfowever the individfual acquirn 9 shy

the~increased level of skills will enjoy a higher level of prod cicitv_

and often a correspondingly higher level of icom e

Eduication as an investm1ent hag been shown to offer a nate of ret~r -on

inetdcptlwihi higher than that~ earna on hysa1 capital

whethe disc~ountad pnesent Ivalue of the additionaloudtpu t utng o

the trinn iscmae it h ot of proviing -that training-n i

7 dvaloped cowunries where labor markets are finely uned an

inivdalsn eangs ma btaeasamsueofis prodict ivity at

employers and comptition among empoyers ensures thatith employea can

a arnate humanselhissre t~ whchrflects the amnirt f capi1 a ea

Possesses an thecost of obta inngit

devlo ing c~ountries theprediction of eannsrom a muiU l

regression analysis of aducation and experience_ actors-s- nsgtra a

-iuunr jaola retifo matin-ad ini an ying-statist

larit-ie ch Sgge eapri nc-ipa Ide- aria of aar ha

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

be e~nns C~ dlvd a -of~ua shya Ia o a o

epeincedo a e~gc s u ean

staisicaignfi ncofEe asoitonb etvie eamr and e 4ar u er Qnal-and joa actei ri-co a-powerful -ude the~orking of the Lowla dsrcur

lt The amounrt of human capiaiposessed by a niiuld sa yrie

by he f ya~sofformal-schooling completed byan ind viduaaio-umbr

Sby~ yearsof jobexperiece in~our Lormulato weiuse a_ser ies

A~Cj~zero-one dummoy variables representing th highest deg9re ac u2 completed by theindividual instead~of ja cot ir~ou_syea re sof oo

vaibebeas ot esne~i urveys willl re or in omati~o compleited degres but not like y~ record yars 6f -schooing

4 prierceis also a very impportant deermian of produativI y especially in8cientificprofessionsan b way inw ch saa rse with experience isverr imora n~4 rin ng theabill ye

Io8ctai31 it experienced researchers In-additionothse

components ofhumnan capital we reuirvarious control var iab08 or

oherinflfuences on the earningsofanindividualThe 6fwse eXoa

pesnal nature (e g whther_ te r ismale Jraace Olta

Insittioalnatre (e~ wether the rsarcher s~d I

researc ni tet o he nis ot agralt re wet h a -iutof the capitaliyplusmntorhet eheLoccp s-a sit of hod

6s ~ nt~ o1i Oa g ny Cesa a a a senior oan ecof gre e te

it cu n a a

r-1 -Uer ~ oo Fe- Ea wa

-coioperei ey~ once 6o nal

-he -earnlng functindn scussed p ne ak s na orn

~nI SALARY =~a + +d e7-OT HER+ b EDU Mc EXPER EXPERSQ RES I

whe re

In ALARY- the natural~ Iog of-monthly salary~in oal Currerc

EDUC IA~ a dummy variable epreoen ig th0 hgestdegreaane

~OR insome fonulaon1 h ac ua number of yero

~rschool ig completed

EXPER~- experienceas mea sured by henumber of ~years of po -en a

wrk epe ncsice qgraduaton

-P~Q yearsjof -experienceq~eof the number of

-= lee

istitutional factiorsand

ii) AL aLeS idare resentingau-e o ae al ae

ca sse a t p e ria a a3

OTHE an array of control var a re sn esonal

aners 0 a a ca eprs a S

F guwo eTIncomeProf11e

EarnnqsMonth 1 NN ~

1Thes ha e o he ncoime Prot le 1S ourC Uizd a6 u1 ar

of~n a da e ereA dicuer ib

e~it1 C) t IMfqc s 1 pa e-- 0 coap e as a etuq-m -e id a~~ls a r ati th 0

ndi 1 sual ar esoa ancs

eoan IanerEity a cofe

a g P r hewkl ra1 nr o

WillCI tie noef 1~o n caretD~b exe

wo tbeca p~a~i egrees wniiy l dxc~dve1~f fona nae dua

ability to profi t from exprience and maintain h poductii late r in

litfe -

Th enec for the curv e to turn downward owards the-en fa pron s

4career is unertndablP when we consider hat~th6 6rve is drawn in

-crosset~io A workor ~imid-career (atthe peak)- s continuinga

receiv triig poruiisi antiipa tiori pf high produc iv oo

the remainnggtyears of h3 ae Th okr c mi towardsheen

hscreer has rela tielyf ee er years ove wi add tio-l ra ini an

provide beifits thus there s~atndency for or er orece ive 9

rainig ae rIthori ft rcardeers

re1rive earn iq toward the -end of the r6aer

I asstate tha t e degre o f nentivepovi d ea

wsa qtion 00th lve and -ucture f I e~

prse ca nadecn a

Earn i osIMont I 7 Im

~2

2lrUEX ec

7Cuve ~ho~ ~ ~o e e ~ u a~Ja~io~~ he

a~a~ Izt eea az~z~ 7h1yeais b~ h~~ndv dult

inevear th earyears of heMir iiduals canowee bndo es ol

Ilab esine 0~ ticcaya reacin wythei t gauate eprad rie 1

-havetheyms danc~iedp e ahe doheas e anrra)ete xeiaxiatehii laea eb nd~ coed t g sarting al ie a

Howeaze o atjh h thi 11g8 tone i d ano if thearatr ybepe i heg brea enst-o 1te

isehih t a heygtcly ehternep teu erap risem~

ealyear sueu o at tothompoy an rnaac rthem

Sklst ye aqien t jo recot ma~ke b en Yhaeh

Vernlnf e ter paea fia be~~ahd Sx ero a

ac r

Vdw-co aqiyUoipri

Th I U~ I ~ Ftyt-W

ot ywt uCr 3 in nuucte iI CQfBc

ise t m e y epdqnies re1

waa the curv 6I het~od hob

os 4~ n bhetweenp be re1nga e en h

eae t 1a a s re_ whcda

Uciet3rG)aftio The oinE1ofKentry totheytm evfQ e t

rs e Ir~n ru (x- tu

9 wi thin h e

we tostmte ea a for

our thre ase studies W~ewishi to- describe te rewar d 3truc ture- of h9

ssein termis of the shaeo th ccoepofI e Ind lien 3eQ f6 si

allows us to say somnething about- he appropria P fs t he ru a e n

the -lght of the staffing- situaia ono e system

In the followi~ng sectin pattro in functins

II THE USE OF ADMINISTRATIVE DATA FOR FOLICYAMAfLYS IS

f aant teppose mto oloqy_ s ha eur Onyoah aes oo ~ a e asoT d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

aners 0 a a ca eprs a S

F guwo eTIncomeProf11e

EarnnqsMonth 1 NN ~

1Thes ha e o he ncoime Prot le 1S ourC Uizd a6 u1 ar

of~n a da e ereA dicuer ib

e~it1 C) t IMfqc s 1 pa e-- 0 coap e as a etuq-m -e id a~~ls a r ati th 0

ndi 1 sual ar esoa ancs

eoan IanerEity a cofe

a g P r hewkl ra1 nr o

WillCI tie noef 1~o n caretD~b exe

wo tbeca p~a~i egrees wniiy l dxc~dve1~f fona nae dua

ability to profi t from exprience and maintain h poductii late r in

litfe -

Th enec for the curv e to turn downward owards the-en fa pron s

4career is unertndablP when we consider hat~th6 6rve is drawn in

-crosset~io A workor ~imid-career (atthe peak)- s continuinga

receiv triig poruiisi antiipa tiori pf high produc iv oo

the remainnggtyears of h3 ae Th okr c mi towardsheen

hscreer has rela tielyf ee er years ove wi add tio-l ra ini an

provide beifits thus there s~atndency for or er orece ive 9

rainig ae rIthori ft rcardeers

re1rive earn iq toward the -end of the r6aer

I asstate tha t e degre o f nentivepovi d ea

wsa qtion 00th lve and -ucture f I e~

prse ca nadecn a

Earn i osIMont I 7 Im

~2

2lrUEX ec

7Cuve ~ho~ ~ ~o e e ~ u a~Ja~io~~ he

a~a~ Izt eea az~z~ 7h1yeais b~ h~~ndv dult

inevear th earyears of heMir iiduals canowee bndo es ol

Ilab esine 0~ ticcaya reacin wythei t gauate eprad rie 1

-havetheyms danc~iedp e ahe doheas e anrra)ete xeiaxiatehii laea eb nd~ coed t g sarting al ie a

Howeaze o atjh h thi 11g8 tone i d ano if thearatr ybepe i heg brea enst-o 1te

isehih t a heygtcly ehternep teu erap risem~

ealyear sueu o at tothompoy an rnaac rthem

Sklst ye aqien t jo recot ma~ke b en Yhaeh

Vernlnf e ter paea fia be~~ahd Sx ero a

ac r

Vdw-co aqiyUoipri

Th I U~ I ~ Ftyt-W

ot ywt uCr 3 in nuucte iI CQfBc

ise t m e y epdqnies re1

waa the curv 6I het~od hob

os 4~ n bhetweenp be re1nga e en h

eae t 1a a s re_ whcda

Uciet3rG)aftio The oinE1ofKentry totheytm evfQ e t

rs e Ir~n ru (x- tu

9 wi thin h e

we tostmte ea a for

our thre ase studies W~ewishi to- describe te rewar d 3truc ture- of h9

ssein termis of the shaeo th ccoepofI e Ind lien 3eQ f6 si

allows us to say somnething about- he appropria P fs t he ru a e n

the -lght of the staffing- situaia ono e system

In the followi~ng sectin pattro in functins

II THE USE OF ADMINISTRATIVE DATA FOR FOLICYAMAfLYS IS

f aant teppose mto oloqy_ s ha eur Onyoah aes oo ~ a e asoT d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

eoan IanerEity a cofe

a g P r hewkl ra1 nr o

WillCI tie noef 1~o n caretD~b exe

wo tbeca p~a~i egrees wniiy l dxc~dve1~f fona nae dua

ability to profi t from exprience and maintain h poductii late r in

litfe -

Th enec for the curv e to turn downward owards the-en fa pron s

4career is unertndablP when we consider hat~th6 6rve is drawn in

-crosset~io A workor ~imid-career (atthe peak)- s continuinga

receiv triig poruiisi antiipa tiori pf high produc iv oo

the remainnggtyears of h3 ae Th okr c mi towardsheen

hscreer has rela tielyf ee er years ove wi add tio-l ra ini an

provide beifits thus there s~atndency for or er orece ive 9

rainig ae rIthori ft rcardeers

re1rive earn iq toward the -end of the r6aer

I asstate tha t e degre o f nentivepovi d ea

wsa qtion 00th lve and -ucture f I e~

prse ca nadecn a

Earn i osIMont I 7 Im

~2

2lrUEX ec

7Cuve ~ho~ ~ ~o e e ~ u a~Ja~io~~ he

a~a~ Izt eea az~z~ 7h1yeais b~ h~~ndv dult

inevear th earyears of heMir iiduals canowee bndo es ol

Ilab esine 0~ ticcaya reacin wythei t gauate eprad rie 1

-havetheyms danc~iedp e ahe doheas e anrra)ete xeiaxiatehii laea eb nd~ coed t g sarting al ie a

Howeaze o atjh h thi 11g8 tone i d ano if thearatr ybepe i heg brea enst-o 1te

isehih t a heygtcly ehternep teu erap risem~

ealyear sueu o at tothompoy an rnaac rthem

Sklst ye aqien t jo recot ma~ke b en Yhaeh

Vernlnf e ter paea fia be~~ahd Sx ero a

ac r

Vdw-co aqiyUoipri

Th I U~ I ~ Ftyt-W

ot ywt uCr 3 in nuucte iI CQfBc

ise t m e y epdqnies re1

waa the curv 6I het~od hob

os 4~ n bhetweenp be re1nga e en h

eae t 1a a s re_ whcda

Uciet3rG)aftio The oinE1ofKentry totheytm evfQ e t

rs e Ir~n ru (x- tu

9 wi thin h e

we tostmte ea a for

our thre ase studies W~ewishi to- describe te rewar d 3truc ture- of h9

ssein termis of the shaeo th ccoepofI e Ind lien 3eQ f6 si

allows us to say somnething about- he appropria P fs t he ru a e n

the -lght of the staffing- situaia ono e system

In the followi~ng sectin pattro in functins

II THE USE OF ADMINISTRATIVE DATA FOR FOLICYAMAfLYS IS

f aant teppose mto oloqy_ s ha eur Onyoah aes oo ~ a e asoT d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Earn i osIMont I 7 Im

~2

2lrUEX ec

7Cuve ~ho~ ~ ~o e e ~ u a~Ja~io~~ he

a~a~ Izt eea az~z~ 7h1yeais b~ h~~ndv dult

inevear th earyears of heMir iiduals canowee bndo es ol

Ilab esine 0~ ticcaya reacin wythei t gauate eprad rie 1

-havetheyms danc~iedp e ahe doheas e anrra)ete xeiaxiatehii laea eb nd~ coed t g sarting al ie a

Howeaze o atjh h thi 11g8 tone i d ano if thearatr ybepe i heg brea enst-o 1te

isehih t a heygtcly ehternep teu erap risem~

ealyear sueu o at tothompoy an rnaac rthem

Sklst ye aqien t jo recot ma~ke b en Yhaeh

Vernlnf e ter paea fia be~~ahd Sx ero a

ac r

Vdw-co aqiyUoipri

Th I U~ I ~ Ftyt-W

ot ywt uCr 3 in nuucte iI CQfBc

ise t m e y epdqnies re1

waa the curv 6I het~od hob

os 4~ n bhetweenp be re1nga e en h

eae t 1a a s re_ whcda

Uciet3rG)aftio The oinE1ofKentry totheytm evfQ e t

rs e Ir~n ru (x- tu

9 wi thin h e

we tostmte ea a for

our thre ase studies W~ewishi to- describe te rewar d 3truc ture- of h9

ssein termis of the shaeo th ccoepofI e Ind lien 3eQ f6 si

allows us to say somnething about- he appropria P fs t he ru a e n

the -lght of the staffing- situaia ono e system

In the followi~ng sectin pattro in functins

II THE USE OF ADMINISTRATIVE DATA FOR FOLICYAMAfLYS IS

f aant teppose mto oloqy_ s ha eur Onyoah aes oo ~ a e asoT d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

ac r

Vdw-co aqiyUoipri

Th I U~ I ~ Ftyt-W

ot ywt uCr 3 in nuucte iI CQfBc

ise t m e y epdqnies re1

waa the curv 6I het~od hob

os 4~ n bhetweenp be re1nga e en h

eae t 1a a s re_ whcda

Uciet3rG)aftio The oinE1ofKentry totheytm evfQ e t

rs e Ir~n ru (x- tu

9 wi thin h e

we tostmte ea a for

our thre ase studies W~ewishi to- describe te rewar d 3truc ture- of h9

ssein termis of the shaeo th ccoepofI e Ind lien 3eQ f6 si

allows us to say somnething about- he appropria P fs t he ru a e n

the -lght of the staffing- situaia ono e system

In the followi~ng sectin pattro in functins

II THE USE OF ADMINISTRATIVE DATA FOR FOLICYAMAfLYS IS

f aant teppose mto oloqy_ s ha eur Onyoah aes oo ~ a e asoT d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

ne~~~~ ~al4co0_o in( ingir

-

bot4Trat ooX

c~beQ~tpU~p (Iication f ean~a

~F tu i~~e~

poronef~nor~~q~or~hthe han1 qwasuy- ve t h in fas -u -1d h 6 N

hc Lco(ani qtprob e atems ls secandjre a ryastpr

gae ra tsion~4 theI-d thtam oflisy sucse countre

orrtapromotu~eflinof e e ocentif1 inco tut per ora -wr

wolhoeu sur_-veycomem eue ofethisin hi per son~ brI

dn ienf r-ec n -onbthe reserLhaniorr saler ichoe1L~o in family

am raphic mobiliyt (a prb x e nsystemst refs(t onshard to atrat

qustonare(seAiPhe ) le n i t sn Unadti of-nomto acrs cutvsadis -cvrgOfal shyn fo on6f apri on -

Voess~ nature imiortant ian sicy gcn t 1n ofase rc

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

t n t

t a callw moreti

ctran abe by te r

4f~dinatrtie rch t eeals 6s

1 petedt lpresea~acc ani-thusservesC

Vysa~1 nuxo~rsu eatneeear a ia6

di udgTTPK a at 1 anatia~ogca oL Pmicrcnue a st

systems-and therefore th stdy)is one1 th~at coud easi~ly be carie d out-

in4 f i e uig eiwof anational agricuiltural research sysem-

ADo inca Republic4gtA44

6K The Division of Agicultural Research (DIA) in the Mnidry of

~WKAgriculture agreed t~oundet_ke a specialsuavey of its efnployeeP using a

qtiueofia proposed by ISNAR This survey provided inhformat ion onthe

~44a~4Be~4 experen o-Acial position andscificeucpon~wrk

outptth o tatached to DIAiniviualscinti meot ec

anaeiosbisynotHoevr 44 t 44 he e

~resuilts The information supplied bj$ th DIA caefom all five Stat4 one

udcer~their jurisadiction~444~4 4 A~ 4 44~~ ~ 44444 x

esimte Linctin reeachr th oWe 4 the~ eanig fo -in4 D h

insteituioal iha enisptici forhe reut Eadches ie tiaOn -A1L lni~npu~Vc ~n~~ir~V~f~e~enco~i m on of ~pe 01

444~ 4i 4(44

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

12

presented in Table 1 below and the corresponding eacnirts profile are

presented in Figure 3

Tab I1 Ealrni ngs Functions for Dominican epuLbl iC

Variable Equation 1 Equation 2 Equationi 3 Equation 4 (n = 72) (n ( =-7272 (n = 72)

CONSTANT 602 609 56 59 (83 )(10861 (1692)h (5989)k

EXPER 003146 0026 0015 0047 (2379)k (197)k (1147) (267)

EXPERSQ -00(j047 -000026 -000035 -000099 (000067) (0397) (055) (109)

LICENCIADO -0038 -01218 -0207 +0553 (0505) (210) (126) (0551)

INGENIERO 0093 bull bull -0085 +0198 (157) (0518) (254)

MS 0261 0186 0088 036 (338) (309)k (0509) (3548)

SHORTCOURSE bull bull 018 -0067 (109) (0284)

NON1NIV bull bull -008 -014

(143) (0917)

DIRECTOR 0498 0534 0505 (743) (705) (698)

EUCARCADlt bullbull 0036 0055 (0815) (128)

F F ALF -0151 -014 -010 -0188 (315) (289) (226)k (291)

AGE +0018

(321)

067 067 072 040

Note The figures in parentheses a-- t statistc all t statistics significant at the 975 level are marked with an asterisk

R 1 a corrtctod R which -l lows us to corrpar eotimatng equations with diffe rent nurit-rz of independent var nab2I+

Descriptive Stit st I-+on ke vriables re found in Anne- 2

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

io

CD

t

A_ 2

70

61

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

10os tasshoY

Of reeac ereu ngasexolan dyvao r ia I EXPET a R t r~~ora wi ot eeaoadf~ij eo ~asF~AW

uueniaan~e~guay~ aoi4~a Lun~ t~ S~9fqu~ fl ~ ~Lt ~ I var~fnbl J ad e oro arbeOamp e typeoa trodlsc a ~ yh~chonrolforoto-~~ -hraris~i fpoiknh_a astp

-a7 d444O~lao~n6

Teneallngs esutin wbea ru ~samet~o in the sesf4 E6ithas themttn--shy

expected shp n epan hig prp in of thokvariance in saa~raies

~Jdamong individoals in terms of the idpnetvariables~used in the

euaions Based on the resul~ts of oretmating equations we can mae

- several observations about the pewa structure4

1T e po itiv and significan tcoef ficient on the expe ren e varia bles

inEutos1 2 and 4 show sarising between3an e

) year- ofOexperience~ InEuton43 the inclusooof a sopara e ae

is~big~ycor~aed wit exper ence expla na t

ucdred4 snificane of the experieaice variableshy

2 Thngtiemefiins onFPE have the expected sin ne ms

of the human capital modl4 butther low r~ognitudo and stat a al

Lnpinficance lead ustobeli veoat he alarya ii o hf s no a vvd t h peakedsucreoam ensn

Yeevo ed t a atur e

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

15

3 The ingeniero agronomo degree is the mninimum training required to

become an agmrcultuval researcher While the lncatura in science

or socia scincee represt C th ia e p rrid uI t i s as the

lflueniero i~Je~t to associ wlkt WthAppedu be in wf1ri lower

4 The masters degree has a strongly positive imact on salay (between

18 and 367) and its effect is statstically significant as shown in

Equations 1 2 and 4

5 he variable SHUGTOURSE measured participation in some form of

non-degree (usually short-term) t raini g after completing formal

train rig It doe not hav a - nficant effect on income and may

underline the fat that howeve ICful they may be short courses do

[lot 3ee tO IlTilPOVe2 Oe CC I

6 The DI[ECT riabl is straIj 1 positive and highly significant

wherever it ham been inclun r the estimate n Equations 12 and

3 however where the r- ind MS variabl- both ipea r the

estimated vetuin to th S 3qr falls This n-cates that there

ia corre l1t1n bt- een iced legre and t e i1oldin of

acidm In strItlve o l n bhold be lit ilt-ited tLl~tiher On

the othr halnd h mr iby It AX ADO Alan - i to classify

ipeople 4h0 - - h i ectioniareWSf n s or

prog le i I iecited with hijher incomes BothNein n

the magtnitiid of the W05-t 1 1lt ml its itit[stical i gnficanc are

Jma l1

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

agea~ rch ei o a n sss a n r I

C0tl e 0n epas _a o b 0 a 3 o ~ficet 6t

Wmaefet1 if~zta ~j

Ai I A rn to tA a e te a i 3o~ so sa rs

th ex e ugtA

es An1 e r4ajn a j 77

u~cijn ~he a1si fwo~The d~ imbabn cPje UZo neere

PacInertQy urve ae a C)e uses opreLg by andA the caag the

The daeausedjo o as eusincame fro a I~ithe Zimabwe I

Inventory sv S AR (see lni on J by hyPregaree by he po

Dr ttfResearcher Aa Sp~~~ func 8n to 5bhveand earis S efl

~ aoui f ar T funane~ tdeoeff c XEa nrOf

unvrst AdUmB anb a ica1-7 EXPRSQ degree the Poi an

resarher ad ex Agai thiedearnings and ehaedng

p~oshigh e5 nts on4~iur

Principai ileva are ofpce h iue a n 6 a Istic

sgiiatA They sumorzeare bellt inTbeA~tea i

~ 0

Prfie ar honil g r 1

A ~

44AA lt4I AIAA 4 A

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

17

Table 2 Earnings Functions for Zimbabwe

Variable Equation I Equation 2 Equation 3 Equatio 4 (n = 174) (n 1174) (n = 174) (n 174)

CONSTNT 856 85G 856 737 (1966) (1465) (1485) (628)k

EXPER 0056 0056 0054 0056 (927) (920) B891) (1066)

EYPERSQ -00011 -00012 -00011 -0001i (62)) (622)k (557) (680)k

BS 0279 0283 0278 (70)8) (706) (702)1

MS 0426 0423 0413 (888) i890)k (869)

PhD 060 0635 0664 (729) (739) (761)

YRSCHOOL 00847 (1363)

DIRECTOR 0428 (246)

PROCRO -0199 (124)

LOCATION 0016 (0484)

FEMALE 00048 0028

(010) (069)

058 058 060 067

Sec notes to Table 1

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Figure

4

-

_ isJ(

J -IJI 1 1 - IrC )(ri( Iri iamp

ii -H

-

84 -

853

82-

8

74

74

Ashy

-

0 2

Elyc I

6 8

t

0 12 14 It f8

Expe n-c iin E2

20 22

YeCL Eq 3

2-1 26 28 30 32

- 1

31

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

19

We note the following results

1 The large coefficient on XPER is indicative of an income profile

whilh is steeply sloped and proviles rapid increasns in salaries in

the early vears

2 The neqatiVe coOMft ent on EXP PltSQ is also highly significant and

rndicateb a norma l patterin of ear ins functlon with diminishing

increages as experience if accumulate

3 Comparing the coefficienuts ar the BS MS an PhD variables ae find

that the sttirattrg equations reflect what one wotd he eve to be

the case i PhD eiin reot than a masters who in turn earinc more

than a bachelors Allioefftcincuts are nigh ly significant

4 In Equation 4 where we used eatrs of Skhooling rather than degree

diunmies the return to a year of schooling is high and statistically

significant it is algo worth noting that the a rnings profile has

the same shape Is the ones estimated using dursmny variables for the

nducaton varitable

5 The DIRECTOR variable shows up as highly posit ye and significant

wi thout taking away explanatory power from the education variables

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Q~ 03

ofacl t on -ene vai

onatioa a gais I wpomen d JI

7 p iia n a tS~ j~ -hi LCTO wss ede cedo not sWea~4er~fon

the ~yiteoai~ntainat t 9epi sopc to~zthsy t

capitald~aFl reare_

radri seAin sa a 4hsintheea r er of one - anduraree

sucesiel higher~azuaiasoywn acrent

Ariutua e arcwthnt h Deatmn~ r ica) pra D i Thiad~ muc longer wihu uptidn taehistoa~ry yjL mig-

lnac toshcb saigt a teye- omthi

~runas t hi humnresorce attern aproac as ledo ba 1 ak-p red

betrke SSand pho centias fha itwants

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

2Vo~ e~q p

t a A1 198 6i C upi a

una erng aatu oc dItion fse rrsac

cond im~ ncl vi seviea ra p it orm - - ee I fa~~ an9 w 1ed n~ u i0 O or that

(Man 6ue~ aabjna ta oruta cpIcda rfd

raeoK1e y PIy ~~~~~~m

weClecere eprdonay ofama re y ~

puroethhan t ra fo b ftee

of he agesrcueuon nydta ffroedeI a4 n istatiethew~a~c og

(Mny sysem iR Latin and cou-rzeeod UCminann Amric

thteAc rgncameproaccessil~9 forpo icy ana s a oii

Thnmeag feerhe~ nheaDeatA gttoA i~tr alosu oetmt annsfntosa oedsgrg dIeeI

tha is-posil nEuto peeinsalrsses 1fal3w

the ~ f-ebsc ua aia nwihtelreut oe

saayi e~mtd safn-too EPREXES lCwn

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

22

Table 3 Earnings Functions for Total Thai Sample

Variable Equation 1 Equation I Equaticn 3 (n 82) (n z 882) (n =882)

CONSTAIT 8017 803 800 (43601)h (41356) (39006)

E 0077 0077 0077 (2472) (2470) (2484)

EYPERSQ -00013 -00013 00013 (1105) (1093) (1119)

MS 0054 00c2 0045 (44) (423) (363)

PhD 0222 0227 0221 (568) (582) (571)

SEX -0023 -0016 (213) (149)

BAGKOK - 0048

(4I0)

R- 0762 0763 0767

See notes to Table 1

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Figure 5 7I_f ILA NIV) - kcp r~ oc - Incirmwic -f k

85 -

84-shy

81 -1 ----shy

8 --- - - -- ------------ - ---shy - -- - -- - - - - ----- - -F - -

1 2 4 8 0 1 14 6 18 20 2 24 2-1 28 W 12 14

L+ El J E 2 Ecq 3

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

24

We observe

1 The earnings functions are all well-behaved and exhibit a relativele

steep slope in the early years and a peak coming between 25 and 26

years of --p Len

2 The earning of a aners iegcee app t be ascociated with a

salary betgtern 5 nd 6 hignr- than the averaqe for all researchers

while the PhD is associated With a 22 qain a deron strated by the

positie and significant coefficients on the degree imus es in

Equat on I

3 When the SEX dunmy is added to the estimating equation (Equation 2)

we find that males appear to earn about 2 less than females

then OU d(ryAT Equation

sigiificance of the 12dbIJIV fh 1 S ead I ng us to he Ii eve that the

reason for ipp-iit lowr ies I es the

I Hoeer the it i included as in 3 the

the s rcOr a is

concenitrati on of wuwn in Balzlkh wh-re sa a ri are between 4 and

51 he (Jr thai1 the aerg

In Equations 1-3 we estimated the earnings function across the total

sample of Thai rc-archer using huruny va Vrtab 1es for e degree level

ind location Th-i i ina th wac thie o I f tot e

rIMOabL n th- - I Lt i1 t ii+ i vtv t-- iag il IifA

[ tpwa Orr [ ilo -I 15s oi-cl ior-i it Jt-

lage wq a r iYi to -25t irnat -rt-nir s ftunc t ion at a maunh finer level

)t l altatin and till obtain statimtically significan results In

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

EjKu~Jlaquon A-- r t ba a orweed srsar

dhe~r mates an ts I( -lav 1gBs y) es a rieee

afto to dehre ever~sexp1d

~lt z~etiltsaen irsn olt~estb p~~gartedkin able ca hecr ~ saggeg~1~ gue6r gt N

Y Taale of the arnngsproie

4 1 4~ ~ gtj ~ W

L ~ ln Equation 4- we haaroe th-t-- a sapeAn o hews

sampesfyleelhihes egee~erng Thlg~o on~ y

eima te a afucto of BSPR EXE S SEX -nd LOAIO

reut are presntedin Tale64 foll e b39the corson g fann18

Equatio 4 E q u 8i5 Eg iaion

1 332 6 )(84 5 1

EXPES9~5 VP _000098~ ~ -000034gt

MPRS - 018

SEX 002 009 0J126 (160) I (028)4 0079)4

LOCATIONN IVIV 00564 -0 159

4 V (4 ~J04 1 1) 4 0 )

074 1 9

Seen0717abl1

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Figure 6 ffY 11L l) kxpc~fvc~c -If~cr)mnc Irrof4cs

bull--- shy - shy -t-- -

858

81

4

JJ

2 4ll 6

E q

L][

4

4L5q 4

0x -

in

32 4

U i 4-- Ar x pound

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

~~~u~ ino~o~~~ i ~ ~ S and 0-ps On gt e6pectocai ea ~a dvanc ge

would4 th~soppof te-anig Eunction anidpesepe

rapid adyancmn thro th~ sytm ~Tidoesno tappea r to eh

2There is no significan difeence -in saariis 6 1en and women as

ampMostrted ythe 16 maudeand statstica1-I _e0

t he sex dwny~~l

3 Fo sc-4eztists with only a BS degree there~aears tle

Kstatistically sgnificant saardat eof the mn dle f 5-6-

I

no apart be siniicnta the Sand Ph levels - perha ps

beas-sinisswt-tee ere 1amporkig out f 0thcapitalh

4 beqn posted tip-cointry to positions of atoiy

r4 Although we proeent~the e i for le hhe~hrruemner o f

sinfiac to the coefficients~

~ V iag~gto b craig eaaeSmlsfo- me F_l -V()e

and MS levels 1heresult C sta ma t 51s n h 3~~~~~~~ a rNPRQ- hd OAINVrp

al-i a fucion o EPE ~ e

Tabe he irannspoi1s FiO5wih coresodi 1plusmn

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

28

Table 5 Earnings Functions for Thai Researchers by Sex_ and Degree

Variable EqUation 7 Equation 8 Equation 3 Equation 10 LISM4OS WomBn Ms Mln MS Women (n 369) ii -255) (n -- 115) (a shy 124)

CONSTAIT 7987 788 8 2 22 (25609) (25252) (12483) (13511)0

YE 0073 0104 0074 0037 ( l6)b (171l)- (83)k (357)

Y[ 1 -000115 -00024 -00013 +000009 (651) (R95)k W448) (021)

LOCAT ION 0068 (0335 --000 010t 317)k (189) (-025) (287)

0692 0837 0696 0753

See notes to Table 1

We observe

1 At the lvel of the BS the earnings of men seem to peak later in

their caree r than those of women inrcating perhaps slightly better

promotion prnOerA t or mn In y ears past

2 At th- level of the BS men seem to benefit more than women from a

Bangkok oti ng ii hown by the hirjher and more significant

coeffii ant on theLCATI ON dIummy in the male equation than in the

femae equation

3 We ohserve aqain the m la rity be tuIen tLe enrnings prifiles of ma e

cIentit with the BS md male scientists with the MS Ths s

cons i t~eat with thel Impreson that there i 7uteimat ic promotion

through the stem

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

Figure 7 77 ITILI J) Ex c4 cc - lIcn irofi

9 4 L4 shy95 shy

9-2 - s

91 shy

89-

2- -- 2

a S -

14 -r

S )( L

A

Lt Eq 7 8A 9E 1

79 shy

0 2 -4 t 8 10 12 11 t 18 20 22 2-4 26 213 30 2 34

E Et 7 t Akj 8 x Ey 9 rj 10

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

9 ~ aI~e e

4W fH noL a foyoen hev~I h64r g- rfil

P7fpe n eal Lareer_ the ea a og

n ~we would4invetqt

iomen MiS z3c3entis ts i vevkth re are no t at shy

fo w reiable ~r is n9portIon

fucton Seod-here m~ayj be imotn persoaal characcrsIE

of theilr attahett h o ri thei raccess to hiher edcatin

gtand rpromotion

~~V

Theprein oe o an S ofanlyis otpretend be exdaustieanays

the4 conditions of service in the Thai Departinen trof Agriculture- Iti

~simply~attempts to demostrate that one can use read yva va1ab 1

~administraive data to~identify impctant~aspects thei conii6in

sevc which merit in-dept examninato

thFro hg ecentace of Eh a arie xpained4 h

sy bemticll rel~ate t iwenqt e h andthe ecc~l) ir4 e this~ ~ ~Aste -gs-~alse~ysemicueiorsu te

a xtrsand Ceward structurenima eaesalizd o-tfr1 0 ha~pr re~d

44

OtQ Sd kdkO e no L~g~J a sr o d

th qvernE and possib a pl~y o trainedpopeisecomes aa bl

~outsd- govrmn0sw ma c o deI the desira01 li

payin premiul salaries to MS and PhD sntistIs rre r i ed d ecEl y and

r m its own- expenditure on t~k~

ISAFDEVELOPMENT AND THE EANNSFUNCTZV_

An assf~ th~e returns to the individual of earning-an advaniced-

degee s ueful way of predicting -wethOr not the sytr-w be-

~abet etini for research$ the scientist a whohaebntrne

governmnt4 odnrs th ytm i yf ioIno the tuIdet earnings

give here high earnvgJo PhDevdene tat are returns to a n

degrees ~gt~ 4 J

In44 i of the systems the earning of an advacd er8~S a3SOc1msI

with inrae bu bohth ih degreeand higq~ r) a ar re

Of ter accomnanied by apointmnt to someadmnistratve fucton 1 1

i~~~rina1~~~ most-qualimteepo t i

____r

ta1~f

Losirtions and perhaps all we cansa It hat a m ovrefEor i b4e

required i ann ocet h ltica

nis imesearch m e hs s~ active infu 1- -Wenhat_

s 7s iIIhv~e evard ue~~vwi e am o look at stu encoag 0eo

reao n-neearc menterp i acemen E of a badi iita vs

Of J5hem TCOharninpa nci can

1 haa lys

32

eve ca th 1113ta

-vanaeoust hav o Le I Elped exn su1t ta a c~reeoe i~1 yg - io act$eco eep~a e arigsfn

availbei th coun th m fatea

uImL1 the syst i radsto he anfdp~ AsISNA tilla~ e the LDeopIenpe retan e~o

ill b1eable t elt he sap o 0tothe t ate oshy~theearnings functon

development of~hunan resoein thedcutry and have a way of making

-~cross-coun~try coprsn ndscuin~iig reiard Bructures

V CONCLUDING OBSERVATIONS

-The uposeof thivs paper w(as to tes a me thodo logy for analyzing readi

available administrative data (U dta collecte v ~arelatively 1pamp

qusionir)togneae-nigt into te character and appropriatenessof a1 systems rewar~d structure~ u~p dlih

eaiig ca itaXlfn baedhe n fuctonofers a we can appraring means -by swhic

~the question of conditions of service in a national Ssse It 0GBoeS6t

preclude the need~to study he institutional processz-dir ~y bL al BU~~t~ X~U yoth s ao A ii~ cond plusmn0 B 0uhoaWay I

seviepeat ro hecases presented-he it in v1 ent t batrqdfe q~al amn countis -B~y Wni er i Iknq c I~s

across a ra o~efcountveISA wi p ernh

Foy be apli cle to particulVar reg mns of h wo6i Q aleativeI

c~ysem a dfferent esatgs in eve lo mn a eZ-n s3

33

of the return to advanced training and tto way in which this return is

gained we will be able to relate the rewad str-cture to human resource

duvelent plans

The UrFcnt documnent is a wocvir pd per foc citical coar-rt and

suqgrtions for impr muynLat Tho 3imale atlnu funutlsm11 est iated

here can etti11 he i111evad throurh th dW It of ttr oxilartatoriy

var iabls or thIlroiqh ai r - oa t tLi patir - i w[ wihuch

rtepartiva thq~A tit n to01 IA oll l lr niut~

ieterMinat 1t of Ir o Lel ite hun- tht ong in its present

torm wI have Iiront-atod n ti mltho prmit a 0111 ta y powerful

ana lysis of reward ttu-tt wth a MeIat 0ey sMaR1 nestnent in data

nol[qction and imilyA n appropciat- form oftw Le most gqneral

the e i irnhart been elop

32

eve ca th 1113ta

-vanaeoust hav o Le I Elped exn su1t ta a c~reeoe i~1 yg - io act$eco eep~a e arigsfn

availbei th coun th m fatea

uImL1 the syst i radsto he anfdp~ AsISNA tilla~ e the LDeopIenpe retan e~o

ill b1eable t elt he sap o 0tothe t ate oshy~theearnings functon

development of~hunan resoein thedcutry and have a way of making

-~cross-coun~try coprsn ndscuin~iig reiard Bructures

V CONCLUDING OBSERVATIONS

-The uposeof thivs paper w(as to tes a me thodo logy for analyzing readi

available administrative data (U dta collecte v ~arelatively 1pamp

qusionir)togneae-nigt into te character and appropriatenessof a1 systems rewar~d structure~ u~p dlih

eaiig ca itaXlfn baedhe n fuctonofers a we can appraring means -by swhic

~the question of conditions of service in a national Ssse It 0GBoeS6t

preclude the need~to study he institutional processz-dir ~y bL al BU~~t~ X~U yoth s ao A ii~ cond plusmn0 B 0uhoaWay I

seviepeat ro hecases presented-he it in v1 ent t batrqdfe q~al amn countis -B~y Wni er i Iknq c I~s

across a ra o~efcountveISA wi p ernh

Foy be apli cle to particulVar reg mns of h wo6i Q aleativeI

c~ysem a dfferent esatgs in eve lo mn a eZ-n s3

33

of the return to advanced training and tto way in which this return is

gained we will be able to relate the rewad str-cture to human resource

duvelent plans

The UrFcnt documnent is a wocvir pd per foc citical coar-rt and

suqgrtions for impr muynLat Tho 3imale atlnu funutlsm11 est iated

here can etti11 he i111evad throurh th dW It of ttr oxilartatoriy

var iabls or thIlroiqh ai r - oa t tLi patir - i w[ wihuch

rtepartiva thq~A tit n to01 IA oll l lr niut~

ieterMinat 1t of Ir o Lel ite hun- tht ong in its present

torm wI have Iiront-atod n ti mltho prmit a 0111 ta y powerful

ana lysis of reward ttu-tt wth a MeIat 0ey sMaR1 nestnent in data

nol[qction and imilyA n appropciat- form oftw Le most gqneral

the e i irnhart been elop

33

of the return to advanced training and tto way in which this return is

gained we will be able to relate the rewad str-cture to human resource

duvelent plans

The UrFcnt documnent is a wocvir pd per foc citical coar-rt and

suqgrtions for impr muynLat Tho 3imale atlnu funutlsm11 est iated

here can etti11 he i111evad throurh th dW It of ttr oxilartatoriy

var iabls or thIlroiqh ai r - oa t tLi patir - i w[ wihuch

rtepartiva thq~A tit n to01 IA oll l lr niut~

ieterMinat 1t of Ir o Lel ite hun- tht ong in its present

torm wI have Iiront-atod n ti mltho prmit a 0111 ta y powerful

ana lysis of reward ttu-tt wth a MeIat 0ey sMaR1 nestnent in data

nol[qction and imilyA n appropciat- form oftw Le most gqneral

the e i irnhart been elop

I LiII r tic~ nterretaton

mret eperince tesuare of years ofs erecedq nubei 0ol$

earninga reatad personal or job cha a cterisuch- occupation se

Theh~n capital approac consiers t an indi viualsa dic~it y

anearings are enancedbyinvesmenineducai ion ad h r

coeftiieai relating the natua1 an yersno

rscoaiof retur Linex~e -one-zca-

~ a~~ne~~ tif e dabyd sona

tbrin both theasimpe formn below and ina math a tca Vappend i extnsonso th odl

Let us define th raeo eunctefrs ero

_wneOY isth icome a pesonwoltdearni a terompetnq e

Ad at zonn she assumel oer-noe eiio aelnwo

onJ

35

This may also be expressed as

Y1 = (I + rj)X0

where inccme in Year I is the initial income increased by the rate of

return cn his investment of foLegone ePocn-ngs of while undergoing

tra ining

In r fashion the rate of return on the second education may be

expressed as

r = YZ - Y

7

3uch that

Yz = Y (i + r) = Xo (1 + ri) (1 + r-)

After S years of schooling earnings will be

Ys = X0 (1 + ri) (I r )(1 + r)

On the assumption that vr r r and that (1 + r) can be

pproximated as e

Y XYenn or

in Yr in X0 + c-S

e~~~~~~~ ~~~n oo A

cain)alary (XhL e of retur on)

Therefo Ve most~ orMUIionj cf e h m3 I Capital model include a

nolnQIeari4 y in the relation Ext~enslpj2i-fLhe model add a numn er

of pesoa or job characteristics which are believed rofaean effeq

S on earnings

4 - q

--~gtF ~ S

37

ANNEX 2

DescLiptive Statistics of Key Variables

1 Dominican Republic Survey of DIA Scienitistsa 1983

Unit Mean SD Min Max

Loq Sa lq 627 280 579 691 YearsE iuce 644 503 1 21 Ver Experience 6683 9813 1 441 Aq 3287 484 23 44

2 Zimbabwe Survey of DRSS Scientists 1984

Unit Mean SD Min Max

Log Salary 907 0333 89 993 Years ExperienceYears Expererce

868 1634

937 31533

i i

41 1681

Arle 3493 1004 21 63 Yeirs Education 1618 239 11 22

3 Thai1and Administ rative Data

38

from DOA 1983

_otal Thai Sample

Lfo Years Yeas Aqo

1 -Iry Exp rience igtxper~encamp

Mean

870 11418

16752 362

SD

032 597

15776 610

Min

776 1 1

23

Max

956 37

13659 59

Thai Women

Log Salary Years3 Experience Year Exerlelce Age

Mea

86876 107601

1-15 186 353881

SD

0303 512274

125309 563396

Mill

776 1 1

23

Max

939 24

576 50

Thai Men

Log Salary Years Eperience Year2- Experience Age

Mean

871 12002

183718

368591

SD

5338 )30084

175624

6 36094

Mm

792 1 1

24

Max

956 37

1369

59

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

35

This may also be expressed as

Y1 = (I + rj)X0

where inccme in Year I is the initial income increased by the rate of

return cn his investment of foLegone ePocn-ngs of while undergoing

tra ining

In r fashion the rate of return on the second education may be

expressed as

r = YZ - Y

7

3uch that

Yz = Y (i + r) = Xo (1 + ri) (1 + r-)

After S years of schooling earnings will be

Ys = X0 (1 + ri) (I r )(1 + r)

On the assumption that vr r r and that (1 + r) can be

pproximated as e

Y XYenn or

in Yr in X0 + c-S

e~~~~~~~ ~~~n oo A

cain)alary (XhL e of retur on)

Therefo Ve most~ orMUIionj cf e h m3 I Capital model include a

nolnQIeari4 y in the relation Ext~enslpj2i-fLhe model add a numn er

of pesoa or job characteristics which are believed rofaean effeq

S on earnings

4 - q

--~gtF ~ S

37

ANNEX 2

DescLiptive Statistics of Key Variables

1 Dominican Republic Survey of DIA Scienitistsa 1983

Unit Mean SD Min Max

Loq Sa lq 627 280 579 691 YearsE iuce 644 503 1 21 Ver Experience 6683 9813 1 441 Aq 3287 484 23 44

2 Zimbabwe Survey of DRSS Scientists 1984

Unit Mean SD Min Max

Log Salary 907 0333 89 993 Years ExperienceYears Expererce

868 1634

937 31533

i i

41 1681

Arle 3493 1004 21 63 Yeirs Education 1618 239 11 22

3 Thai1and Administ rative Data

38

from DOA 1983

_otal Thai Sample

Lfo Years Yeas Aqo

1 -Iry Exp rience igtxper~encamp

Mean

870 11418

16752 362

SD

032 597

15776 610

Min

776 1 1

23

Max

956 37

13659 59

Thai Women

Log Salary Years3 Experience Year Exerlelce Age

Mea

86876 107601

1-15 186 353881

SD

0303 512274

125309 563396

Mill

776 1 1

23

Max

939 24

576 50

Thai Men

Log Salary Years Eperience Year2- Experience Age

Mean

871 12002

183718

368591

SD

5338 )30084

175624

6 36094

Mm

792 1 1

24

Max

956 37

1369

59

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

e~~~~~~~ ~~~n oo A

cain)alary (XhL e of retur on)

Therefo Ve most~ orMUIionj cf e h m3 I Capital model include a

nolnQIeari4 y in the relation Ext~enslpj2i-fLhe model add a numn er

of pesoa or job characteristics which are believed rofaean effeq

S on earnings

4 - q

--~gtF ~ S

37

ANNEX 2

DescLiptive Statistics of Key Variables

1 Dominican Republic Survey of DIA Scienitistsa 1983

Unit Mean SD Min Max

Loq Sa lq 627 280 579 691 YearsE iuce 644 503 1 21 Ver Experience 6683 9813 1 441 Aq 3287 484 23 44

2 Zimbabwe Survey of DRSS Scientists 1984

Unit Mean SD Min Max

Log Salary 907 0333 89 993 Years ExperienceYears Expererce

868 1634

937 31533

i i

41 1681

Arle 3493 1004 21 63 Yeirs Education 1618 239 11 22

3 Thai1and Administ rative Data

38

from DOA 1983

_otal Thai Sample

Lfo Years Yeas Aqo

1 -Iry Exp rience igtxper~encamp

Mean

870 11418

16752 362

SD

032 597

15776 610

Min

776 1 1

23

Max

956 37

13659 59

Thai Women

Log Salary Years3 Experience Year Exerlelce Age

Mea

86876 107601

1-15 186 353881

SD

0303 512274

125309 563396

Mill

776 1 1

23

Max

939 24

576 50

Thai Men

Log Salary Years Eperience Year2- Experience Age

Mean

871 12002

183718

368591

SD

5338 )30084

175624

6 36094

Mm

792 1 1

24

Max

956 37

1369

59

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

37

ANNEX 2

DescLiptive Statistics of Key Variables

1 Dominican Republic Survey of DIA Scienitistsa 1983

Unit Mean SD Min Max

Loq Sa lq 627 280 579 691 YearsE iuce 644 503 1 21 Ver Experience 6683 9813 1 441 Aq 3287 484 23 44

2 Zimbabwe Survey of DRSS Scientists 1984

Unit Mean SD Min Max

Log Salary 907 0333 89 993 Years ExperienceYears Expererce

868 1634

937 31533

i i

41 1681

Arle 3493 1004 21 63 Yeirs Education 1618 239 11 22

3 Thai1and Administ rative Data

38

from DOA 1983

_otal Thai Sample

Lfo Years Yeas Aqo

1 -Iry Exp rience igtxper~encamp

Mean

870 11418

16752 362

SD

032 597

15776 610

Min

776 1 1

23

Max

956 37

13659 59

Thai Women

Log Salary Years3 Experience Year Exerlelce Age

Mea

86876 107601

1-15 186 353881

SD

0303 512274

125309 563396

Mill

776 1 1

23

Max

939 24

576 50

Thai Men

Log Salary Years Eperience Year2- Experience Age

Mean

871 12002

183718

368591

SD

5338 )30084

175624

6 36094

Mm

792 1 1

24

Max

956 37

1369

59

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

3 Thai1and Administ rative Data

38

from DOA 1983

_otal Thai Sample

Lfo Years Yeas Aqo

1 -Iry Exp rience igtxper~encamp

Mean

870 11418

16752 362

SD

032 597

15776 610

Min

776 1 1

23

Max

956 37

13659 59

Thai Women

Log Salary Years3 Experience Year Exerlelce Age

Mea

86876 107601

1-15 186 353881

SD

0303 512274

125309 563396

Mill

776 1 1

23

Max

939 24

576 50

Thai Men

Log Salary Years Eperience Year2- Experience Age

Mean

871 12002

183718

368591

SD

5338 )30084

175624

6 36094

Mm

792 1 1

24

Max

956 37

1369

59

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

MMM RESOURCE NVENTOR~i MAALYSIS

Cien Names4

a r i a1 Status Aiore

Single 4

~ MarriecdWidowed

S p a r at e d V

Nuie of dependn

io (bgi with highest degre o

I University ~A~

Name of tA~~Locatiorn Ye 3rs-At t n de d University ACityCuntry From To

ede

DegreeO Otie

Spec1iliation

A 3 AAp

4 ~

2 Shot d

4-

o-se

Pt

-

(les tha

A

9 months)-

AA

44

2~~~~lo1 n iih1 ~ ~ ~Sh11016rem 01

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

0

Dreoe pio r

4 Type of Wark Prf rm d b unicoate the pe~cntage ofyour time devote to

TOTA I 01 0

5 Ifthere is a formal scheme of serviie ilYy n st itu e Pleas indicterad Step

6 lRemuneration StructureiA 4I

a~ Current base salay14

c4~~bVlu ofllowaniug provided __or _________

1seiliainT~ ~ 1 d preiums for funcItin~ i4~4

~7 1P0 tion at entry Title ~ 11

Grade StepBase saari~y al entry

w~

outside of our curren ns UOrmiistry-o 0(13 o1 ee 8 Othe ork xper ernce~ Please li a period of proressional ernblo~i

o

Namie of~ rocatio Dates -bf P Sa Em yr To~ A11 cnrm 4--

I I~I k

N -A

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

GA pe 16i thagtof -ouini - era ig th ry ut ii t

raoDi-i

CoA ca

atta~ch adtoa pae-feesay

atos Pleas YQ pdca

ofpilication (bookjunlatcevsa hrp te6ber of pages adyear of publication (Atchx~ta pages i nc sr

ln- e m nat n- he r 31aI

bw 16 oun

kLY

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4

enJ- 1 iJo f

-Treatd Untvri Godya Pu antheMofDtrwda 1Laboi h a

Plan oor the Deatent frpm e5oear adC pali1985 Th aue The NeZ~t erandsI

Mazular Dpk Te-r~-L~ r~AM kersand

Mohnr-e Raksh eermiatbof Laourics Metropoli Est~~imtsfo BooaadClC

14

41-7

MI

stolombia o d Barmc Saw

3 Distri OreioIncom on

n ien Develoing om ia or an Saf

4