Trends in International Migration - OECD · sion and control policies. ... tion, administrative...

SOPEMI

TRENDS IN INTERNATIONAL MIGRATION

Continuous Reporting System on Migration

ANNUAL REPORT

1999 EDITION

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

229

Part III*

CLANDESTINE IMMIGRATION: ECONOMIC AND POLITICAL ISSUES1

Introduction

Considered over a long period of time (since the1970s), and against a backdrop of otherwise moder-ate growth in international migration (Tapinos andDelaunay, 1998), the increase in the variety of formsof clandestine residence and the growing length ofstay of clandestine immigrants have had the effect ofvirtually reducing the debate on immigration policyto one on the issue of clandestine immigration. Thecurbing of illegal inflows and of foreigners’ irregular-ity with regard to residence and employment hasbecome one of the priorities of migration policy. Thispolicy orientation is now common to all the countriesof Europe, particularly the new countries of immigra-tion in the south (i.e. Italy, Spain, Greece and Portu-gal), as well as to North America, although the UnitedStates and Canada are still open to regular immigra-tion.

To examine the economic and political chal-lenges deriving from clandestine migration is toinvestigate whether clandestine migration possessesspecific features as compared to regular migration. Itis from such a perspective that this paper considersthe problems of measurement, the economic dimen-sion and control policies.

A. ILLEGAL MIGRATION

1. Illegal migration: a multi-faceted concept

The extent of States’ sovereignty defines thescope of clandestine migration flows. The clandes-tine nature of such migration is defined by referenceto the rules of law (and their shortcomings), therestrictions on entering and leaving a country, and

the regulations governing access to the labour mar-ket. By definition, clandestine immigration eludesregistration and statistical coverage. The first diffi-culty confronting one is that of how to define andmeasure clandestine immigration.

Convention No. 143 adopted by the 1975 ILOConference defines clandestine or illegal migrationmovements as those where migrants find themselves“during their journey, on arrival or during their periodof residence and employment [in] conditions contra-vening relevant international multilateral or bilateralinstruments or agreements, or national laws or regu-lations” (Moulier Boutang, Garson and Silberman,1986). This definition places the stress on the diverseaspects of irregularity: entry, residence in the hostcountry and the undertaking of an occupation.

In a world without restrictions on entering orleaving a country, illegal immigration would be aconcept without reference. Unless there simulta-neously exists both restrictions and a degree of toler-ance illegal immigration cannot take place. It is amanifestation of an imbalance between the effec-tively unlimited supply of candidates for emigrationand the limitations placed by recipient countries onthe acceptance of new entrants.

In countries like the United States and Canadawhich remain essentially open to immigration, illegalimmigration is an alternative entry procedure forthose who do not meet the required criteria, for thosewho would have to wait longer than they would wish inorder to obtain an immigrant’s visa, as well as for thosefor whom unauthorised immigration is less expensive.Where opportunities for entering and staying in acountry are limited, as in Europe at the present time,illegal entry is, excepting being the subject of a family

* This special chapter was written by Mr. Georges Tapinos, Professor, Institut d’études politiques de Paris, Consultant tothe OECD Secretariat. The views expressed are those of the author and do not commit either the Organisation or thenational authorities.

OECD 1999

Trends in International Migration

230

reunion application or applying for asylum, the onlyoption available to candidates for migration. In thecases of both types of country, once there exists astatutory maximum length of stay, whether thisdepends on the type of visa (e.g. for tourists, studentsand temporary workers) or whether indeed there isany obligation to have a visa at all, overstaying2 placesthe person concerned outside the law.

This fundamental distinction is easier to draw inthe United States where there exists at entry a statu-tory dividing line between immigrants and non-immigrants, and where the statistical system esti-mates the number of non-immigrants by countingthose who do not amend their status after entry,whose departure from the country is not registered,and who are assumed to be still in the country afterthe expiry date of their visa.

In European countries, where no such distinc-tion is made on entry, overstaying usually occurswhen an immigrant is turned down after applying forthe obtention or renewal of a residence permit. Thissituation, the reflection of rules governing legal resi-dence, can result not only from wilfulness on the partof migrants but also from the ambiguity or incoher-ence of the rules themselves. For example, in coun-tries like Italy and Spain where there have beenfrequent regularisation programmes and where apermit obtained at the time of a regularisation exer-cise is normally limited to a period of one year, withthe possibility (non-automatic) for it to be renewed,repeated regularisations of the same people havebeen observed. By contrast, in the United States,those who benefited from the amnesty law of 1986[the Immigration Reform and Control Act (IRCA)]obtained permits enabling them to remain in thecountry until they met conditions for the obtention ofa permanent residence permit (Green Card).

Illegal immigration also has a time dimension asit can be defined by length of stay. From the immi-grant’s point of view, illegality may represent either atemporary phase in the migration cycle (as was thecase in France during the 1960s where there existedthe possibility for illegal immigrants to be regu-larised) or a permanent state (in the absence of reg-ularisation or when regularisation is exceptional andnon-renewable – which is what American regularisa-tion in 1986 and the regularisation programmesrecently undertaken in European countries areassumed to be). This distinction also holds from thepoint of view of the recipient country, which may beeither a country of first entry or one where the immi-grant intends to settle permanently. In this regard,

the situation develops over time. When Spain andItaly, emigration countries par excellence, began toaccept immigrants at a time when traditional Euro-pean immigration countries were closing their bor-ders to immigration, the authorities in thesecountries liked to think they were countries of firstentry. By the late 1980s and early 1990s, it hadbecome clear that these countries had in turnbecome settlement countries with a significant num-ber of clandestine immigrants.

The examination of clandestinity cannot bereduced to the clandestine migrant himself. Theanalysis of the phenomenon and the formulation ofpolicies to combat it should look at the whole lengthof the chain of clandestinity, which brings into play aseries of agents: the migrant, the intermediary whofacilitates his passage or placement, the company forwhom the migrant works and, according to the case,the principal contractor. One is therefore confrontedby a wide array of complex situations and circuitswhich range from the arrangement of the meetingbetween the clandestine immigrant and the employerthrough to far-reaching and powerful organisationswhich control a veritable trade in labour. In America,workers from Mexico or Central America endeavour tocross the border with the help of a smuggler, whomight also arrange their obtention of employment.Elsewhere, nationals of sub-Saharan Africa enterAlgeria by aeroplane, continue their journey by landthrough to the Spanish enclave of Melilla in Moroccoin the hope of being transported to Spain at the timeof a regularisation programme.

The organisations involved in smuggling opera-tions can be of considerable scale. In the United States,the INS dismantled in 1998 an organisation which hadarranged the passage to the United States of approxi-mately 10 000 people; in November of the same yeartwo million fake identification papers were seized inLos Angeles (INS, March 1999). The clandestine chainsometimes goes further back than the migrant himself.For example, the Moroccan youths who are known tohave entered Europe were not, as one might have sup-posed, unoccupied youths who had taken the initiativeto emigrate clandestinely, rather they had been sent bytheir parents who had elected to take advantage oftheir children’s minority which, in Europe, protectsthem against expulsion. This extreme heterogeneityshould rule out judgements on this phenomenon madeon purely humanitarian or, conversely, purely criminalbases.

These distinctions have considerable implica-tions for the measurement and characterisation of

OECD 1999

Clandestine Immigration: Economic and Political Issues

231

illegal migrants, the economic impact of illegal migra-tion, administrative effectiveness and the aspect ofcontrol policy.

2. Measurement methods

Estimating the number of people in an irregularsituation is inherently problematic. Account must betaken not only of the legal system of immigration, thesystem of statistical observation and mechanismsinvolved in undocumented living, but also, giventhat such estimates rely to varying degrees on surveyevidence, of the perception that the clandestinemigrants have of the acceptance of their clandestin-ity by the local population and of the nature of therisks that they are taking.

In that it concerns the measurement of an unob-servable event, the measurement of irregular migra-tion runs into well-known problems. The difficultiesare particularly severe in the case of clandestinemigration for whilst measurement in this field canonly be indirect, the absence of regularities in thephenomena that might be linked to this form ofmigration limits the usefulness of indirect methods.Indeed, the classic demographic distinction betweendirect and indirect methods is hard to sustain.Strictly speaking, one rarely has direct observationsat one’s disposal. At best, observations are incom-plete; in every case complementary information isrequired – which is itself often estimated – for thephenomenon to be satisfactorily understood.

Figures for apprehensions in border regions, forexample, have to be adjusted to take account of thenumber of times that individual migrants attempt tocross. Figures derived from regularisation pro-grammes need to be amended to take account ofthose excluded for failing to meet the criteria (partic-ularly those relating to the date of arrival, the obliga-tion to remain in the country continuously, and thepresentation of a work permit). Similarly, surveys thatfocus directly on the clandestine population (“snow-ball” sampling or the Delphi technique) which pre-sume to identify from the outset people who areknown to be in an irregular situation contain substan-tial biases. Measurements made in the frontier zonesof sending countries before migrants cross the fron-tier (cf. the Mexican COLEF survey) constitute anexception.

It is important first of all to know what one ismeasuring. Is it numbers (stocks) or movements(flows)? Is it workers or the whole of the population?To measure clandestine entries, is to attempt to assess

the volume of the flows of people entering the countrywithout the required legal administrative documen-tation, whether they cross the frontier with falsepapers or at a point that is not controlled. By con-trast, irregularity in respect of those who reside con-cerns stocks which comprise both the net cumulativeflows of people who have entered without authorisa-tion and are not regularised, and those who haveentered clandestinely but have lost their right of res-idence and are still in the country. It should also beconsidered as essential, in so far as the employmentstatus of those in an irregular situation will also, logi-cally, be irregular too, to attempt to measure thenumber of foreigners in irregular employment.

Defining the groups in these ways allows one todistinguish the possible estimates, those of: clandes-tine entrants (inflow of illegal migrants over a givenperiod), clandestine residence (the total foreign popula-tion residing in the country illegally at a givenmoment) and clandestine workers (the number of clan-destine workers in employment at a given moment).

Chart III.1. summarises the different entry, resi-dence and employment situations in which foreignmigrants can find themselves. It is possible to iden-tify six categories of clandestinity:

A. Migrants who have entered the countrylegally with a legal residence permit, but whoare working illegally either because the job isnot declared or because their residence per-mit does not allow them to work.

B. Migrants who have entered the countrylegally, who are living in the country illegally(either because their work permits are invalidor have expired, or because they do not haveresidence permits), and who are working ille-gally. It is assumed that a migrant without aresidence permit cannot work legally underthe legislation in force.

C. The same category as above, but coveringinactive migrants.

D. Migrants who have entered the country clan-destinely, who have no residence permit, andwho are working illegally.

E. The same category as above, but coveringinactive migrants.

F. Migrants who have entered the countryclandestinely, who have a residence permit(e.g. following regularisation, or by variation intheir status through marriage) and are workingillegally.

OECD 1999

Tre

nd

s in In

tern

atio

na

l Mig

ratio

n

232

OE

CD

1999

selves

legalntity papers

tine passage of inspection

Legal residence– Regularisation– Adjustment of residence

status (through marriage...)

F

employmentlared activityidence permitot allow to work

Inactivepopulation

Legalemployment(work permit

in order)

selves

legalntity papers




F


Inactivepopulation


in order)

selves

legalntity papers




F


Inactivepopulation


in order)

Chart III.1. Presentation of the various situations in which foreign migrants can find them

ENTRIES

ENTRY

RESIDENCE

EMPLOYMENT

Administrativelylegal

(papers in order)

Il– false ide– clandes– absence

Legal residence(residence permit in order)

Illegal residence– invalid or out-of-date residence permit– lack of residence permit

Illegal residence(lack of residence permit)

A C B D E

Illegalemployment(undeclared

activity)

Inactivepopulation

Illegal employment– undeclared activity– the residence permit

does not allow to work

Inactivepopulation


in order)


activity)

Inactivepopulation

Illegal– undec– the res

does n


ENTRIES

ENTRY

RESIDENCE

EMPLOYMENT


(papers in order)





A C B D E


activity)

Inactivepopulation



Inactivepopulation


in order)


activity)

Inactivepopulation


does n


ENTRIES

ENTRY

RESIDENCE

EMPLOYMENT


(papers in order)





A C B D E


activity)

Inactivepopulation



Inactivepopulation


in order)


activity)

Inactivepopulation


does n


233

The problem is how to go from administrativestatistics, which are marked by characteristics spe-cific to each country, to a method of measuring thephenomenon.3 Most methods rely on data from therecipient country, and they are based on two kinds oflogic with the estimate of the stock of clandestineimmigrants made either by comparing statistics onentry and modifications to status with data relating tostocks, or by using comprehensive registers of demo-graphic events such as births and deaths.

Methods of estimating the clandestine popula-tion were first developed and applied in the UnitedStates. Direct calculation, through comparing entrystatistics and modifications to status from INS (Immi-gration and Naturalization Service) data, permittedthe inference of the number of visa overstayers. Tothis component of the stock of clandestine immi-grants should be added clandestine entrants (entrieswithout inspection) estimated by means of an indirectprocedure (Warren, 1994).

“Residual” estimating is based on census data.From the total foreign-born population counted at thetime of the census, 1990 say, and the theoretical legalforeign-born population estimated for the same pointin time from INS figures, an estimate of the popula-tion of clandestine immigrants picked-up by the cen-sus can be obtained by subtracting one from theother. To be of analytical interest, the censuseswould have had to have counted a high proportion ofthe illegal migrants (Clark, Passel, Zimmermann andFix, 1995). Mention should also be made of estimatesof clandestine entries arrived at by adjusting thenumber of arrests on the basis of the likelihood ofbeing apprehended at the United States-Mexico bor-der (Espenshade, 1995).

In Europe, measurement of clandestine migra-tion has been attempted only relatively recently. Thewide range of methods contemplated and used is areflection of the similarly wide range of statisticalobservation systems employed but is above all tes-tament to the exploratory nature of these firstattempts. Tables III.1 and III.2 summarise the resultsof comparative research carried out by Eurostat insome ten European countries; its objectives were tocompile a register of methods used in the variouscountries and to suggest viable procedures in the lightof the kind of data available (Delaunay and Tapinos,1998). In practice, the measurement is based ondirect observation of the individual in a an irregularsituation or considered likely to be so, either at themoment of his crossing the border (for example dataon apprehensions and forced departures), during his

residence (for example, from regularisations andrefusals of asylum). Such information constitutes thebasis for only an imperfect measure. This is whyrecourse to indirect estimation methods is neces-sary.

Estimates based on the population of sendingcountries are really an application of classicalexpected-population and gender ratio methodsintegrating additional information derived fromdata from immigration countries. The “expectedpopulation” method involves a comparison betweenexpected and observed stocks broken down by ageand gender in two successive censuses. Starting fromthe population distribution on that date (t), andallowing for deaths and births registered or esti-mated during the period (t, t + k), the differencebetween the population actually observed and thatextrapolated in (t + k) gives an estimate of the send-ing country’s net migration during the inter-censusperiod. Evaluating clandestine migration involvescomparing this net migration estimate with variationsin stocks (or in net migration) registered world-widefor nationals of these countries which necessitatesknowledge of the stocks in recipient countries of thenationals of the country concerned. Such a method isappropriate only under the hypothesis that thenationals of the country concerned migrate to a verysmall number of countries.

The second method focuses on the high propor-tion of men among clandestine migrants; this isobserved by default in gender ratios (women appearto be in a majority) of populations counted in send-ing countries. An estimate of those missing is thenpossible if one knows the ratios of men among thoseremaining and among all nationals registered both intheir own countries and abroad. These indirect meth-ods are not reliable. A small weakness in one of thelinks of the estimating chain (e.g. in census coverage,or in measuring the death rate) can easily lead to asubstantial variation in the measure of clandestinity.Furthermore, they only provide an overall estimateof the number nationals who are living abroad; theirdistribution by recipient country would remainunknown.

In so far as clandestine immigrants work in thehidden economy, it will be possible to consider anestimate of clandestine immigration as a by-productof a measurement of the hidden economy,4 definedas those activities which evade payment of tax andsocial security contributions. This approach has twolimitations: firstly, with a few exceptions, clandestinemigrant workers account for only a small proportion

OECD 1999


234

of those clandestinely employed; secondly, themethods used to measure the hidden economyreflect concerns linked to taxation and national sta-tistics (e.g. the amount of, and growth in, GDP), andare not directly concerned with the legal status ofmigrant workers. Moreover, in the hidden economythe dividing line between clandestine employmentand registered employment tends to be very difficultto draw (see below).

The most elaborate attempts to estimate thenumber of clandestine foreigners by measuring thehidden economy have been made in Italy. They haverelied on observed divergences in the nationalaccounts between declared added value and esti-mates of added value based on average productivitythroughout a given sector and anomalies observed inchanges in participation rates.

Three concluding remarks on these measure-ment methods:

It is not possible to place on the same plane, onthe one hand, the innumerable guestimates (guessestimates) that tend rather to reflect the feeling ofanxiety (or the need to give reassurance) that theirauthors feel with regard to clandestine immigrationand, on the other hand, the difficult and imperfect

attempts, conducted in conformity with the proce-dures of scientific research, to arrive at a statisticalmeasurement. Most certainly, there is no guaranteethat the measurements arrived at on the latter basiswill more closely correspond to reality than thosebased on the former approach. However, that theirmethod of calculation, the field covered, the hypoth-eses used and statistical biases are all explicit doespresent a considerable advantage. Such anapproach, in offering the possibility of subjecting tocomparison the results obtained through a variety ofstatistical methods, is the sole means of coming closeto an accurate estimate.

The issue is not so much the number of clandes-tine immigrants but rather the characteristics and themechanisms of the reproduction of clandestinity:these give real meaning to the estimate of the stockobtained. The stock of clandestine immigrantsresults from the difference between their inflow andoutflow and depends therefore on the averagelength of time spent in the state of clandestinity. It isof some importance to know the relative extents towhich a change in the stock of clandestine immi-grants is due to a change in inflows and a change inthe average duration of clandestine stay.5

Table III.1. Typology of the methods used to estimate clandestinity

Direct measures

Administrative statistics

Statistics on refusals

Visas/entries 1Political asylum 2Residence and work permits 3

Infractions

Entry/border 4Stay 5Work or job 6

Regularisations

Mass 7Exceptionnal 8

Surveys

Without sample Delphi “method” 9

With sample

“Snowball” technique 10At the time of the regularisation 11On employment 12On mobility/border flows 13Biographical 14

Indirect estimates

Comparison of sources

Expected populations 15Pairing of files 16

Inferences from secondary events

Sex-ratio 17Common law crimes 18Births and deaths 19Statistics on school attendance 20Social/health benefits 21

Work statistics National data 22

Source: Delaunay, Tapinos, 1998.

OECD 1999


235

Irregular migration is by definition a breach ofimmigration law, and we have long pretended that itwas an exceptional phenomenon. This allowed us tobelieve that we were dealing with a problem to whichwe had to find a solution. This confusion betweennormative and positive had the effect of concealingthe need to put in place a statistical apparatus thatwould provide for both measurement and the carry-ing out of longitudinal surveys with a view to under-standing the mechanisms of migration and thecharacteristics of clandestine migrants. Clandestineimmigration is not an exceptional, non-renewableevent. It is an inevitable phenomenon, one that islikely to grow as globalisation, economic transitionand inter-ethnic conflict continue.

B. THE ECONOMIC IMPACT

Is there an economic dimension that is specificto illegal migration as compared to legal migration?To what extent are the economic implications of ille-

gal labour immigration linked to the foreign workers’legal status?

In the standard model of the labour market, theincidence of immigration, legal or illegal, depends onthe nature of the migration process and the degree towhich immigrants and nationals complement or sub-stitute one another. In this way, one may contrast thesituation according to which immigration derivesfrom an economy-wide or sectoral labour shortageand may therefore be seen as endogenous, with thatwhere immigration is exogenous and likely to engen-der competition with the pre-existing domesticlabour supply. Under this latter hypothesis, whichstandard labour market theory assumes to be thecase, the equilibrium wage falls; the quantitativeimpact on the employment of nationals depends onthe (downward) elasticity in the supply of nationals’labour compared with the (upward) elasticity ofemployers’ total labour demand. A significant dis-placement effect is only observed where a lowerequilibrium wage rate leads to a substantial with-

Table III.2. Tested or concieved methods to estimate clandestinity

Method

Eu

rop

e

Un

ite

d K

ingd

om

Be

lgiu

m

Fra

nce

Gre

ece

Ne

the

rlan

ds

Ital

y

Po

rtu

gal

Cze

ck R

ep

.

Sw

itze

rlan

d

Entry/Exit

Visa refusals 1

Entry refusals 1

Passenger cards

Apprehensions at border 4

Biographical inquiries 14

Surveys on flows 13

Residence

Breaches of residence regulations 5

Refusals of residence permits 3

Refusals of political asylum 2

Regularisations 8

Common law crimes 18

Comparison of sources 16

Deaths and births 19

Expected populations 15

Sex-ratio 17

School, social assistance 21

Delphi Method 10

Surveys by stage 11

Work

National data 22

Employment infraction 6

Surveys on irregular employment 12

Methods being applied

Collaborators' proposals

OECD 1999


236

drawal by nationals from the labour market without asignificant increase in labour demand.

The empirical scope of such a prediction isdependent on the extent to which we can determinethe elasticity of the various segments of the work-force’s labour supply (e.g. with regard to age, gender,marital status and skills) and the nature of changes inlabour supply (e.g. variations in participation rates,hours worked, etc.). Another school of thought seeslabour market segmentation as a key factor underly-ing the resort to immigrant workers. Layard, Nickelland Jackman (1991) suggest, given the differing char-acteristics of jobs in the primary and secondary sec-tors, that a fall in wage rates in the primary sectorcauses labour demand to rise, whereas a fall in wagerates in the secondary sector causes labour supply tofall. This could be expected to have a positive impacton the employment of foreigners in the latter sector.

Could the anticipated effects be extended toirregular immigration? The highlighting of an effectthat is specific to irregular immigration or irregularemployment would necessitate an examination ofthe behaviour of labour supply and demand and theinstitutional variables that influence it (in particularthe existence of an informal sector) and the specificimpact of illegal immigration at a macroeconomiclevel with regard to resource allocation, distributionand taxation.

1. The behaviour of labour supply and demand

For the migrant, there can be advantages in clan-destine employment, but illegality is rarely the resultof choice.6 When a regularisation exercise is inprogress, the eagerness of most people who meet thecriteria to file an application shows that illegal workershave a marked preference for legal status.7 However,the argument that clandestine workers (who areassumed to be more interested in financial gain thanin working conditions) are prepared to accept longerworking hours needs qualification. In an attempt toisolate the sole impact of incentives and constraintson working hours, Dunn (1960) sought to assess theextent to which pay rates match the marginal rate ofsubstitution between income and leisure. To that end,he compared the results obtained from three samplesdistinguished by the existence and amount ofbonuses linked to long working hours and the fixedcosts linked to obtaining employment. It appearedthat in agriculture, the economic sector that comesclosest to the hypothesis of economic theory (i.e. theabsence of overtime bonuses, low fixed job-obtention

costs, plentiful information on wages, and competitionbetween employers), there was no differencebetween the working hours of legal and illegal work-ers. From this, the author concluded that preferencesin respect of working hours are independent of aworker’s legal status.

The advantages of illegal migration tend to beon the employers’ side. In that the migrant’s illegalstatus vis-à-vis his residence or employment placeshim in a situation of marked dependence renderinghim more willing to accept a very low wage, oftenbelow the legal minimum, the employer mightexpect to benefit. The threat of being caught weigh-ing on the migrant is for the employer a source of pro-tection against being penalised himself. Andconversely, legal residence status combined withsemi-clandestine employment status can representfor the employer an advantage in so far as they allowa part of the work undertaken by the migrant to goundeclared or for the migrant to be considered as anindependent worker. The monograph recently pub-lished by Iskander (1999) on the garment fabricationsector in Paris and the Parisian suburbs shows thatemployers’ preference for one or the other of thesesituations depends on whether the firm’s strategy isoriented more towards producing a quality productor to the minimisation of wages. There exists for theemployer, leaving aside the cost of sanctions againstunauthorised employment (this issue is examinedbelow), three reasons for preferring clandestine for-eign workers: differences in pay, differences in socialcharges, and flexibility in the production process.

For the employer to benefit from lower labourcosts, undocumented migrants have to accept wagesbelow the current market rate. Taking into account dif-ferentials in pay, standards of living and job opportu-nities in the sending and host countries, both legaland illegal migrants normally have a lower reservewage. Do clandestine migrants have a lower reservewage than documented migrants?

On the one hand, those embarking on clandes-tine emigration have to evaluate the expected costsof their illegality; these mainly comprise the cost ofentering the country (e.g. the likelihood of beingarrested in the course of each attempted entry, theaverage number of attempts made before a success-ful entry, brokerage fees) and similarly the expectedcost of being penalised for taking up clandestineemployment. On the other hand, the threat of beingcaught coupled with the desire, perhaps obligation,to pay back rapidly the monies spent on entering thecountry limit the incentive for prolonging job-search

OECD 1999


237

efforts. Clandestine immigrants are therefore inclinedto accept a lower rate of pay than nationals and regularimmigrants of the same skill category.

As far as social charges and non-wage costs areconcerned, ceteris paribus, the advantage for employ-ers in hiring undocumented workers increases in linewith the proportion of labour costs that would other-wise be accounted for by mandatory deductions andconformity with health and safety regulations. In thiscontext, we might talk of a “welfare magnet”, but in aquite different sense from how it is normally under-stood. Moreover, given the precariousness of his sit-uation and weak negotiating power, clandestinestatus also renders him more susceptible to discrim-inatory practices.

2. Undocumented migrant workers and the hidden economy

The structure of the labour market and the hid-den economy constitute for both the employer andthe worker the critical element in their decision mak-ing processes. It is by no means true to say that thehidden economy and clandestine foreign employ-ment cover one and the same domain, but equallythey are not unrelated given that the hidden econ-omy is the only way for foreigners in an illegal situa-tion and with no work permits to enter the labourmarket. Clandestine work by foreigners is the pointat which two domains converge: on one side, irregu-lar migration (which is located within the dynamic ofinternational migration) and, on the other, the hid-den economy (which is linked to changes in thelabour market, but cannot be simply reduced to thework of foreigners in an irregular situation) (Moulier-Boutang, 1991). For clandestine foreign workers, thesetwo domains are superimposed upon one another.Work carried out by foreigners in an irregular situationis therefore located in the overall dynamic of thelabour market, independent of the dynamics of inter-national migration. The emergence of a hiddeneconomy is part of a complex process of labour mar-ket hierarchisation and the bypassing regulationsdirected at responding to competition through achiev-ing greater flexibility in production and pay. It is alsoaccompanied by formal “desalarisation” and frequentmovement from employee status to non-employee sta-tus (from salaried employment to self-employment, forexample in the building industry), and the activation ofcommunity (i.e. family, ethnic and political) links inemployment relations. These strategies for managingwork are particularly important in highly seasonal jobs

that demand a rapid response to variations in demandand changing tastes.8

These ‘atypical’ jobs are performed by differentcategories of the labour force: young school-leavers andstudents, women looking for a second job, home helps,and foreigners in a precarious situation (whetherregular or otherwise in terms of residence status)are the most commonly employed on such a basis.9

Irregularly employed foreigners constitute one ofthe elements of the hidden economy; they are notthe cause of its existence. However, the existence ofa hidden economy that is broadly tolerated by soci-ety as a whole makes the recruitment of unautho-rised migrants more likely, particularly as networks ofmigrants make it easier to hire undocumented work-ers in this sector.10 However, though the recourse toclandestine foreign labour can regarded as an ele-ment of a flexibility strategy, were precariousness tobe extended to other segments of the labour forcethe relative advantage of hiring clandestine foreignlabour would diminish (Iskander, 1999).

Clandestine employment is also observedamong sole traders (such as door-to-door salesmenand people who sell goods in underground railwaystations). In such cases, clandestinity can be both anecessity and a choice for immigrants.11 It wouldappear that these activities are more the conse-quence of an exogenous inflow than a response to ademand for labour in the host country. That said, itcannot be denied in that the ‘African’ products soldby Senegalese street-sellers in Italy are in fact pro-duced locally within the informal sector.

There is an interrelationship between declaredand undeclared employment, that reflects both theprohibition/tolerance dialectic of control policy andthe complementarily/substitutability debate on thelabour market. Whether it involves legal activities thatbreach tax and employment legislation or illegal activ-ities (e.g. the sale of prohibited goods), the informalsector develops from the formal sector in the sameway as immigration is a consequence of the existenceof a system of legal immigration. The informal sectorresults from institutional rigidities in the formal sector(e.g. tax regulations and working time rules) and couldnot thrive unless tolerated by society.

3. Impact on the labour market. Empirical results

To what extent does the employment of clan-destine immigrants affect the employment andwages of nationals and legal immigrants? What dif-ferences in wages are observed between nationals,

OECD 1999


238

legal migrants and illegal migrants? For otherwiseidentical individuals, are these differences the out-come of market adjustments or an expression of dis-criminatory practices?

Ten years of scholarly debate on the subject inthe United States from the end of the 1970s through tothe end of the 1980s have highlighted the problemsone encounters when one attempts to isolate the spe-cific effect of migrants’ legal status and the difficultiesone meets when attempting to reconcile apparentlycontradictory results. One requires, in fact, data whichallow one to take into account self selection effects(age, sex, marital status, education, time spent in theUnited States, social capital, etc.), the self selection ofthose who return and also unobservable events (forexample, attitude towards risk). The kind of data thatwe have at our disposal rarely allow us to exploit thesophisticated statistical techniques that might elimi-nate these biases. The research undertaken by Massey(1987) on four Mexican communities observed theUnited States and in Mexico is exemplary in thisregard. It shows that the resulting status of inequalitydoes not of itself lead to a lower wage but does exertan indirect influence in the sense that it is associatedwith interrupted periods of residence and to shortertotal periods of stay. Experience in the United Statesis consequently limited and the workers tend to beconfined to badly paid jobs, factors which have theeffect of diminishing wages, be the persons legal orillegal. One of the earliest studies in the United Stateson the relative impact of illegal and legal immigrantson the wages of nationals shows that illegal migrantshave a smaller negative impact (Bean, Lowell andTaylor, 1988). The explanation for this outcome may liein the fact that most legal immigrants enter the countryirrespective of labour market conditions in the UnitedStates, whereas illegal migrants respond more to anendogenous demand for labour in that country. Otherstudies concluded that the results obtained should besubject to some qualification. On the whole, theresults are not clear-cut, and the impact, if there is one,is moderate.

In Europe, in the absence of appropriate datathe results have been more fragmentory and con-trasted.12

In addition to the methodological problemsalready mentioned, these divergences may beexplained by the heterogeneity of the forms of illegalmigration, differing contexts, as well as by the differ-ing dynamics underlying the process of unauthorisedmigration. In France during the 1960s, for example,when regularisation programmes were frequent,

enterprises had a two-tier strategy whereby theydirectly recruited irregular workers who were alreadyin the country whilst at the same time, as an insur-ance mechanism for times when they might findthemselves unable to recruit such labour, they alsowent through legal procedures to hire migrant work-ers (Tapinos, 1973).

For migrants, too, it was a short-cut comparedwith the legal recruitment system; it enabled them tocut down their waiting time and, in some cases, eventhe cost of migration. Illegal immigrants would acceptthe first job they were offered, and usually at the low-est rates of pay. As soon as they were regularised,they had little difficulty, at a time of substantialindustrial growth and labour shortages, in findingjobs that were sometimes better paid than those ofthe legal immigrants who continued to accept mini-mum wage rates.

The impact of a migrant’s legal status can also bemeasured indirectly by looking at how migrants’wages have changed following their regularisation.Figures are generally lacking, and where they existthey come from cross-country surveys that containbiases and are therefore difficult to interpret. Anexception to this has been the exploitation of twoAmerican surveys, the Legalised Population Survey(LPS) and the Current Population Survey (CPS) whichincorporate a longitudinal follow-up (Tienda &Singer, 1995). This analysis seeks to break down sal-ary variation since regularisation into what can beattributed to characteristics at entry that are specificto illegal workers (e.g. younger, more likely to comefrom non-English-speaking countries, lower standardof education, a shorter period of residence in theUnited States), and to factors that may have influ-enced increases in their wage in the United States(e.g. education, command of English, professionalexperience, changes in the labour market, entrycohort, the first sector of activity, economic indicatorsin the geographical area of activity, and changes inthe economic situation in the United States). Resultspoint to a substantial rise in the wage of workers fol-lowing regularisation, correlated with the year ofentry and the duration of residence, with major vari-ations according to country of origin.

4. The macroeconomic impact, distribution and taxation

What, then, is the total macroeconomic impactspecifically attributable to the employment of illegalforeign labour on distribution and taxation. Accord-

OECD 1999


239

ing to standard economic theory, immigrationreduces the equilibrium wage, increases totalemployment (i.e. nationals and immigrant workers)and alters the distribution of income between labourand capital. The greater the fall in wages rates due toimmigration, the higher is the surplus and return oncapital investment. Do variations in the amount ofsurplus and its distribution depend on the legal sta-tus of the migrant workers? Insofar as the effects ofillegal immigration are linked to a reduction in labourcosts and the equilibrium wage rate, illegal immigra-tion, by offering the possibility of labour costs lowerthan would be the case under legal migration, has cet-eris paribus a more pronounced positive effect on GDPthan legal migration.

However, it is also important to take account ofthe characteristics of migrants, their skills, the sectorsand branches in which they are employed, the condi-tions of supply and demand for nationals, and thetime horizon in question. For example, Djajic pre-sents an open economy model with full employmentin two sectors: unskilled workers, both nationals andunauthorised migrants, are employed in brancheswhose product is an input for branches where onlynationals are employed. The effects on wages andflexibility in a sector brought about by the employ-ment of undocumented workers increase the totalproduct and per capita product for the workforce as awhole. In the long term, the employment of illegalforeign workers may promote the acquisition of skillsand the upward mobility of the domestic labourforce, and thereby raise incomes for society as awhole (Djajic, 1997). However, results are linked tomigrants’ skill levels and to their exclusive concen-tration in a branch or sector, rather than to their ille-gal status. A structure of this type provides anequally good description of both legal and illegalmigration in Europe during the 1960s and early 1970s(Lutz, 1963; Tapinos, 1973).

By contrast, given the current situation in Europe,particularly in southern Europe where the laboursupply and unemployment are both high, substantialnon-wage labour costs in the formal sector and theshortage of capital can lead to a displacement of capi-tal and workers towards the low-cost informal sector.This is an undesirable consequence in the longterm (Dell’Aringa and Neri, 1989). It is a hypothesisthat needs checking; the detailed sectoral analy-ses required for this have yet to be carried out(cf. Venturini, 1997).

The most sensitive area concerns the effects ofundocumented migration on the most disadvan-

taged workers in society. Insofar as immigrationincreases the total product, specific groups affectedby immigration could receive compensation. Theminimum wage and other forms of guaranteedincome seek to meet this redistributive objective forthe most disadvantaged. However, such measurescan themselves encourage illegal immigration. Theargument is not that undocumented migrants areattracted by the “welfare magnet”, but rather that byraising the reserve wage of nationals, welfare-statepolicies increase the return for employers on the useof illegal immigrant workers and, to a lesser extent,on the use of legal immigrant workers.

It is precisely considerations of this type that areadduced in support of restrictive policies. However,restrictions on legal and illegal immigration that aimto protect the least skilled groups on the labour mar-ket have to be compared in an open economy witheffects on the same groups that would result from anequivalent increase in labour-intensive imports.Potential migrants unable to leave their countrieswould specialise in the production and export oflabour-intensive products; the employment andearnings of unskilled workers in what would other-wise have been immigration countries would therebybe affected.

The fear of a crowding-out effect on nationalsand legal migrants is reinforced when illegal immi-gration occurs during situations of rising unemploy-ment. According to a neo-classical approach, anincrease in unemployment would be analysed as afurther voluntary withdrawal by nationals and legalimmigrants from the labour market, their reservewages being higher than the now lower equilibriumwage brought about by the employment of illegal for-eign workers. Though what exactly voluntary unem-ployment and reserve wage mean in such a model is,to say the least, somewhat ambiguous, particularly inan open economy with the possibility of temporarylabour migration.

Let us imagine a large inflow of undocumentedforeigners ready to accept jobs at a wage 50% belowthe market rate – a movement that could lead to amassive withdrawal of nationals from the labour mar-ket. To consider these nationals as voluntarily unem-ployed because their reserve wage is too high wouldbe to render the notion of a reserve wage meaning-less. The reserve wage is established in relation tocertain expectations that workers have with regard totheir standard of living – taking into account time andplace, to use the academic definition of the rightprice. Migrants are prepared to accept wages which,

OECD 1999


240

given the constraint of an irreducible level consump-tion in the country where the jobs are, provide a stan-dard of living desired for family members who stayedbehind through remittances that increase in valuewith improvements in the terms of trade. The reservewages of residents of immigration countries and ofpotential migrants are associated with different refer-ence spaces and life styles. Differentials may explainwhy people resort to unauthorised immigration, butthey also pose the problem of an inter-personal com-parison of utilities.

The question of labour market competitionneeds to be examined together with concerns for thefiscal impact of illegal immigration. This brings usback to an examination of the extent to which deduc-tions and benefits are linked to territoriality, nation-ality and regularity in respect of residence.

In terms of taxation, whatever calculation tech-nique is used to itemise the expenditures andresources put into immigration as a whole (it is a cal-culation that cannot be simply reduced to a cross-sectional analysis, and has aroused a debate notablefor its extremely divergent positions), the fiscalimpact of unauthorised foreign workers and theirfamilies is generally positive as far as the nationalbudget is concerned: illegal migrants pay indirecttaxes in the same way as all other consumers andalthough they avoid paying income tax their reve-nues are generally such that the loss to the publicexchequer cannot be significant and is unlikely toexceed what they obtain in return, given that theiruse of community services tends to be even lowerthan their incomes. The only real “cost” linked toirregularity is for services the use of which is notdetermined by the legality of residence status; thesemainly concern the education of children of illegalimmigrants.

As far as social protection is concerned, contri-butions and entitlements are linked to regularity inrespect of residence and work. On the hypothesisthat regular migrants realise a net gain from the socialprotection system, a relative increase in the propor-tion of illegal immigrants in the immigrant popula-tion who do not contribute – but do not draw benefitseither – actually reduces the cost of the system. Onthe hypothesis that the contributions of regularmigrants are greater than the benefits they receive,which is usually the case, a similar relative increase inthe proportion of illegal immigrants represents a lossto the system. Whichever of these hypotheses isretained, the claim that the money that illegalmigrants take out of the social protection system is

the determining factor behind their decision tomigrate should be treated with caution. There is,however, an argument in favour of reallocatingresources between the various administrative tiers inorder to pay for higher charges at the local level.

The research by McCarthy and Vernez (1997)that reports on the results of surveys conducted bythe Rand Corporation in California can be cited as anexample. According to United States law, illegalimmigrants are not eligible for certain assistance pro-grammes, although they do qualify for education andfor food programmes. Children qualify for assistancein the same way as all children born in the UnitedStates. In fact, the drawing of social benefits dependsmore on migrants’ socio-economic status rather thanon their legal status. However, it is noteworthy that:a) hospitalisation is higher among children born inthe United States of illegal immigrants; b) many ille-gal immigrants qualify for access to certain social pro-grammes (e.g. AFDC, food stamps and Medicaid)insofar as their children born in the United States areeligible or one of the parents has the status of legalimmigrant; and, c) illegal immigrants tend not toapply for reduced tax liability.

The final element to be considered in this calcu-lation is the cost of control, which can indeed be veryhigh. Entry control seeks to counter a wide range ofillegal activities including drug trafficking and terror-ism. Not all the costs incurred in entry control can beattributed to illegal migration. Moreover, the specificcost of applying laws (e.g. frontier controls, the impo-sition of penalties on employers, and the investiga-tion and sending back of foreigners lacking theproper documents) are an integral part of generalimmigration policy and as such is the price to be paidfor operating a legal migration regime – unless, thatis, we were to imagine a totally open-frontier system.In other words, for illegal immigration to represent afiscal burden, it would have to be assumed that thecost of control (properly adjusted) exceeded thepositive net balance of fiscal revenues less socialexpenditures attributable to it. That is a highlyunlikely outcome given the extent of illegal immigra-tion’s positive impact on GDP.

In conclusion, in order for the phenomenon ofclandestine immigration to be comprehended it isessential that its economic dimension be taken intoaccount. It would appear however that economicanalyses of the labour market have not identifiedand incorporated within its models the specific fea-tures of clandestine immigration. This explains whythe majority of the studies of illegal immigration, in

OECD 1999


241

particular to the Unites States, analyse the employ-ment of illegal immigrants as if it were coterminouswith that of unskilled legal immigrants. The differ-ence in the economic incidence of illegal and legalimmigration appear then as a matter of degree ratherthan of nature

C. COMBATING CLANDESTINE IMMIGRATION

Before turning to an examination of the administra-tive effectiveness of controls, the questions of the legit-imacy of restrictions on movement and settlement andtheir economic efficiency, from both the world andnational point of view, naturally pose themselves.

With regard to economic efficiency, as definedby some criterion of world or national welfare, thereare arguments in favour of both an open frontier anda control policy. Using very restrictive hypotheses,particularly an otherwise fixed national supply oflabour, it is possible to demonstrate that total worldwelfare would increase were all frontiers opened up(Hamilton and Whalley, 1984). At a national level, onemay more realistically say that, for the recipientcountry, immigration has distributive and “free rider”effects in that it allows new entrants to benefit fromthe past accumulation of capital and technologicalprogress, the cost of which has been borne by previ-ous generations. However, looking over the longterm, there are too many non-measurable parame-ters. That is why immigration policies are usuallyfounded on short-term economic arguments, or onarguments of political philosophy that have nothingto do with the issue of economic impact.

In this regard, two observations can be made.First, in a democratic society, whilst the regulation ofinflows and the definition and implementation ofrules for accepting new members of the communityare seen as forming part of the legitimate attributesof a nation-state, the issue of sending immigrantsback against their will is much more problematic.There is an asymmetry here which, for the individual,recalls the asymmetry between the right to leaveone’s country and the absence of an equivalent rightto enter another. Liberal societies have also made acommitment not to close their frontiers to refugeesand asylum seekers. What, then, in these conditions,are the options for the State?

The combating of clandestine immigration iseffected by measures that attempt to control inflowsand length of stay, and curb the clandestine employ-ment of foreign labour. These two elements areclosely linked: curbing the clandestine employment

of foreign labour aims to limit clandestine inflowsand residence; the main justification for the controlof flows is to protect the domestic workforce fromunfair competition thought to come from the clan-destine employment of foreign labour. To these twoadministrative activities, we must add free trade andco-operation policies that seek to reduce the pro-pensity to emigrate.

1. Frontier control and controlling the length of stay

To measure the effectiveness of administrativemeasures requires that the policy’s objective hasbeen clearly defined. If the objective is to reduceillegal immigration and settlement in the country, arethe entry restrictions efficient, do they represent thebest option? To answer this question, we need totake account not only of the direct effects on inflows,but also of the effects on net migration. However,entry control is only one of the possible ways of lim-iting inflows and settlement in the country. It follows,therefore, in relation to the established objective,that we need to compare the cost-effectiveness offrontier controls with other indirect methods, forexample the imposition of penalties on the employ-ers of clandestine foreign workers or policies of co-operation with origin countries.

The policy of controlling inflows responds to gen-eral principles of immigration policy regarding thenumber and characteristics of legal migrants. In set-tlement countries, in particular the United States,this involves defining immigration criteria and mea-sures for combating clandestine immigration; in theEuropean Union, it takes the form of severelyrestricting labour immigration from non-membercountries. Whether they limit access to the countrythrough a policy of visas or prevent the clandestinecrossing of frontiers, these policies of controllingmigration flows have revealed their effectiveness butalso their limitations. Although they have reducedthe number of entries of the categories concerned,they have also led to their substitution, to somedegree, by those categories whose entry continues tobe possible and/or less controlled, and who havesubsequently transformed themselves into clandes-tine residents. Clearly, policies for controlling inflowscannot on their own hope to regulate immigration.The international opening of economies, the volumeof cross-border movements, the maintenance(indeed the heightening) of economic imbalancesbetween rich and poor countries, and the existenceof networks sustained by the presence of estab-

OECD 1999


242

lished foreign communities ensure that control is notsimply a matter of visas and frontier police.

In the context of globalisation, entry control pre-sents a real problem for developed countries. First,outside a closed system (such as that of the SovietBloc or the Berlin Wall), and in a democratic environ-ment, a selective, partial system of control is difficultto implement. Moreover, it is generally the case thatcountries that attract migrants are often those thatalso attract entrants of all kinds, mainly tourists andfamily visitors. The problem is how to distinguishclearly at the point of entry between potentialmigrants and other flows. No administrative measure[e.g. visa policies that distinguish between differentcategories of inflow, the insistence on special condi-tions for tourists such as the obligation to have areturn ticket and/or a certain amount of money, andthe recently withdrawn French system of an “accom-modation certificate” (certificat d’hébergement)] hasproved itself sufficient to meet the problem.13

In this context, the United States’ migrationpolicy, which focuses mainly on entry (i.e. the visasystem)14 and the surveillance of land frontiers,provides a particularly good illustration. How effec-tive is the threat of being apprehended by the Immi-gration and Naturalization Service (INS)? Researchby Espenshade (1995) does not offer a firmresponse.15 There is a 30% chance of being appre-hended each time an attempt is made to enter theUnited States, but of course migrants can simplykeep trying until they succeed. There is a 1-2%chance of being apprehended on American soil. It isclear that, when migrants reach the American border,they will cross it eventually, and once they are insidethe country, there is very little likelihood of theirbeing arrested. It appears, then, that there is no linkbetween the perceived risk of being arrested at theborder and the total number of illegal immigrants inthe United States. However, Espenshade concludesthat implementation of INS policies “does havesome dissuasive effect”. Over recent years, althoughthe strengthening of INS controls has increased therisk of apprehension it has also been accompaniedby an enlargement in the number of crossing points.

Assessing the administrative effectiveness ofcontrol policy by its impact on flow reduction can bedeceptive as soon as there exists ambiguity in thepolicy’s objective. In North America and Europe, thestepping up of controls does not rule out a degree ofacceptance of irregular entries. In the United States, atleast until 1994, the declared control policy at theMexican border was exercised somewhat selectively.16

In Europe, after frontiers were closed the admissionof asylum seekers was for a time and without callingthe principle of closure into question an indirect andmore efficient means of selecting entries.

Control over entry in no way guarantees controlover the length of stay. The strengthening of externalcontrols, however effective they may be as far asinflows are concerned, does have an impact on therate of migration return, on length of stay and, there-fore, on the stock of illegal immigrants in these coun-tries. In the United States, where the risk of beingapprehended at the frontier has increased overrecent years, migrants have tried to compensate forthe additional cost by crossing the border less oftenand spending more time in the United States. InEurope, the closing down of frontiers has had theeffect of slowing down the rate at which labourmigrants return home (with marked differences bynationality), consolidating the family reunion pro-cess, and extending the length of stay.

Democratic and liberal states face a range ofpolitical, legal and administrative difficulties. Sendingback long-standing foreign residents, particularly ifthey are in employment is, irrespective of their legalstatus, open to question. Among developed coun-tries, no recipient country has introduced plans for amass-scale repatriation. Public opinion is in favour ofrestrictive measures at entry but is much more reluc-tant to accept coercive repatriation measures. Thereare also legal and technical difficulties that hinder theimplementation of forced departure measures. Thelaw requires compliance with strict conditions forrepatriation or expulsion: these take the form of indi-vidual legal or administrative decisions, notifying theforeigner of the decision personally, and the possibil-ity of verifying the individual’s identity and citizen-ship. There is no shortage of examples of peopleagainst whom decisions have been taken, but whichhave proved impossible to carry out: these includeindividuals who have had their identity papers takenaway from them and cannot be expelled or escorted tothe border, and cases where, despite strong evidenceof a person’s nationality, it has been necessary, in theabsence of formal proof, to obtain acknowledgementof citizenship from the consulate in the country of ori-gin, the latter’s co-operation in such matters beinglinked to considerations outside the field of migration.Readmission agreements are specifically aimed atdealing with this type of problem.

The asylum applications system is a typical illustra-tion of the limitations affecting the control of inflowsand length of stay. The asylum system seeks to guar-

OECD 1999


243

antee legal protection for asylum seekers in situationswhere they are few in number and where there existsan indisputable means of assessing personal risk inthe country of origin. In the late 1980s and early 1990s,the use of the asylum procedure by candidates formigration, when all other means were out of the ques-tion or more difficult, has been reflected in a substan-tial growth in the number of asylum seekers in westernEurope, particularly Germany – to the point ofseverely hampering the functioning of the system.This crisis has led to reform of national legislation.

Let us take the case of someone entering a coun-try and applying for asylum. He files an applicationbut, because of the large number of requests out-standing, the processing can take months, sometimeslonger than two years. In most cases, the request isrejected in the first instance and the application goesthrough to appeal, a procedure designed to ensure sofar as possible that no legitimate request for refugeestatus is turned down. The appeal procedure is sus-pensive; the person can therefore stay in the countryfor the duration of the procedure. During this time, thecountry has to decide between the following alterna-tives: either the asylum seeker obtains a work permit,which means that if the appeal is rejected (which iswhat usually happens), the asylum seeker is to allintents and purposes settled and the implementationof the judicial decision (which should involve sendinghim to a foreign country) turns out to be technicallyand humanely out of the question; or the country doesnot grant the asylum seeker a work permit, and thisinvolves giving him financial assistance and enablinghim to qualify for social security benefits while heawaits the final decision. In the latter case, the chargeincreases over time and becomes increasingly difficultto justify, especially so given that some asylum seek-ers manage to find jobs in the hidden economy.17

Countries faced with this dilemma have initiallychosen to speed up the process by increasing thebudgets of the government departments responsi-ble for refugees. This has been effective. However, itultimately proved necessary to make the asylumscheme more restrictive: in Germany, this resulted inan amendment to Article 16 of the Basic Law; inEurope more generally, it took the form of imple-menting the Dublin Convention whereby candidatesmay make only one application in the EuropeanUnion. These measures have led to a significantreduction in the number of asylum seekers. Yet,greater financial resources and stronger controlshave only a limited effect on the principle of the asy-lum system; there is therefore reason to believe that

the asylum route will continue to be a one of the keycomponents of migration flows. The asylum proce-dure, which is specific with regard to entry, illustratesthe general difficulty of controlling the length of stay.

It appears therefore that there is an inter-relation-ship between entry control and the regulation of migra-tion return. A country which makes entry more difficultthereby reduces the incentive to leave for thosealready present. However, when confronted with thedifficulty of exerting control over length of stay, coun-tries react by raising barriers to entry, often going fur-ther than the degree of closure that was initiallyplanned and seen by the country as desirable. Immi-gration control can only be carried out at entry, but theeconomic, social and political problems of immigrationare more the result of the settlement of migrants.

As soon as unauthorised migration is, in theabsence of reliable data, assumed to reach propor-tions deemed “large”, governments face a dilemma:either they apply the existing legislation and therebycreate a group of marginalised individuals, or theydecide on an amnesty which has to be presented andperceived as an exceptional and “one-off” operation.In the latter instance efforts are made in the hope ofensuring that the amnesty does not attract newinflows,18 and in particular that it does not stimulateexpectations of another amnesty.

The European experience, particularly since1973-74, and the American experience (IRCA, 1986)show that no regularisation exercise has ever put anend to the presence of clandestine immigrants, inthe sense that none have affected the decisive fac-tors and mechanisms underlying illegal migration.This does not mean that regularisation programmesshould be ruled out, rather that it is important to rea-lise that amnesties (partly) efface the past, but haveno effect on the future. However, the American exam-ple shows that, even in the absence of any prospectof a new amnesty, illegal migration has once againincreased; this makes it quite clear that there doesnot need to be the prospect of a new amnesty for thestock of clandestine immigrants to increase. Recentexperience in Europe shows that it is not an option todecide whether to have an amnesty or not; rather, thechoice is between repeated amnesties and discreetamnesties carried out on a case-by-case basis.

2. Employment controls and penalties against employers

The imposition of penalties against employerswho use clandestine labour, that is to say the second

OECD 1999


244

form of control, aims to influence the determiningfactors of unauthorised migration but, like regularisa-tion, it is a means after the fact of dealing with theproblem.

In the light of the most recent European andAmerican experiences, sanctions against employerswould appear to be of very limited effectiveness.There are a number of reasons for this. First, there areintrinsic difficulties in implementing such a system;some are legal or technical, while others are linked tothe socio-cultural environment. In most cases, employ-ers are not asked (or, more accurately, they cannot beasked) to check the validity of documents producedby migrant workers. As a result, even if the employeris under a duty to check on a worker’s status, he canonly be punished for not ensuring that the docu-ments were valid. More importantly, the system ofsanctions can only be effective if it is perceived bysociety, particularly at local level, as a legitimate ele-ment of labour market intervention.

In reality, there is usually quite a gap betweenbeing found breaking the law and being found guiltyin a court of law. It is hardly surprising, therefore, thatthe legal process comes to a halt somewherebetween the two as a result of local considerationsand political pressures that highlight the difficultiesthat enterprises would face if they had to pay forlabour at the current market rate and the risks that acessation of activities would have on the employ-ment of nationals.

Sanctions against employers also have an eco-nomic dimension. The effectiveness of the system ofsanctions depends on the extent to which a sanctionreduces the demand for unauthorised workers byadding a risk premium to their wage. This effective-ness also depends on the extent to which sanctionsencourage the replacement of clandestine immi-grants by nationals or legal immigrants therebyreducing clandestine entries. There is a price thatentrepreneurs are prepared to pay to be able to hireunauthorised workers; the problem may be formu-lated within the theory of optimal taxation (Hill andPearce, 1990). Entrepreneurs will risk employingunauthorised foreigners for as long as the expectedcost of being penalised (a function of the expectedsize of the penalty and the likelihood of beingcaught) is lower than the difference between thelabour costs of illegal and legal migrants.

The unknown element for the employer is thelikelihood of being checked up on. It can however becalculated, or at least its determinants can be

assessed on the basis of such factors as the amount ofmoney invested by the competent governmentdepartment (the INS, for example), the degree of geo-graphical and sectoral concentration of illegal workers,and the size and location of the company. This eco-nomic analysis needs qualification in the case of Euro-pean countries. The vision of the market economyignores the shame that can go with being found guiltyof employing illegal workers. Clandestine workers canbe hired by small enterprises in the formal and infor-mal sectors, but not by large firms. These latter pos-sess, however, the political influence to recruit themigrants they need through other, more widelyaccepted, means. It follows that the perceived concen-tration of illegal workers in certain jobs, and in firmsmarked by certain characteristics, can be deceptive.

There remains the question of the discrimina-tory effect that employer sanctions have on legalmigrant workers. An honest employer who does notseek to hire illegal migrants might refuse to hireworkers who display characteristics that point to ahigh probability of their being clandestine (e.g. a lowlevel of skills, or a poor command of the language).Paradoxically, law-abiding employers who have nointention of discriminating are more likely to operateon this basis thereby actually discriminating againstlegal workers. However, this legitimate concern mustnot be used as an excuse for relaxing a system ofsanctions that aims to prevent unfair competitionamong unskilled workers (North, 1994). In any case,that administrative effectiveness is limited does notjustify the conclusion that there should be no con-trols at all. After all, that the police will never succeedin completely stamping out criminality would not jus-tify disbanding their force.

The significance and scope of these measuresalso depend on the context in which they operateand on the importance attached to the socio-economic and legal dimensions of the offence. InEuropean countries such as France, the combating ofthe employment of foreigners in an irregular situa-tion forms part of the wider campaign against clan-destine employment. It aims above all to securecompliance with employment legislation with a viewto preventing the unfair competition that wouldresult from hiring clandestine labour. The penaltiesare directed at employers, not at workers. To thisend, drawing upon the lessons learned from applyingthe law, in particular the variety of legal devices usedwith the aim of evading employment and taxationregulations, attempts have been made to strengthenand reposition control provisions (i.e. focus on pre-

OECD 1999


245

vention), the justification for which goes beyondadministrative and economic efficiency. The ques-tion of the efficiency of sanctions, notably with regardto how much they reduce clandestine immigrationand what impact they have at the macroeconomiclevel, has been accorded rather less attention.

The perspective in the United States is different.The imposition of penalties on employers seeks toreduce clandestine entry and residence; it is not firstand foremost a legal instrument of regulating thelabour market. Here, as a result of experienceacquired since 1986, less effort has been made toalter labour market regulation in response to theinventiveness of clandestine workers and theiremployers. Instead, questions have been askedabout the relative efficiency of implementing lawsagainst employers as compared with measuresaimed directly at curbing clandestine inflows at fron-tiers. Taken to their logical conclusion, were it to beshown that, by comparison with legal immigration,illegal immigration is of greater benefit to the econ-omy then such considerations of economic efficiencywould lead to the conclusion that implementing lawsagainst employers has a negative economic impact.

This means that sanctions are efficient only ifthey reflect society’s view of the legitimacy andnecessity of combating clandestine employment andform part of a more general policy directed to thisend. Real control is social control, as the Swiss exam-ple testifies. It necessitates at the local level a con-sensus on the need to combat the phenomenon andimplement sanctions against employers. The newstrategic orientation adopted by the INS, which pro-poses to associate local communities and the imple-menting agencies in its actions in order to “identifyand define the problems linked to illegal immigra-tion” is in line with such an view (INS, March 1999).

3. Policies designed to reduce the propensity to emigrate

In a world where demographic and economicimbalances produce a supply of emigrants out of allproportion to the reception capacities and demand indeveloped countries, it has become necessary to re-sit-uate migration in the context of development, and topromote a strategy of trade liberalisation, incentives forprivate investment, aid and co-operation, from whichone might eventually hope to see a reduction ininequalities between nations and less incentive toemigrate. The realism of this approach, and theconfidence we can have in it, depend on the

answers to two lines of inquiry: What effect can theliberalisation of national economies and tradeliberalisation be expected to have on growth inrevenues and employment? and, What effect canstronger economic growth in origin countries beexpected to have on the incentive to emigrate(Tapinos, 1999)?

The impact of external opening on the growth ofeconomies is highly debatable. The analysis mainly(and often solely) focuses on the impact on exportsand imports and on growth in the economy; it is morediscreet about the effects on employment andincome distribution.19

The relationship between economic growth andemigration is even more uncertain. Economic theoryand empirical observation suggest two possibleapproaches. If the explanation for migration is mainlyto be found in income differentials (this is the onlyscenario produced by the theory of internationaltrade that assumes full use of factors of production)and employment (the neo-classical theory of alloca-tion of factors), and if the opening up to trade andmovements of capital is likely to encourage growth inthe revenues and well-being of all trading partnersand stronger growth among the less developed coun-tries, it is possible to predict that the surplus work-force will be absorbed and that there will beconvergence in wages.

A reduction in the incentive to emigrate followsfrom this; free trade and development aid are thenan alternative to population movements. On theother hand, if it is true that the origin of migration liesin the breakdown of the demographic-economicequilibrium, which results notably from the take-offof the development process, then an increase ininternal mobility, which is inherent to this process,normally extends outside the confines of the country.According to this hypothesis, development andinternational migration go hand in hand (Tapinos,1974; Diaz Briquets and Weintraub, 1991, Massey,1988). This hypothesis finds confirmation in empiri-cal studies of the American and European cases. Thetwo viewpoints may be reconciled through the intro-duction of a time dimension. In the medium term, thekind of development fostered by trade liberalisationstrengthens the incentive to emigrate; in the longterm, improved standards of living make emigrationless advantageous for potential candidates.

The decrease in clandestine migration is envis-aged in this way as the consequence of a fall in thegeneral propensity to emigrate. However, the selec-

OECD 1999


246

tive character of migration needs to be taken intoaccount. If those who are disposed to migrate clandes-tinely are different from other potential migrants,whether these characteristics be observable or other-wise, then the question of the specific impact of devel-opment and trade liberalisation will remain open. Apartial indication in this context was provided byHanson and Spilimbergo (1996), who showed thatillegal migration as measured by apprehensions is afunction of wage differentials between Mexico and theUnited States, but that it is more responsive to varia-tions in real wages in Mexico than to variations in realwages in America. Generalisation on the basis of thisobservation would not, of course, be warranted.

Conclusion

What lessons can we draw from this examinationof the economic and political issues of internationalmigration in the light of experiences in North Americaand Europe? What is the scope for public interven-tion in the context of accelerating globalisationalongside the persistence of demographic and eco-nomic disequilibria between the rich and poor coun-tries and the instability resulting from intra- andinter-national conflicts?

The phenomenon of clandestine migration is notconfined to any one migration regime in particular.Neither a regime open to regular migration such asthat in North America nor one of closed borders suchas that prevailing in Europe guarantees the contain-ment of clandestine entries and the prevention ofunauthorised residence by foreigners in an irregularsituation. The diversity of the potential forms of clan-destinity, the multiplicity of the objectives in the fightagainst clandestine immigration, the conflicts of inter-est between different groups in recipient countriesand the ambivalence of public opinion necessitate theimplementation of a range of measures. The effective-ness of these measures rests on the assumption that

migration is a process with inflows and outflows. It isthe obstacles to entry which have the effect of ulti-mately inducing an increase in the number of illegalimmigrants greater than would be the case under amore liberal system of entering and leaving.

The problem of the control of migration flowsshould not be viewed as being uniquely bilateral. Theemergence of new geographic areas for migrationimposes on governments forms of co-operation rang-ing from the exchange of experiences through to co-ordination in the methods they employ to regulatemigration flows.20 Yet in a world increasingly open tothe circulation of goods and capital, where the reduc-tion in transport costs leads to considerable cross-border movements linked to tourism, to the acquisi-tion of tertiary education and professional training andto business travel it is anachronistic to conceive ofimmigration control exclusively in terms of sovereignty.

A distinction needs to be made between move-ment, the expression of the human right to come andgo, and settlement which is conditional on the legitimateprerogative of societies to decide the extent to andthe conditions under which they are ready to acceptnew members. For all that, the difficulty of managingthe risk that incomers will remain beyond the durationof their visa cannot legitimise the imposition of restric-tions on the circulation of persons for this would bedetrimental to the majority of them who desire merelyand with honest intent to undertake a brief stay.

It should be noted also that migrants belong nei-ther to their country of origin nor to their country ofdestination. It is an illusion to imagine that irregularmigration can be eliminated solely by state interven-tion and improved co-operation between the originand destination countries. Clandestine migration, aswell as being a breach of the law is also a manifesta-tion of individual liberty. Between the inspector whocarries out his duty and the migrant in search of hisdestiny, the issues are not of the same nature. This isthe root of the problem.

OECD 1999

247


NOTES

1. The OECD and the Dutch authorities with the supportof the United States organised a seminar on prevent-ing and combating the employment of foreigners in anirregular situation. The seminar took place in TheHague on 22-23 April 1999. The seminar was a continu-ation of the work carried out since 1995 by the OECDSecretariat under the auspices of the Working Party onMigration. This work has facilitated the study of illegalimmigration in several OECD Member countries inrelation to labour market and other programmes toregularise foreigners in an irregular situation. The sem-inar comprised three sessions. The first session analy-sed the impact of irregular migration on the economyand the labour market. The second session was dedi-cated to the comparative analysis and the evaluationof measures undertaken to prevent and combat theemployment of foreigners in an irregular situation. Thethird session was composed of two round tables. Par-ticipants in the first round table included OECD policymakers, representatives of governmental administra-tions in charge of the implementation of sanctionsagainst labour traffickers, dishonest employers andforeigners in an irregular situation. Representatives ofTrade Unions also contributed to the discussions onthe difficulties encountered in implementing sanc-tions at national levels and proposals for actions toovercome these difficulties. The second round tablewas devoted to the extension and improvement ofinternational co-operation directed at effectively pre-venting and combating the employment of foreignersin an irregular situation. The proceedings of the semi-nar as well as the conclusions and recommendations ofthe OECD Secretariat will be published by the OECDin January 2000. The document presented at the semi-nar by Georges Tapinos (Irregular migration: economicand political issues) has been chosen for the SpecialChapter of this edition of the annual report "Trends inInternational Migration".

2. The relative proportion of these two components inthe number of illegal migrants varies from country tocountry. In the United States, it has been estimatedthat illegal entries (entries without inspection) andoverstayers each account for 50% of total illegal immi-gration (Warren, 1994). A more recent estimate by theINS puts the figures at 60% and 40% respectively.

3. One can not always distinguish these different catego-ries in the administrative statistics. Indeed, theserecord either legal movements or infractions which,when they are penalised, measure clandestinity at thevery moment that it has ceased to exist. Moreover,

administrative statistics, çross-sectional by nature,record the events or migrants affected by the eventswithout, in general, it being possible to attach theevents to a reference population. Further, there is dis-tinction to be drawn as to the reliability of the informa-tion: between administrative records which caninclude a fiscal dimension or an element of sanctionwhich might incite inexact declarations, and data whichrespond to a desire for socio-economic knowledgesuch as censuses and some surveys; they do not nec-essarily have the same utility when one seeks to deter-mine the number of clandestine immigrants. Whenclandestine immigration is tolerated, it is quite possi-ble that a census will record a non-negligible propor-tion of the clandestine immigrants.

4. It is a gross over-simplification, but we can identifythree overlapping types of procedure. Firstly, there aresurveys that have the advantage of including charac-teristics of economic sectors and the workers con-cerned, but do not identify variations over time. Then,there are numerical methods that focus on anomaliesbetween observed and assumed developments; thevariables customarily used are participation rates, thedisparity between national revenue and expenditures,the proportions of expenditure using cash and large-denomination notes, and household electricity con-sumption. These methods need a departure point atwhich time it is reasonable to assume that the hiddeneconomy is, to all intents and purposes, non-existent.It is then possible to identify growth in the hiddeneconomy over time. A more ambitious econometricprocedure uses a theoretical model of the informaleconomy that involves the introduction of a series ofvariables defining determiners (e.g. fiscal pressure orregulations covering employment) and indicators ofthe hidden economy (e.g. participation rates). Becauseof the wide variety of procedures used and specificfeatures of the statistical factors requested and thefield covered, it is difficult to make comparisons. As apercentage of GDP, estimates range between 2% and4%, with particularly high levels for certain sectors suchas building, hotels and catering, domestic services andagriculture.

5. Despite the multiplicity of conceivable procedures, formost of the countries concerned we do not have a sin-gle estimate worthy of the name. For the Europeancountries, one is obliged to content oneself with verypartial measurements, typically estimates inferredfrom the number of asylum or regularisation applica-tions refused. In the United States, however, the civil

OECD 1999


248

service furnishes one of the most reliable estimates ofthe stock of clandestine immigrants. The number inOctober 1996 was put at 5 million, there having takenplace between 1992 and 1996 an annual growth of275 000 (US Statistical Abstract 1996 and GAO,April 1999).

6. This can vary for nationals and established regularmigrants. Although there is certainly more of an incen-tive for nationals to work in the informal economy thanfor undocumented foreigners, the advantages for theformer outweigh the risk; this is not so for illegal work-ers. Contini (1982) cites the example of Fiat whose staffbetween the mid-1960s and the early 1970s left foremployment with smaller enterprises where there wasmore work and higher capitalisation and profits. Regu-lar migrants, too, have accepted restrictions that lim-ited their mobility.

7. However, it should be pointed out that there are caseswhere undocumented migrant workers have lost theirjobs following regularisation; they are less likely torepeat the experience another time.

8. The decision to sub-contract part of production rein-forces, and often conditions, the opportunity to reducewages and avoid mandatory deductions (i.e. taxes andsocial security).

9. Sub-contracting, a practice that responds to a logic ofefficiency, also allows artificially created companiesto expand by employing unauthorised labour. Forexample, a French legal judgement refers to a sub-contractor who employed eight people, five of whomhad no residence permits or work permits, and one ofwhom had no work permit. Work was supervised by acharge-hand from the main company, the workers werelodged in huts hired by the same company, and all theequipment and materials used for production were alsothe property of the main company. Sub-contractingenabled the main company to avoid paying taxes andsocial charges (Pupier, 1992).

10. Unauthorised migrants often have a better under-standing of conditions in the informal labour marketthan nationals and immigrants who have been in thecountry for a long time, particularly in the services sec-tor. This gives them a comparative advantage.

11. For example, some Sri Lankans who had no real inten-tion of settling, applied for asylum in order to takeadvantage of the slow pace at which the applicationprocess is conducted, and used the time to build uptheir savings before returning home (Costes, 1991).

12. We have indeed very little research material of the sta-tistical quality necessary to make an assessment. Oneof the monographs worthy of note, that by Venturini(1997) measured the elasticity of regular employmentcompared with unauthorised, irregular employmentand threw doubt on the hypothesis that there is com-petition from irregular foreign labour. Using ISTAT dataon the hidden economy during the period 1980-1994,she showed in 14 branches in 5 sectors (agriculture,industry, construction, tradable services and non-tradable services) that there are variations between

sectors, in particular that the depressive effect is morepronounced in agriculture, but that the overall impactis small. However, in another context, other studiesshow that the pay of undocumented workers is lowerthan that of nationals (allowing for different levels ofproductivity), and that they perform lowly regardedjobs with few differences in pay determined by quali-fications, gender or age (Lianos, Sarris and Katseli,1996; Markova and Sarris, 1997).

13. However, as statistics on flows into European countriesshow, it would be wrong to under-estimate the impactof the visa policy.

14. Under American law, obtaining a visa from an Americanconsulate abroad does not remove the right of immi-gration officers to refuse entry at the point of entry.

15. n 1993, the INS spent $US 362m on implementing thepolicy; to this sum, we must add the $US 12.2m spenton equipment used in achieving the same objective.That year, 1 282 000 apprehensions were carried out byborder patrols; another 45,000 were made inside theUnited States. In 1998, the INS spent around $US 900mfor the Budget Authority for Border Patrol (cf. Tight-ened Controls and Changing Flows: Evaluation the INSBorder Enforcement Strategy by R. Suro in ResearchPerspectives on Migration, Vol. 2, n°1, 1999).

16. This has been well documented by the Zapata CanyonProject (Bustamante, 1990). Illustrations of this includethe fact that most migrants do not choose to cross thefrontier where the geographical conditions are leastdangerous, but where there is a large labour market onthe American side (e.g. the Tijuana-San Diego border),and the large number of 'illegal commuters'.

17. It is interesting to note that countries choosing the formeroption (e.g. France) have subsequently gone for the latter,and countries choosing the latter (e.g. Germany) havesubsequently adopted the former.

18. This assumes that provision is made for deadlines,proof that the applicant is indeed residing in the coun-try, and possibly further proof that he has a job.

19. Given the wide range of experiences, it is particularlyinteresting to identify the factors likely to influence thedirection of variations and to measure their effects. In areview of the literature, Edwards (1993) selects five ele-ments: liberalisation measures, in particular the distinc-tion between, on the one hand, the removal of quotaand tariff barriers and, on the other, devaluation that canbecome necessary as a result of the opening; the time-tabling and speed of the opening; the stage of develop-ment that the country has reached; the internationalenvironment; and labour market distortions.

20. Hence the lifting of internal controls within theEuropean Union and the drafting of common regula-tions for external border control (the Schengen inter-governmental agreement and the Dublin Conventionon asylum seekers). For all this, none of the countriesconcerned has ruled out the possibility of applyingspecial conditions to certain categories of would-beentrants.

OECD 1999

249


BIBLIOGRAPHY

BAGANHA, M.I. (1998),“Immigrant involvement in the informal economy: thePortuguese case”, Journal of Ethnic and Migration Studies,Vol. 24, No. 2, April, pp. 367-385.

BEAN, F.D., LOWEL, B.L. and TAYLOR, L.J. (1988),“Undocumented Mexican immigrants and the earningsof other workers in the United States”, Demography,Vol. 25, No. 1, February, pp. 35-52.

BENEDETTI, L. (1991),“Le travail clandestin entre le travail choisi et le travailpossible”, in Montagne-Villette, S. (ed) Espace et travailclandestin, Masson, Paris, pp. 65-70.

BERANEK, W. (1982),“The illegal alien work force, demand for unskilledlabour and the minimum wage”, Journal of LabourResearch, Vol. III, No. 1.

BHAGWATI, J.N. (1981),“Alternative theories of illegal trade: economic conse-quences and statistical detection”, WeltwirtschaftlichesArchiv., 117, pp. 409-427.

BOND, E.W. and CHEN, T.J. (1987),“The welfare effects of illegal immigration”, Journal ofInternational Economic, 14, pp. 315-328.

BORJAS, G., FREEMAN, R.B. and KATZ, L.F. (1986),“Searching for the effect of immigration on the labormarket” in American Economic Review, May, pp. 246-251.

BRATSBERG, B. (1995),“Legal vs. illegal U.S. immigration and source coun-try characteristics”, Journal of International Economics,pp. 715-726.

BUCCI, G.A. and TENORIO, R. (1996),“On financing the internal enforcement of illegal immi-gration policies”, Journal of Population Economics, Vol. 9,pp. 65-81.

BUSTAMANTE, J. (1990),“Undocumented migration from Mexico to the UnitedStates: preliminary findings of the Zapata canyon Project”in Undocumented Migration to the United States. IRCA and theExperience of the 1980s, Santa Monica, Calif., Rand Corpora-tion, Washington, D.C., The Urban Institute.

CHISWICK, B.R. (1986),“Illegal aliens: a preliminary report on employee-employer survey”, American Economic Review, Vol. 76,No. 2, May, pp. 253-257.

CHISWICK, B.R. (1986),Illegal Aliens: Their Employment and Employers, WE UpjohnInstitute for Employment Research.

CHISWICK, B.R. (1988),“Illegal immigration and immigration control”, Journal ofEconomic Perspectives, Vol. 2, pp. 101-115.

CLARK, R.L., PASSEL, J.S., ZIMMERMANN, W.N. and FIX, M.with MANN, T.L. and BERKOWITZ, R.E. (1995),Fiscal Impacts of Undocumented Aliens: Selected Estimates for SevenStates, The Urban Institute, Washington D.C., January.

COBB-CLARK, D. A., SHIELLS, C. R. and LOWELL, B. L. (1995),“Immigration reform: the effects of employer sanctionsand legalization on wages”, Journal of Labor Economics,Vol. 13, No. 3, July, pp. 472-498.

COMMISSION FOR THE STUDY OF INTERNATIONALMIGRATION AND COOPERATIVE ECONOMIC DEVELOP-MENT (1990),Unauthorized Migration: An EconomicDevelopment Response,Report, July.

CONTINI, B. (1982),“The second economy of Italy” in Tanzi V. (ed.), TheUnderground Economy in the United States and Abroad, DCHeath and Co., Boston, pp. 199-208.

COSTES, L. (1991),“Immigrés et travail clandestin dans le métro” inMontagne-Villette, S. (ed) Espaces et travail clandestins,Masson, Paris.

CORNELIUS, W.A., MARTIN, P.L. and HOLLIFIELD, J.F. (1994), Controlling Immigration: A Global Perspective, Stanford Uni-versity Press, Tanford, California.

DELL’ARINGA, C. and NERI, F. (1989),“Illegal migrants and the informal economy in Italy” inGordon, I. and Thirwall, A.P. (eds) European Factor Mobil-ity (Trends and Consequences), Basingstoke (Macmillan),pp. 133-147.

DELAUNAY, D. and Tapinos, G. (1998)La mesure de la migration clandestine en Europe. Vol. : Rapportde synthèse, Eurostat Working Paper (Population etconditions sociales).

DJAJIC, S. (1997),“Illegal immigration and resource allocation”, InternationalEconomic Review, Vol. 38, No. 1, February, pp. 97-117.

DJAJIC, S. (1997),“Dynamics of immigration control”, CEPR Worshop,14-15 February, Athens.

DONATO, K.M., DURAND, J. and MASSEY, D.S. (1992),“Stemming the tide? Assessing the deterrent effects ofImmigration Reform and Control Act”, Demography,Vol. 29, No. 2, May, pp. 139-157.

OECD 1999


250

DUNN, L. F. (1990),“An empirical study of labour market equilibriumunder working hours constraints”, Review of Economics andStatistics, Vol. 72, No. 2, May, pp. 250-258.

ESPENSHADE, T.J. (1995),“Using INS border apprehension data to measure theflow of undocumented migrants crossing the U.S.-MexicoFrontier”, International Migration Review, 29, pp. 545-565.

ETHIER, W.J. (1986),“Illegal immigration. The host country problem”, TheAmerican Economic Review, Vol. 76, pp. 57-71.

ETHIER, W.J. (1986),“Illegal immigration”, The American Economic Review,Vol. 76, No. 2, May, pp. 258-262.

GARSON, J.P. and THOREAU, C. (1997),“Lessons from regularisation programmes and measurestaken to combat employment of foreigners in irregularsituations”, CEPR Workshop, 14-15 February, Athens.

GONZALES, J.G. (1994),“Illegal immigration in the presence of labor unions”,International Economic Journal, Vol. 8, pp. 57-70.

GROSSMAN, J.B. (1984),“Illegal immigrants and domestic employment”, Indus-trial and Labour Relations Review, Vol. 37, No. 2, January,pp. 240-251.

HAMILTON, B. and WHALLEY, J. (1984),“Efficiency and distributional implications of globalrestrictions on labour mobility”, Journal of DevelopmentEconomics, 14, pp. 61-75.

HANSON, G. H. and SPILIMBERGO, A. (1996),“Illegal immigration, border enforcement, and relativewages: evidence from apprehensions at the U.S.-Mexico Border”, NBER Working Paper n° 5592, May.

HAUT CONSEIL A L’INTÉGRATION (1992),L’emploi illégal des étrangers (rapport particulier).

HILL, J.K. and PEARCE, J.E. (1990),“The incidence of sanctions against employers of ille-gal aliens”, Journal of Political Economy, Vol. 98, No. 1,pp. 28-44.

ISKANDER, N. (1999),“Immigrant workers in an irregular situation: the case ofthe garment industry in Paris and its suburbs”, TheHague Seminar, OECD, 22-23 April.

KARAMESSINI, M. (1996),“Labour flexibility and segmentation of the Greek labourmarket in the eighties: sectoral analysis and typology”,European Association of Labour Economists, 8e con-férence annuelle, Crete, 12-22 September.

LAYARD, R., NICKELL, S. and JACKMAN, R. (1991),Unemployment, Macroeconomic Performance and the LabourMarket, Oxford University Press.

LEVINE, P. (1995),“The welfare economics of legal and illegal migration”,CEPR Workshop, 14-16 September, Halkidiki.

LIANOS, T. P., SARRIS, A. and KATSELI, L.T. (1996),“Illegal migration and local labour markets. The case ofNorthern Greece”, International Migration, Vol. XXXIV,No. 3, pp. 449-484.

LUTZ, V. (1963),“Foreign workers and domestic wage levels, with anillustration from the Swiss case”, Banca Nazionale delLavoro Quarterly Review, March, pp. 3-68.

McCARTHY, K. F. and VERNEZ, G. (1997)Immigration in a Changing Economy – California’s Experience,RAND, Santa-Monica.

MARIE, C.V. (1994),La verbalisation du travail illégal, les chiffres de l’année 1994,MILUTMO, Mission de liaison interministérielle pourla lutte contre le travail clandestin, l’emploi nondéclaré et les trafics de main-d’œuvre, Paris.

MARIE, C.V. (1997),“La loi du 11 mars 1997 sur le travail illégal : les don-neurs d’ordre en ligne de mire”, Regards sur l’actualité,No. 235, November, pp. 31-53.

MARKOVA, E. and SARRIS, A. H. (1997),The Performance of Bulgarian Illegal Immigrants in the GreekLabour Market, Department of Economics, Athens Uni-versity, February.

MASSEY, D. S. (1987),“Do undocumented migrants earn lower wages thanlegal immigrants? New evidence from Mexico”, Interna-tional Migration Review, Vol. XXI, No. 2, pp. 236-274.

MASSEY, D.S. (1998), “Economic development and international migration” inPopulation and Development Review, September, pp. 383-413.

MIGRATIONS-SOCIÉTÉ (1995)“Marché de l’emploi et clandestinité”, Vol. 7, No. 39,May-June, pp. 30-91 (Suite d’articles).

MOULIER-BOUTANG, Y., GARSON, J.P. and SILBERMAN, R.(1986),Économie politique des migrations clandestines de main-d’œuvre,Publisud, Paris.

MOULIER-BOUTANG, Y. (1991),“Dynamique des migrations internationales et écono-mie souterraine : comparaison internationale et pers-pectives européennes”, in Espaces et travail clandestin,Montagné-Vilette (S.), Paris, Masson, pp. 113-121.

NORTH, D. (1994), “Enforcing the minimum wage and employer sanc-tions”, Annals AAPSS, Vol. 534, pp. 58-68.

NORTH, D. and HOUSTON, M. (1976),The Characteristics and Role of Illegal Aliens in the U.S. LaborMarket: an Exploratory Study, Report, U.S. Department ofLabour, Washington D.C.

OECD (1990), “Comparative analysis of regularisation experience inFrance, Italy, Spain and the United States” in SOPEMI,Paris, pp. 65-101.

PALLIDA, S. (1990),“L’immigration entre l’économie ethnique et l’écono-mie souterraine”, in Abou Sada, G., Courault, B. etSeboulou, Z. (ed) L’immigration au tournant, CIEMII,L’Harmattan, Paris.

PHILIBERT, J.P. and SAUVAIGO, S. (1996),Rapport sur l’immigration clandestine et le séjour irrégulierd’étrangers en France, Commission d’enquête; Rapport

OECD 1999


251

n° 2699 de l’Assemblée nationale, Tome 1 : Rapport,Tome 2 : Audition, Paris.

PORTES, A., CASTELLS, M. and BENTON, L. A. (eds.) (1989), The Informal Economy – Studies in Advanced and Less Devel-oped Countries, The Johns Hopkins University Press,Baltimore and London.

PUPIER, A. (1992),“La fausse sous-traitance dans le bâtiment et les travauxpublics”, Sociétés contemporaines, No. 10, pp. 153-170.

REYNOLDS, C.W. and MCCLEERY, R.K. (1988),“The political economy of immigration law: impact ofSimpson Rodino on the United States and Mexico”,Journal of Economic Perspectives, Vol. 2, No. 3, pp. 117-131.

RICCA, S. (1984),“Administering migrant workers in an irregular situa-tion in Greece, Italy and Spain”, Working PaperNo. 11E, ILO, Geneva.

ROBIN, S. and MARIE, C. V. (1995), “A comparative analysis of legislation in several Euro-pean countries aimed at the illegal employment of for-eigners”, OECD Working Party on Migration, Paris.

ROTHMAN, E.S. and ESPENSHADE, T.J. (1992),“Fiscal impact of immigration to the United States”,Population Index, Vol. 58, No. 3, pp. 381-415.

SLIMANE, L. (1995),L’immigration clandestine de la main-d’œuvre dans la régionbruxelloise, Bruylant, Brussells.

TANZI, V. (ed.) (1982),The Underground Economy in the United States and Abroad,Lexington Books, Lexington (Mass.).

TAPINOS, G. (1973),L’ économie des migrations internationales, Fondation natio-nale des Sciences politiques/Armand Colin, Paris.

TAPINOS, G. (1992),“La pression migratoire : sentiment d’inquiétude ouconcept analytique ?”, in Tapinos, G. and Keely, C.B.La migration internationale : deux approches, Programmemondial pour l’emploi, Document de travail, ILO,Geneva, May, pp. 1-14.

TAPINOS, G. (1999), “Migration, trade and development.The European Union and the Maghreb countries”, inEldorado or Fortress? Migration in Southern Europe, editedby Russel King, Gabriella Lazaridis and C. Tsardanidis,Mac Millan, London.

TAPINOS, G. (en collaboration avec de RUGY, A.) (1993),“The macroeconomic impact of immigration. Review of

the literature published since the mid-1970” in Trendsin International Migration, OECD, Paris.

TAPINOS, G., LACROIX, P. and RUGY, A. de (1996),Les méthodes d’évaluation de l’immigration clandestine dans cer-tains pays étrangers, Fondation nationale des Sciencespolitiques, December, Paris.

TAPINOS, G. and DELAUNAY, D. (1999),“Can one really talk of the globalisation of migrationflows?” in Globalisation, Migration and Development, OECD,Paris.

TIENDA, M. and SINGER, A. (1995),“Wage mobility of undocumented workers in theUnited States”, International Migration Review, Vol. XXIX,No. 1, Spring, pp. 112-138.

TODARO, M.P. and MARUSZCO, L. (1987), “Illegal migration and U.S. immigration reform”, Popula-tion and Development Review, Vol. 13, No. 1, March.

UNITED NATIONS (1998),World Population Monitoring, 1997: International Migrationand Development, United Nations, New York (Economicand Social Affairs).

UNITED STATES GENERAL ACCOUNTING OFFICE (1997)Illegal Immigration Southwest Border Stragegy Inconclusive:More Evaluation Needed, GAO, December.

UNITED STATES GENERAL ACCOUNTRIG OFFICE (1999)Illegal Aliens-Significant Obstacles to Reducing UnauthorizedAlien Employment Exist, GAO, April.

US DEPARTMENT OF JUSTICE (1999),Backgrounder, Immigration and Naturalization Service,Interior Enforcement Stragegy, March.

VAN HAMERSFOORT, H. (1996), “Migration: the limits of governmental control”, NewCommunity, Vol. 22, No. 2, April, pp. 243-257.

VENTURINI, A. (1997), “Do immigrants working illegaly reduce the natives’slegal employment? Evidence from Italy”, Paper pre-sented for the CEPR Seminar on Illegal Migrants, Athens.

WARREN, R. (1994),“Estimates of the unauthorized immigrant popula-tion residing in the United States by country oforigin and state of residence”, October, P.A.A.Meeting, April 1995.

YOSHIDA, C. (1996),“The global welfare of illegal immigration – no discern-ment case”, Indian Journal of Economics, Vol. LXXVI,No. 303, April, pp. 537-550.

OECD 1999

Trends in International Migration - OECD · sion and control policies. ... tion, administrative...

Documents

Transcript of Trends in International Migration - OECD · sion and control policies. ... tion, administrative...