Real Wages and Standards of Living in the Ottoman Empire
-
Upload
muadh-abdin -
Category
Documents
-
view
217 -
download
0
Transcript of Real Wages and Standards of Living in the Ottoman Empire
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 1/30
Economic History Association
Real Wages and Standards of Living in the Ottoman Empire, 1489-1914Author(s): Süleyman Özmucur and Şevket PamukSource: The Journal of Economic History, Vol. 62, No. 2 (Jun., 2002), pp. 293-321Published by: Cambridge University Press on behalf of the Economic History AssociationStable URL: http://www.jstor.org/stable/2698182
Accessed: 10/12/2009 04:08
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=cup.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Economic History Association and Cambridge University Press are collaborating with JSTOR to digitize,
preserve and extend access to The Journal of Economic History.
http://www.jstor.org
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 2/30
Real
Wages
and
Standards
of Living
in
the Ottoman
Empire,
1489-1914
SOLEYMAN
OZMUCUR
AND
UEVKET
PAMUK
Utilizing
a
large
volume
of archival
documents,
this
study
establishes
for
the
first
time the long-term
trends
in real wages
of
skilled
andunskilled
construction
workers
in Istanbul
and other
Ottoman
cities
in southeastern
Europe and
the Middle
East,
from
the second
half
of the fifteenth century
until
World War
I.
A
detailed
consumer
price
index
and nominal
wage
indices
are constructed
for the city
of Istanbul
for this
purpose.
These price
and wage
series
are then
inserted
into a
larger
framework
of
price
and
wage
trends
in
European
cities during
the
same
period.
uringthe last two decadeseconomistsandeconomichistorianshave
Jdevoted
much effort
o estimating
he
percapita
real product
of
differ-
ent countries,
and to
analyzing
trends
in the
gap
between
"leaders"
and
"fol-
lowers."'
We
know that
the
gap
is
large
today,
and
we can
infer from
the
growth
record
of both groups
of countries
that
it was
smaller-or
did
not
exist at
all-prior
to the onset
of
modern
economic
growth.
Recent
research
by
Angus
Maddison
and others
has confirmed
the
existence
of
a
gap
as
of
1820. We
have
also
learned that
the gap
between
developed
and
developing
countries as a whole widened not only in the era of the IndustrialRevolution
from
1820
to 1914,
but
also from
1914
to 1950.2
There
is little
information,
however,
about
the
period
before
1820.
How large
was this
gap
in 1750,
in
1600, or
in 1500?
These inquiries
inevitably
give
rise to questions
about
the
prevailing
trends
in
per
capita
incomes,
productivity,
and institutions
during
the early
modem
era,
not only
in
Western
Europe
but also
in
today's
devel-
oping
regions.
The
Journal
of Economic
History,
Vol. 62,
No.
2
(June
2002). ?
The Economic
History
Association.
All
rights
reserved.
ISSN
0022-0507.
Siileyman
Ozmucur
is Professor, Department
of
Economics, Bogazi,i
University,
Istanbul,
Turkey,
and
University
of
Pennsylvania,
Philadelphia,
PA 19104.
E-mail:
$evket
Pamuk
is Professor,
Department
of Economics
and
Ataturk
Institute
for Modem
Turkish
History,
Bogazi,i
University,
Bebek 80815,
Istanbul, Turkey.
E-mail:
This research
project
was
supported
by
the State
Institute
of Statistics
of
Turkey,
the Academy
of
Sciences
of
Turkey,
and
the
Bogazi,i
University
ResearchFund through
Projects
99HZ02
and
0
1
Z
0
O.
In addition
to these
institutions,
the authors
would
like
to thank
Figen
Ta?kln,
Nadir
Ozbek,
and
Bagi
Erten
for
valuable research
assistance.
Thanks
are also
due
to Robert
Allen for
sharing
his
data
and
insights. Earlierversions of this article were presented at the London School of Economics and at the
Third Conference
of the
European
Historical Economics
Society
held in
Lisbon,
October
1999.
The
authors
are
pleased
to
acknowledge
the helpful
comments
made
by
the
participants
of those
meetings,
especially
those by Jeffrey
Williamson
and Jan Luiten
van
Zanden.
They would
also like
to
thank
Jan
de Vries
and two anonymous
referees
for valuable
comments
and suggestions.
'
Abramowitz,
"Catching
Up";
Barro,
Determinants;
and
Maddison,
World
Economy.
See
also
Hanson,
"Third
World
Incomes";
and
Crafts,
"Gross
National
Product."
2Maddison,
"Comparison"
and WorldEconomy.
293
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 3/30
294
Ozmucurand Pamuk
With
the
possible
exception of a handful of countries,estimates of
per
capitaGDP for the period
before 1820
are
difficult
o construct
and unreli-
able. Moreover, t hasnot been possibleto constructdetailedestimatesfor
any
of the
developing
countries or
any period
before 1870. An
alternative
approachorstudying
he
gap
in
levels of
per capita
ncome or standards f
livinghas been to compare
real
wages
in
specific
occupations,
most
often
in
skilledandunskilled
constructionwork
n
urban
areas.Real
wage
dataare
of
far better
quality
han
per capita
GDP
estimates
especially
for
the
period
before
WorldWar
I,
and
they
are available or
a wider
range
of countries.
Infact, real wage seriesarevirtually he only solid
informationwe have on
standards f living in the developingcountries orthe periodbefore 1870,
if not 1914. In short,realwages continue o be the
most reliable source of
information bout
iving standards efore the veryrecentpast.
Although hey
cannot
be
described
as "national"
n
any sense,
urban
real
wage
series exist for many regions, and large
interregionaldifferences
within the same
country
are not
apparent
n
these series. Nonetheless,
real
wage
series
are
open
to valid
objections.
Even
if we
accept
the
representa-
tive
wage
as an
adequate
roxyfor
the
per capita
earningsof labor, his does
notmakeit a good proxyfor per capita ncome,which includes the returns
to
capital
and
and;
ongitudinal
nd
comparative
tudiesarefurther
ompli-
cated
by
the fact thatfactor
shareshave varied
greatly
over time and across
societies.
Moreover,
households ncomeswere often determined
y changes
in
employment evels,
and
in the
participation
ates
of
men, women,
and
children,and,
above all, by nonmarketncomes.
Despite
hese
qualifications,
he
historical
ink
between
wages
and he stan-
dardof
living
remains.
A
decline
n
realwages
did
result
n a decline n
house-
hold welfare,because each unit of laborcommanded ewer consumption
goods-or, alternatively,
ecausemore
abor
had to
be
supplied
o command
theaccustomed asket
of
goods.
Either
way,
hehousehold uffered
y
sacrific-
ing consumption,
onmarketncome, eisure,
r somecombination
f the hree.3
A
largepart
of our
knowledge
about
ong-term
rends
n
real
wages
dur-
ing
the
late
medieval
andearly
modem
periodsgoes
back
to
the vast
re-
search
ffort
undertaken y
members fthe
International
cientificCommit-
tee on Price
History (ISCPH),
established
n
1929. Most studies to
date,
including hose of theISCPH,have focusedon westernandcentralEurope.4
Arguably,
he most influential
project
n this
respect
has
been
that of
E.
H.
Phelps
Brown and
Sheila
Hopkins,
who constructed
rice
and
wage
series
for southern
England
over
seven centuries
beginning
n
1264.
The
Phelps
Brown-Hopkins
indings
werenotable ortheir
pessimism.They
confirmed
3Van Zanden, "Wages"; and de
Vries, "Between Purchasing
Power."
'
Amongst the classic studies
were those by Beveridge ("Wages"), Hamilton (American Treasure),
and Phelps
Brown and
Hopkins ("Seven
Centuries" and
Perspective).
For an overview of
the
main
results, see Allen, "Great Divergence."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 4/30
Ottoman
Living
Standards
295
long-term
luctuations,
ut
did not
revealany
secular mprovement
n
urban
real wages
from
the fifteenth
o the mid-nineteenth
entury.'
n recent
de-
cades,however,evidencehasbeen accumulatinghatproductivity,tandards
of living,
and per
capita
GDPwere
rising in many
parts
of western
Europe
before
he
Industrial
Revolution.6
or his
reason,
researchers
n
recentyears
have
been reviewing
the old series
and
constructing
new ones
in order
to
reassess
the
history
of real wages.
Theyhave
alsobeen
debating
how
well
the real
wage
of urbanconstruction
workers
can
serve
as a proxy
for the
living
standard
f the population
at
large,and
whether
more representative
indices can
be constructed
with the available
data.Changes
n
occupational
structure ndincomedistributionmayhavebeentwo important ndrelated
reasons
behind
the
divergence
of GDP
per capita
and
realwages
in
western
Europe.
n
other
words,
t is
possible
that
he
wages
of
construction
workers
declined
elative
o other ncomesduring
hisperiod.
In addition,
households
may
have increased
heir
abor
supply
during
his
period.7
In contrast
o ouradvancing
tate
of
understanding
f
Western
Europe,only
a few real wage
series have been
constructed
o
date
for
cities
on the
Euro-
peanperiphery
and in other
continents.
The Ottoman
Empire
stood at
the
crossroads f intercontinentalrade, tretchingrom heBalkansand heBlack
Sea
through
Anatolia,
Syria,
Mesopotamia,
he
Gulf, Egypt,
and most
of
NorthAfrica
from
the
early
sixteenth
enturyup
to World
War
I.
At its
peak
in
the
seventeenth
nd eighteenth
enturies,
ts
population
xceeded
30
mil-
lion. One
might
have expected
hat
the economic
nstitutions hat
sustained
this
large,
multiethnic
entity
for so long
wouldbe
of interest
o economic
historians.
Unfortunately,
mainstream
economic
historians
have long
neglected
the
land
regime,
manufactures,
conomic
policies,
and the
daily
existenceof ordinarymen andwomen.As aresult, helongevityof the Otto-
man
Empire
remainsananomaly
and even a
mystery
or
many.
For
most
of its
six-century
xistence,
he
Ottoman
Empire
s best charac-
terized
as a bureaucratic, grarian
empire.
Its
economic
institutions
and
policies
were
shaped
to a
large
degree
by
the
priorities
andinterests
of
a
central
bureaucracy.
Recent literature
has
emphasized
that
the Ottoman
empire
did not
simply
and irreversibly
decline
after
the
sixteenthcentury.
On
the
contrary,
he
Ottoman
tate
and
society
showedconsiderable
ability
toadapt o changingcircumstancesn Eurasiaromthe seventeenth hrough
the nineteenthcenturies.
The central
bureaucracy
managed
o contain
the
many
challenges
it faced
with
pragmatism,
lexibility,
and
a
tradition
of
negotiation
o co-opt
and incorporate
nto
the state
the social
groups
that
rebelled
against
t. The Ottoman
tate also
showed considerable
lexibility
'Phelps
Brown and
Hopkins,
"Seven
Centuries"
and Perspective.
6
Maddison,
WorldEconomy;
Persson,
Pre-Industrial
Economic
Growth;
and de
Vries and van
der
Woude,
First Modern Economy.
7Van
Zanden, "Great Convergence," p. 12; and de Vries, "Industrious Revolution."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 5/30
296
Ozmucurand Pamuk
not only
in
military
echnology,but
also
in
its
fiscal, financial,
andmonetary
institutions.Despite
wars and internal
conflict
from the 1770s through he
1830s, the Ottomansmanaged o regroupandsurvive into the modem era
with a
strong
central tate
and
many
of their
central
nstitutions
ntact,
while
many
of
their contemporaries
n
both Europe
andAsia collapsed.
Even before
the nineteenth
century,
he
Ottoman economy
was
well
linked
o
Europe
via
tradeacross heMediterraneannd
he
Balkans.Thanks
toresearch
n
recentdecades,
we now
have a
reasonably ood pictureof the
long-term
rends n
economic activity
n the OttomanEmpire.Most eco-
nomic historiansagree
that
the
sixteenthcenturywas a periodof demo-
graphicand economicexpansion,at leastin thecore regionsof the empire.
In contrast, he seventeenthcenturyhas been
characterized s a period
of
"crisis
and
partial
recovery"when population
and economicactivitystag-
nated
(or worse)
in
many parts
of the
empire.Contrary
o earlier
assump-
tions,
it
has been shown
that
theeighteenth entury
was a
period
of
relative
peace, stability,
and economicexpansion
hat
came to
an
end
with
the
de-
cades of
war,
fiscal
difficulty,
and
inflation
after
1770. Between 1850
and
1914,
the overalltrendwas
once
again
owards
imited
ncreases
n
average
incomes. Despitethe accumulation f valuableevidence for different ime
periods,
we are still
unable
to
compare
evels of per capita ncome in 1914
with those
in
1750 (or 1600
or
1500),
for
different
egions of the Ottoman
Empire
or for the
empire
as
a whole.8
Utilizing
a
large
volume of archival
documents,
his
study
will
establish
the
long-term
rends
n
wages
of
skilledandunskilledconstructionworkers
in
Istanbul
and other
Ottoman
ities
(including
Belgrade,Salonika,Edirne,
Bursa,Damascus,andJerusalem)rom he late
fifteenthcenturyup to
World
WarI. These trendswill offerimportantnsights nto the long-termperfor-
mance of the
Ottoman
economy. The price
and wage series will then be
inserted
nto a
larger
ramework
f
price
and
wage
trends
n
European
ities
during
he
same
period.
This
comparison
will
provide
new
and
important
evidence
regarding
tandards
f
livingandgrowth
n
the early
modem era.
It
will
also shed
light
on the
origins
and
evolution of a
gap
in
real
wages
between
developed
and
developing
countries.
PRICES
Sources
and
Methods
We
begin
with a
summary resentation
f the
results
of
a
study
of
prices
n
Istanbul ndother
eading itiesofthe Ottoman mpire rom hefifteenth
o the
twentieth
enturies.Additional etails
bout he ndicesare
presented
n
Appen-
8
Inalcik
and Quataert, conomic and Social History;or a recentoverview see Pamuk,Monetary
History,pp. 228-29.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 6/30
Ottoman
Living
Standards
297
dix 1. Our
studyutilized
dataon theprices
of standard
ommodities food
and
nonfood
tems)
collected
rom
more
than6,000
account
books and
price ists
locatedn theOttoman rchivesnIstanbul. hefirst tepwasto constructhree
separateood-price
ndices.One
of these
wasbasedon
the accountbooks
and
prices
paidby
the
manypious
foundations
vaklf),
both
largeand small,
and
their
oupkitchens
imaret);
nother
ndexwasbased
on the account
books of
theTopkapi
alacekitchen;
thirdutilized
he
officially
stablished rice
ceil-
ings
(narh) or the
basic items
of consumption
n
the capital
ity.9
To the extent
possible,
standard ommodities
were used
in constructing
the
indices
in
order
to minimize the effects
of
quality
changes.
The
food
indices included the prices of ten leading items of consumption:flour
(mostly
wheaten),rice, honey,
cooking oil,
mutton,
chickpeas, entils,
on-
ions, eggs,
andfuel-grade
olive oil.
Amongthese,
flour,
rice, cooking
oil,
mutton,
olive
oil,
and
honeyprovided
he
mostreliable
ong-term
eries
and
carried
he highestweights
in our food budget.
In
cases
where the
prices
of
one or
more of these items
were not available
for a given year,
missing
values
were estimated
by an algorithm
hatapplied
regression
echniques
o
the
availablevalues.
The medium- and long-termtrendsexhibitedby the three food-price
indices are
quite
similar.
In
addition,
accountbooks showing
the private
purchases
by the
householdsof high-ranking
ureaucrats
lso point
to simi-
larlong-term
rends
n food prices
at the
retail level.10
The fact
that narh
prices
closely approximated
rices
paid by
the palace,pious
foundations,
and
private
ndividualssuggests
thatgovernment
ntervention
did not
alter
prices,
especially
in
the medium
andlong term.
It also pointsto
a high de-
gree
of government
pragmatism
n
economic
affairs.Historians
have
long
argued hat the permanentapplicationandenforcementof narh in urban
areas
was a
typical
example
of Ottoman
nterventionism
and
rigidity
in
defense of a traditional
order.
There
s,
however,
a
good
deal
of evidence
that he Ottomans
became
ncreasingly
onscious,
after
1600,
of
the
limita-
tions
of interventionism
n
economicaffairs.
They
earned hat
price
ceilings
9'Most of
the account
books
of the pious foundations
used
in this study are available
from
the
Maliyeden Miidevver
(MM), Bab-i
Defteri,
Ba?muhasebe
Kalemi
(B6M),
Evkaf Defterleri (EV),
and
Evkaf Nezareti, Haremeyn Muhasebeciligi (EV.HMH) Collections at the Prime Ministry Ottoman
Archives
(Ba?bakanlik
Osmanli
Ar?ivleri
[BOA])
in
Istanbul.
The account
books of the
Topkapi
Palace kitchen
are
available
from
the Bab-i
Defteri, Matbah
Emini (D.B$M.MTE),
Bab-i
Defteri,
Baamuhasebe
Kalemi (D.B$M)
Maliyeden
Miudewer (MM),
Kamil Kepeci
(KK),
Cevdet Saray (CS),
Maliye Nezareti,
Mesarifat
Muhasebesi (ML.MSF),
Bab-i
Defteri, Odun
Emini (D.B6M.ODE)
and
Bab-i Defteri, Istanbul
Agasi (D.B6M.ISA)
Collections
also at the
BOA, and the Dosya
(D)
Collection
at the Topkapi
Palace Archives.
The
narh lists were approved
by the local judges
(kadi) and
were taken
from
the Istanbul, Uskudar
and Galata Islamic court archives
in Istanbul.
A
complete
list
of
all
archival
documents used
in this
study
is
presented
on
year-to-year
basis
in the
appendices
of Pamuk,
Istanbul.
' Annual
values of these four
food-price
indices are not
provided
here
due
to limitations
of
space.
For further details see Pamuk, Istanbul, figures 2.2 and 2.3.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 7/30
298
Ozmucur
and
Pamuk
thatdiverged ubstantially
rom he underlyingmarket ealities
couldnot
be
enforcedover long periods
of time.Forthisreason, nterventionism ecame
increasinglyselective. It was used primarily or the provisioning of the
capitalcity and the army,and for selected
commodities.Perhapsmore im-
portantly, fter1650
the
narhcame
to be considerednot aspermanent olicy
butan nstrument eserved
or
extraordinary
onditions uch as
wars,excep-
tional
difficulties
n the provisioningof
the
capitalcity,
or
periods
of
mone-
tary nstability.With
ncreasing iscal andmonetary
difficultiesandgrowing
price instability, he price
ceilings once againbecame
a regularfeatureof
Ottoman
urban ife from 1789 to
1850.11
Because he
palace
and
narh
prices
might be consideredofficial or state-controlledprices, this study gives
greaterweight
to the
prices
paidby
the
soup
kitchensand
the
pious founda-
tions.
The narh
prices
andthe retail
pricespaid
by
the
households
of
high-
level bureaucratswere not included
n the
final
index.
Since
the
availability
and
quality
of
price
observations
aried over
time
for most of the foodstuffs
n
our ist,theperiod1460-1860 was
divided nto
five
subperiods
and indices were calculatedseparately
for
each.
In each
subperiod
ome commoditieshad to be excludedfrom
the index due
to the
unavailability f priceobservations.
Based
on the
availableevidence
regarding
he
budget
of an
average
urban
consumer,
he
weighting
of food items in the overall indices was fixed be-
tween 75 and 80 percent.
The weightingof eachcommodity n the overall
index was then based
on the sharesof each in totalexpendituresof the re-
spective nstitutions.
Tocite two prominent xamples, he weightingof flour
varied
generally
between 32 and
40
percent
of food
expenditures
nd
24 to
32
percent
of
overall
expenditures, epending
on the fluctuations
n
prices.
Similarly, heweightof muttonvariedbetween 5 and 8 percentof the over-
all
budget.
Diets
of
private
households
n
the
capitalcity probably
differed
from those offered by the soup kitchens.At this
stage, however,
it is not
possible
to
approximate
he
private
diets more
closely.
The
following
commodity
basketwas used most
frequently
n
the
food
price
index:
11
kile
of flour
(288.6 kg)
4 kileof rice (51.2 kg)
9 okka
of animal-based
ooking
oil
(11.5 kg)
13 okka
of
honey (16.6 kg)
40 okka
of
mutton
51.2 kg)
0.2
kile
of
chickpeas
2.56 kg)
5 okkaof olive oil
(6.4 kg).
"
Pamuk, Monetary History, pp. 9-19.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 8/30
OttomanLiving Standards
299
One
mportant
eason
why
the
weightings
of the individual ommoditieswas
kept constant over time was the unavailability
of consumerbudgets
for
differentpointsin time. It is likely,however, hatchanges n the commodity
compositionof the consumerbudgetwere small,
in partbecause
increases
in
average ncomes
were limitedduring hesecenturies.On the other
hand,
some
substitutionmust have occurredover time, away
from commodities
whose
prices rose
faster towardsthose whose prices declined
in relative
terms.It is thusreasonable
o expectthatourfixed-weights
ndex overstates
the
long-termprice
increasessomewhat,as
will be further
discussed
below.
In
the secondstep,various
nonfood
prices(of
soap,wood, coal,
andnails)
obtainedfrom a varietyof sources, most importantly he palace account
books, were addedto the price indices. From
thevariousaccountbooks of
the imperialpalaceit was possible
to obtain
ong-termprice
series on
two
types of woolen cloth:
the locally produced
Vuha,
and the
Vuha
Londrine
imported
rom
England.
However,
because the woolen cloth
described
n
palace
sourcesbelongedto moreexpensivevarieties,
cloth
prices
were not
included
n
the overall index
until
1860.12
Price data for
many
othertypes
of cloth have also been
collected,
but none of these are available
for long
periodsof time.A cost-of-living ndex shouldalso includethe rentalcost of
housing,
but an
adequate
eries for standard
ousing
is not available.
It is well known thatpremodemprices exhibited
wide short-term luctu-
ations due to harvest
conditions,
difficulties
in
transportation,
wars,
and
other causes.
In
the case of our
indices, however,
a number of factors
worked
to reduce such fluctuations.
First,
the
capital city
was
subject
to a
higherdegree
of
government
nvolvement
n
price formation.Secondly,
it
is
likelythatsoupkitchens andthe imperial
palacewere subjectto
a lesser
degreeofpricevariation hanwereprivatehouseholds.Thirdly, nprocess-
ing price observations rom archivaldocuments
we excluded prices
more
than 100
percent
higher or more than 50 percent
ower than the
value of
the
previous year,
unless the new price
level
was
observed
again
in later
years. Although most
of the excluded observations
reflected errors
n the
originaldocument,some of
them must
have
represented
actual
prices.
As
a result
we
do not have a
high degree
of confidence
in the
year-to-year
fluctuationsof our
indices,
but we are
quite
confident
about
the level of
pricesin the medium and longer term.
For
the
period
1860-1914,
data romthe
palace, vakif,
and
narh
sources
are
very
limited. For this reasonwe used
the detailed
quarterly
wholesale
prices
of the
CommodityExchange
of
Istanbul, overing
about
two
dozen
commodities,
and a separateseries
for
imported
cotton textiles. Indices
based
on
these
prices
were
then
linked
to those for the earlier
period,
with
2
We compared the
woolen cloth prices paid by
the palace with the prices of more ordinary varieties
worn by the janissaries, as given in Sahillioglu, "Yeni9eri (;uhasi."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 9/30
300
Ozmucur
and
Pamuk
the help
of
detailed
data orboth
retail
prices
of individual ommodities
and
prices
at
the Commodity
Exchange
or 1860-1862
and
1913-1914.
Results
We have
thus obtained,
or the firsttime
forsoutheasternEurope
and
the
Near East,
detailed
andreliable
price
series for these four-and-a-half
entu-
ries
(Table
1).
The indices
show thatconsumerprices
in Istanbul
ncreased
by
a
total
of
about
300-fold from
1469
until WorldWar
I.
This overall
in-
crease
corresponds
o anaverage
ncrease
of 1.3
percent
per year,
endpoint-
to-endpoint,and1.1percentperyearfor the fittedline. They also indicate
that
Istanbul
xperienced
a
significant
wave of
inflation
rom
the
early
six-
teenth
o
the late
seventeenth
entury,
when
prices
ncreased evenfold.
This
is the
period
usually
associated
with the Price
Revolution.
The indices also
show,
however, hat
here
occurred
much
stronger
wave
of inflation
begin-
ning
in the 1770s
and lasting
into the
1860s, when
pricesrose
as
much
as
25-fold.
Most of
the
latter ncreases
were associated
with the debasements
that
began
in
the
1780s
and accelerated
during
the
reign
of
the reformist
sultanMahmud11 1808-1839). In contrast,priceswere relativelystable
from
the 1690s
to
the
1760s,
and
from he 1860s
to
1914.
(For
a
period-by-
period
review
of the
price
trends,
see Appendix
1.)
Obviously
there
were
many
causes
of inflation
during
he
early
modern
period,
as evidenced
by the large
literatureand the
extensive
debates
on
the
subject.
From
the
long-termperspective
offered
by
theseprice
indices
and
by
$evket
Pamuk's
recent study
of
the
Ottoman
currency,
however,
it
is clear that
debasements
or the
reduction
of
the
silver content
of
the
unit
of account by the monetaryauthoritieswere the most importantcause of
Ottoman
price
increases.13
In
fact,
even
though
nominal
prices
increased
about
300-fold
over these four-and-ahalf centuries, prices
expressed
in
grams
of silver rose
less than
threefold
(Table
1). There
were
medium-
term
movements in
the
latter
series: it
increased
from
1500
until
1640
during
he
period
of the PriceRevolution,
declined
until the
early
decades
of the
eighteenth
century,
and ncreased
againuntil
the middle
of
the
nine-
teenth century.
These
fluctuationsoccurred,
however,
around
a secular
trend thatwas rising only modestly. The so-called silver inflation thus
contributed
o the
changes
in
the
overall price
level,
but its
impactpaled
in relation
to that of
debasements.
Istanbulwas
chosen
primarily
ecause
pricedata
aremost detailed
or the
capitalcity.14
However, price data
from
the account
books
of pious foundations
13
Pamuk,
Monetary History.
14
Detailed indices on the prices of basic foodstuffs expressed in grams of silver, and the
terns
of
trade between foodstuffs and manufactured goods,
have
also been calculated from the Istanbul data
(Pamuk, Istanbul, ch. 2).
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 10/30
Ottoman
Living
Standards
301
TABLE
1
PRICESAND WAGES
N
ISTANBUL,
1469-1911
(decennialaverages)
Consumer
Price NominalDaily
Wages
Index
Real
Daily
Wages
(1469
=
1.0)
UnskilledWorker
SkilledWorker
(1489/90
=
1.0)
Period InAk,e
In Silvera
In
Ak,e
In
Silvera
In
Ak,e
In
Silvera
Unskilled Skilled
1469-1479 1.20
1.17
n.a. n.a.
n.a. n.a.
n.a. n.a.
1480-1489
1.30
1.02 4.9
3.3
9.8 6.6 0.93
1.04
1490-1499
1.09
0.85 4.7
3.1 7.7
5.2 1.06 0.96
1540-1549
1.60
1.00 5.3
3.5
9.6 6.3 0.86
0.92
1550-1559
1.64
1.25 5.0 3.3
9.7 6.4
0.78 0.86
1560-1569
1.86 1.32
5.9
3.8 9.3
6.0 0.83
0.73
1570-1579
1.99
1.41
5.6
3.4
9.9
6.0 0.68
0.66
1580-1589
3.35 1.47
8.1 3.5
12.4
5.4 0.65 0.53
1590-1599
4.45 1.46
11.7 2.6
20.7 4.6
0.51 0.50
1600-1609
5.43 1.81
13.9 4.0
22.5
6.5
0.62
0.56
1610-1619
4.44
1.47
14.2 4.1 22.8 6.6 0.76
0.67
1620-1629
5.56 1.54
15.0
3.4 18.3
4.1 0.84
0.57
1640-1649
4.37
1.34
15.2
4.2
29.5
8.2
0.79
0.84
1660-1669
5.82
1.56
19.5
4.4 29.9
6.8
0.81
0.69
1670-1679
7.06 1.69
18.0
3.7 23.0
4.8 0.71
0.50
1680-1689 7.46 1.79 21.8 4.5 33.4 6.9 0.73 0.62
1690-1699 7.96
1.21 26.2
3.4
36.9
4.8 0.78
0.61
1700-1709
6.05 0.92
23.8 3.1
38.0 5.0
1.02 0.90
1710-1719
6.87 1.03
25.2 3.3
38.2 5.0
0.92 0.77
1720-1729
6.98 1.06
26.4
3.5 42.3
5.6 0.97 0.85
1730-1739 7.62
1.10 30.6
3.8
44.8
5.6 0.94 0.76
1740-1749 9.93
1.39
32.5 3.9
51.1
6.2
0.91 0.79
1750-1759 10.1
1.30
32.4 3.6 58.6
6.6 0.76 0.75
1760-1769
11.1 1.23 30.3 2.9
50.8 4.8 0.70 0.64
1770-1779
17.8
1.93 38.7
3.6
71.2
6.7 0.52
0.51
1780-1789
17.2
1.58 56.9
4.7 113.4
9.4 0.86 0.94
1790-1799 24.5 1.50 80.2 4.1 148.4 7.7 0.84 0.84
1800-1809 34.7
1.91
114.4 5.3 217.7
10.0 0.84
0.88
1810-1819 49.5
1.86
202.6
6.9
401.2 13.7 0.95
1.04
1820-1829
62.8 1.26 275.9
5.0
475.4
8.9
1.13
1.10
1830-1839
130.1
1.12 608.8 4.5 1,054
7.8
1.11
1.05
1840-1849 181.0 1.70
717.2 5.9
1,238
10.1
0.99
0.95
1850-1859
240.5
2.32 935.3 7.8 1,599
13.3
1.38 1.30
1860-1869
285.7 2.76 964.7
8.0 1,716 14.3
0.81 0.80
1870-1879
265.6
2.56
941.4
7.8 1,989
16.6
0.86 0.99
1880-1889
238.1 2.30
944.0 7.9 2,061
17.2 0.91 1.04
1890-1899 235.6 2.27 1,063 8.9 2,308 19.2 1.12 1.35
1900-1909
220.9
2.13 993.5
8.3
2,150
17.9 1.10 1.31
1910-1914
294.2 2.84
1,861
12.6
3,450
26.6 1.30
1.51
a
Grams.
Sources: See the
text.
(similarly
btained
rom
he account
books of
pious
foundations
n
the Otto-
manarchives
n
Istanbul)
areavailable or
other
cities of
the
empire
as well.
Observations
or
a shorter ist
of commoditieswere used to construct
epa-
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 11/30
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 12/30
OttomanLiving Standards
303
long-term
rends
n rental
prices
of
residential
housing
elsewhere
n
Euro-
pean cities
for which dataare available.
n
view of Istanbul's
growthas an
urbancenter,one should expectrents on residentialhousingto have risen
over
time.
Taking
nto accountpatterns
n housingownership,
we assumed
as a firstapproximation
hat rent
paymentson urbanhousing
doubled
in
relation o the overall
ndex, andthat
he shareof rents n the
averageurban
budget
rose fromabout
5
percent
o about10
percent,
betweenthe sixteenth
andearly
wentiethcenturies.
The inclusionof
rents nour ndex undersuch
an assumptionwouldincrease he
overall ndex
by 5 percent, herebyoffset-
ting
the inclusion
of cloth prices.In the absence
of anymeaningful time
series dataon rents,it is difficult o be more specificon this issue.
Finally,
we consider
he inclusionof
sugar
and
coffee
in our
price
index.
Istanbul
sugar prices declined by
about 50 percent
relativeto the overall
index
during
he sixteenthcentury,
and
by
an
additional50
percentduring
the period 1820-1860.
The available price
series for
coffee in Istanbul
tracked
the overall
index until the nineteenth century,
but declined by
about60
percentrelative
o
the
index during he period 1820-1
860.20Since
the
share
of
these two items
in
the
averageconsumerbudget
was limited
until 1860, however, our overall results would not change significantly.
We
estimatethat the inclusion of
these two items
in the
consumerbudget
would reduce the
terminal value of the
index
in
1914
by
no more
than
1
percent.21
RelativeCosts ofLiving
Thus
the
cumulative mpact
of
including
cloth, sugar,coffee,
andrents
n
the consumerbudgetwould be quite small. The impactwould be much
greater
n the relativecosts of living
of lower- andhigher-income
onsum-
ers,
however.
Since
cloth, sugar,
andcoffee
had
larger
hares n
the
budgets
of
higher-income
onsumers,
he decline
in their
prices
reduced
heir cost
of
living
in
particular.On
the
other hand,
the inclusion of
rent
payments
would tend to
raise
the costs
of
living
of lower-incomegroups
more
than
those facedby higher
income
groups,
since
the latterwere
likelier
to own
their
dwellings.
We thus estimate
that the inclusion
of these four items
would shiftrelativepricesin favorof higher-income onsumers.Withre-
spect
o
the two
groups
whose incomes
will
be
examined
n
the
next
section,
for
example,
we
expect
that the
expanded
price
index would widen
the
20
Before
the eighteenth century, sugar and coffee
were brought to Istanbul from Egypt. It
was the
arrival of imports
from Atlantic sources that
led
to the
sharp decline in the prices of both commodities
during the nineteenth century.
21 For the period 1860-1914 our
index includes prices of imported cotton cloth, sugar, and
coffee.
All price observations
used
in
this study, as well as the indices constructed
with them, are presented
in Pamuk, Istanbul.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 13/30
304
Ozmucurand Pamuk
difference
between
the
purchasing
power
of skilled
and unskilledworkers
by
about
10
percent
or the entire
period.
Oneimportant ndrelatedcharacteristicf the Istanbulpriceseriesis the
important
ivergence
n relative
prices
over the course
of these
four-and-a-
half
centuries.
Prices of some foodstuffs
meat,
flour,
milk, eggs)
rose
400-
to600-fold
in
nominal
erms
and
our- o sixfold
in
silverterms)during
his
period,while prices
of other oodstuffs olive
oil, honey,
coffee,onions)
and
all
manufactured
tems for which
price
series are available
rose less than
200-fold. Since
the commodities
with
higher
rates of
price
increaseswere
essentials, hey
had a
higher
share nthe
budgets
of
lower-income
consum-
ers. Evenafterallowingforsomesubstitution wayfromcommoditieswith
morerapidprice
ncreases,we estimate
hatasa resultof these relative
price
movements
he overall
rate of inflation
aced
by
lower-incomeconsumers
was
higher,
and
he overall nflation aced
by higher-income
onsumers
was
lower,
thanthe
averagesprovided
by
our index.
If
we
had
adopted
eparate
consumer
budgets
for skilled and unskilledworkers,
for
example,
these
movements
n
relativeprices
would
have led, roughly,
o
cumulative
price
increasesthat were
5
percent
higher
for unskilled workers
and
5
percent
lowerfor skilledworkers.
When
we combine hese different
ffects,
we
arriveat a
cumulative
differ-
ence
of
about
20
percent
between
the
prices
faced
by
skilled
and
unskilled
workersduring
hese four-and-a-halfenturies.
n other
words,
we
estimate
thatthe overall
rate of inflation aced
by
unskilled
workers
was 10 percent
higher,andby
skilledworkers
10
percent ower,
thanthe averagesreported
in
our
consumer
price
index
in
Table
1.22
WAGES
For the second
part
of this
study, daily
wage data
were
gathered
from
morethan
5,000
accountbooks
for construction
nd
repairprojects
n
Istan-
bul
and
other
cities. These accountbooks contain
daily wages
for both
un-
skilled
and a
variety
of
skilled
construction
workers.
Urban
construction
work
was
a
relativelyhomogeneouscategory
of
labor
over time and
space.
Moreover,
n
contrast
to the
payments
made
to
other
employees,
urban
constructionworkersreceivedmuchof theirpayin cash ratherhan n shel-
ter,
food,
and
clothing.
As
a
result,
their
wages permitcomparisons
with
other
preindustrial
ocieties.
The construction ccountbooks
prepared
y
the
state
or
by pious
founda-
tions
usually
consisted
of a series of attendance ecords
isting
the work-
22
A recent study of prices
and inequality
in
Europe since
1500
(Hoffman
et al., "Real Inequality")
similarly argues
that relative price movements
favored higher-income groups and
increased income
inequality in Westem Europe from 1500 to 1800.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 14/30
Ottoman
Living Standards
305
man's
name,
his
craft,
his rank
(master,
common laborer,
etc.)
and
his
wages.
Sometimes
the accountsprovide
a separate
record
for each day;
sometimesseveraldaysor weeksarecoveredby asingleattendanceheet.23
Information bout
the
length of theworkday
s rare in these records,
as is
information
n theprovision
of meals. For thatreason,we have
chosen to
ignore
those aspects
of the
daily
wage.
We
have also
decided to
ignore
the
seasonal
variations
n daily wages. In any
case, the overwhelming
majority
of the available
observationsbelong
to the construction
season (April
throughOctober
n Istanbul).
The
wages
for unskilled
workersreferredmostly
to one
type
of worker,
called rgadintheearlyperiodandren berafterabout1700. Incontrast,he
accountbooks
revealdailywagerates
ormorethan
a half-dozen
categories
of
skilled
construction
workers.
n
order o utilize
additional
nformation,
n
index
was constructed
ncluding
the wages of carpenters,
abinetmakers,
masons, stonecutters,
ditch-diggers,plasterers,
and
others. Based
on its
predominance
n
the
accountbooks,
the shareof carpenters
neccar)in the
index fluctuated
between 50
and 60
percent.
Figure
1
and
Table
1
present
realdailywage
series for skilled and
un-
skilled constructionworkers n Istanbul,obtainedby deflating he nominal
daily
wage
series
by
the
consumerprice
ndex.
It is not
easy
to
judge
to what
extent
daily
wages
were influencedby
institutional actors,
and to what
extent
by
market forces. However,
the fact that during periods
of
rapid
debasement
eal
wages initially
declined,
but soon
were
adjusted
upwards,
suggests
that
the
process
of
wage
formation
was
open
to market
orces.
Our
indices indicate that real
wages
of unskilled construction
workers
declined
by
30 to 40
percent
during
he sixteenthcentury.
They
then
re-
mainedroughlyunchangeduntil the mid-eighteenthcentury,afterwhich
they
increased
by
about
30
percent
up
to the
mid-nineteenth
entury,
and
then
by
another
40
percent
during
he late nineteenth
and
early
twentieth
centuries.On the
eve of
World
War
I,
real
wages
of
unskilled construction
workers
were
about
20
percent
above
their levels
in
1500.
But because
relativeprices
had shifted
in
favor
of
goods
consumed
by higher-income
consumers
during
hese
centuries,
andbecause he skill
premium
had
begun
to
rise
late
in
the nineteenth entury, y
1914 the
real
wages
of skilled
work-
ers stood at more than50 percentabovetheir evels in 1500.
23
Most of the account
books of the pious foundations
and the state used
in
this study
are
available
from the Maliyeden Mudewer
(MM), EvkafNezareti,
Haremeyn Muhasebeciligi (EV.HMH),
Evkaf
Defterleri (EV),
Bab-i
Defteri, Baamuhasebe Kalemi (D.B$M),
Bab-i
Defteri,
Bina Emini
(D.B$M.BNE)
and Maliye Nezareti, and Mesarifat
Muhasebesi
(ML.MSF) collections in the BOA
archives
in
Istanbul. Pamuk (Istanbul,
appendices
7-10) provides a full list of the archival
sources
on
an annual basis, as well as the
nominal daily
wages for each wage category in the
form of annual
averages for Istanbul and other
cities. Many of
the account books also included
lists and prices of
materials purchased,
such as iron, lime, and
nails, the last of which
was utilized
in
the construction
of
the price indices.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 15/30
306
Ozmucurand Pamuk
3.5
Skilled
3.0 x Unskilled
30-Year Moving Averages,
Skilled
2.5
-
30-Year Moving Averages, Unskilled
2.0
1.0
0.5
Nx
x
#
0.0
.. . . . . .. . . . . .
1450 1500 1550 1600 1650 1700 1750 1800 1850
1900
FIGURE
REAL DAILYWAGEOF CONSTRUCTIONWORKERS: STANBUL, 1489-19 14
(1489/90 [unskilled] 1.0)
Sources:See the text.
When we take into account he differential mpactof the items excluded
from our price indices, as well as the differential mpactof the more rapid
rise in the prices of essentials,we mustadjust he cumulative osts-of-living
increasesof skilled andunskilledworkersby a further 1 0 and+10 percent,
respectively. (See
Appendix
1 for a period-by-period analysis of the trends
in realwages.) These adjustmentso the series in Figure 1 and Table 1 give
a betterview of trends n the wage premium or skilled labor.After declin-
ing duringthe sixteenth and seventeenthcenturies,the wage differential
began to increase n the second half of the eighteenthcentury,reaching
ts
peak on the eve of WorldWar . Not only changingdemand
but also decline
in supply(due to the emigration f skilled construction rtisans)must have
contributed o this trend.
The purchasing ower of the dailywages of both classes of workerwere
reasonably igh during hese four-and-a-halfenturies.During he sixteenth
centuryanunskilledconstructionworker ouldpurchasewith his dailywage
8 kg of bread, or 2.5 kg of rice, or more than 2 kg of mutton. The daily
wages of skilled workers were 1.5 to 2 times higher. At these levels of pay,
skilled
construction workers must have enjoyed standards of living well
abovethe average or the populationas a whole, and
also
above
the
average
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 16/30
Ottoman
Living
Standards
307
of
the urbanareas,
even
if
they
worked
fewer
than 200 daysper year.
For
purposes
of international omparisons,
t is also
possible
to
express
daily
wages in Istanbuland otherOttomancities in gramsof silver simplyby
multiplying
he
nominal
wages
by
the
silver
contentof the Ottoman
ak
e.
Around1500,
for
example,
daily
wages
of unskilled
construction
workers
in
the
capitalcity
equaled3.4 grams
of
silver;
hose
of skilled
construction
workers
equaled
6.6
grams
of silver (Table
1).
Data
were also
collected
on
thedailywages
of constructionworkers,
both
skilled
andunskilled,
n other
Ottoman
ities
around he eastern
Mediterra-
neanand the
Balkans
forthe
same period
1490-1914.
These
observations
were obtainedfrom the accountbooks of pious foundationsoperating n
these cities,
and
are available
romthe
Ottoman
archives n Istanbul.
They
showclearly
thatnominalwages
in
otherOttoman
ities also
increasedby
more than
300-fold
during
his
period.24
In orderto establish
whether
there
was
any
relationship
between daily
wages
in
the
capitalcity
and
hose
in
second-tier
urban enters
of the empire
(Bursa,
Edirne,
Belgrade,
Salonika,
Damascus,
erusalem,
nda
few others),
we regressed
w1
wo
(the
ratio of
nominal wages
in Istanbul
o nominal
wages in othercities) on the timetrend,usingdatafor each yearin which
observations
were
recorded or both
Istanbuland
any
other
city.
Our sepa-
rateregressions
or
the
wages of
skilled andunskilled
workers,as
well as a
composite
wage
definedas
a
weighted
combination
f
the
two,
indicate hat
the
null
hypothesis
of an intercept
of 1.0
for
w,
/
wo
cannot be rejected.
Moreover,
the
slope
coefficient
is
not significantly
different from
zero,
indicating
no discernible
ime trend
n this ratio.
SHORT-AND LONG-TERMDETERMINANTSOFREAL WAGES
In
the short
and
medium erm,
hemost mportant
etenninants
f Ottoman
realwages
were changes
n
prices
andthe
speedwith
which
nominalwages
adjusted
o them.
Because his adjustment
as
slow,price
shocks
ed to
fluc-
tuations
n real
wages.
Consumer
prices
exhibited
arge
short-term
luctua-
tions
due to harvestconditions,
difficulties
n
transportation,
ars,
andother
causes.
The
most
important
ause
of
changes
n the
price
evel,however,
was
debasement.Evenif pricesdid not risequickly,becauseof a specieshortage
or someother
eason,
ong-distance
radeactedas
theultimate
qualizer.
rice
adjustments
fter
a
debasement
ended o be
more
rapid
he more
open
the
economy,
and
the more
frequently
ebasement
was resorted
o.
Recentresearch
has shown
that he urbanpopulation
and some
segments
of
the
countryside
were
alreadypart
of the money
economyby
the
end of
the
fifteenth
century.
Even
more
significantly,
here occurreda
substantial
n-
24
Pamuk, Istanbul, figures 5.1-5.4 and appendix table 9.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 17/30
308
Ozmucurand
Pamuk
crease
n the use of money during
he sixteenth
century,
because of both the
increasedavailability
f specie and
he commercializationf
the
ruralecon-
omy.The evidencefor thisimportant evelopment omesfrom a numberof
sources. First, recent
researchhas pointedout thatpopulationgrowth
and
urbanization uring
he sixteenthcenturywereaccompaniedby
the
growth
of economic linkages between urban
and
rural
areas.
As
a result, there
emerged
n
the
Balkans and
in
Anatolia an intensive
pattern
of
periodic
marketsand market
airs, where peasantsand larger andholders
old part
of theirproduce o urbanresidents.
These marketsalso
provided
an impor-
tantopportunity
or nomads o comeinto contact
with
peasants
and towns-
people. Largesectorsof the ruralpopulation ameto use coinage,especially
thesmall denominations f silver
ak( e
and he copper
mangir,through heir
participation
n
these markets.25
Secondly,
mall-scalebut ntensivenetworks f credit elations
developed
in
and around
the urban centers
of
the Balkans,
Anatolia, and probably
elsewhereduring
he same period.Evidencefrom thousands
of courtcases
in
these towns and cities involvinglenders and borrowers
eave no doubt
thatthe
use
of
credit,
small
and
large,
was
widespreadamong
all
segments
of urbansociety andpartsof thecountryside. t is clearthatneither he Is-
lamic
prohibitions
on
usury
nor the absenceof formalbanking
nstitutions
prevented
he
expansion of credit
n
Ottoman ociety during
he sixteenth
andseventeenth enturies.
Thirdly,
he provincial aw codes,
most of which
were ssued
between
he
mid-fifteenth nd
he mid-sixteenth enturies, how
very clearly
that
for every province
the
Ottoman
tate issued a long list of
activities that were
subject
o
taxation,
and the rate to be assessed on each
of
them. These
codes
point
to
an
economy
with
strong
urbanandrural
ink-
ages, considerablemarketorientation,and frequentcollection of small
amounts
of
cash taxes
from
artisans,
merchants, omads,
and
peasants.
The
availability
nd
use of
coinage
declined
during
he
secondhalf of
the seven-
teenth
century,
but increased
again
during he eighteenth.26
As
a
result,
debasement ad
animpacton virtaally
all
groups
n
Ottoman
society;
and each
group,
n
turn,adapted
o
it.
Court
ecordsmake
clear
that
mostmen and
women,
both
urban
ndrural,
were
clear
about he
distributional
consequences
f
different
ways
of
dealing
with
the
coinage.
In
general,
all
persons acing uture bligations xpressedn theunitof account-most impor-
tantly,
borrowers
ndtenants
paying
ixed rents
n
cash-stood
to
gain
from
debasement.
Conversely,
hose
who
expected
o
be paid
fixed
sums-notably
bureaucrats,anissaries,
nd
wageearners-stood
to
lose
from
debasement.27
25
Faroqhi, "Early History,"
"Sixteenth Century Periodic Markets," and "Rural Society."
26
Pamuk,
Monetary History,
pp. 74-82.
27
There existed a large overlap between the guild
members who were wage eamers and the
janissaries, after the latter
began to moonlight as artisans and shopkeepers
in
the seventeenth
century
(ibid., pp. 55-58, 193-200).
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 18/30
Ottoman
Living
Standards 309
We examined he short-and
long-term
changes
n real
wages through
a
time-series
analysis
of
our
price
and
wage
indices. For
the
purposes
of this
analysis themoretechnicalaspectsof which arepresentednAppendix2),
we utilized a
compositewage defined as
C
1S+ 2U
(1)
3
whereS
is the
daily wage
of a skilledconstruction
worker
and
Uis
the
daily
wage
of
an unskilled construction
worker.
Our
results indicate that
in re-
sponse
to
debasements,
both
prices
andnominal
wages adjusted
with
a
lag.
However,becausetheadjustmentf nominalwageswas slower than hatof
prices, debasements
ed to declines
in
realwages
in
the short and even
the
medium erm.
These results
help explain
he
unpopularity
f-and
the
often
strongpolitical oppositionto-debasements
among
the urban
population,
especially
n the
capital
city.
For
the
longer term,
it would
be
best to
investigate
he determinants f
realwageswithin the framework
f supply
anddemand or laborand a real
wage equation.One
would
expect
that
changes
n the
size
of the
laborforce
would be
an
important
determinant
f the
long-termchanges
in
the real
wage rate, especially
in
the
early
modem
period.28 pecifically,
one
would
expect
that
strong population
growthduring
he sixteenth
century
would
tend to
depress
real
wages,
whereas the
depopulation
of the seventeenth
century
would have
the
opposite
effect.Yetnone of the
dummy
variables
we
employed
o reflect these
demographic
rendswere
statistically
ignificant
when inserted nto
the
real-wageequation.
t
is of course
possible
that our
knowledge
of the
prevailingpopulation
rends
in
different
subperiods
s
faulty.29
It is
likelier,
however,
that
the
population
variables
were
over-
whelmed
by
the
powerfulprice
and inflationvariable.30
t
is also
possible
that
our
regressionmodel
is
misspecified
n
some
way,
or
that
changes
in
population
did not
have
any impact
on
Ottoman
eal
wages;
but
we do
not
attach
high degrees
of
probability
o
these
explanations. Further
details
of
the
real-wageequationare
presented
n
Appendix2).
The
only
other
variablethat is
statistically ignificant
n
our
real-wage
equation s the time trend(t). Itscoefficient ndicatesa long-termupward
28
We are inspired here by
a
similar study
of the
long-term trends
in
English
real
wages
undertaken
in Lindert, "English Population."
29
Annual population series are not available
for the Ottoman Empire, or for any part of it, for the
period before 1850. Instead, we have some estimates of questionable quality for different points
in
time.
On
the basis of the
latter,
we are
certain
that the sixteenth century was a period of relatively high
population growth. In contrast, 1580-1700
was a period of stagnation or, more probably, decline
in
population. Trends for 1700 to 1840 are also not entirely clear. They may have varied from
region to
region; a modest upward trend is our best guess at the moment.
30
In his study of long-term trends
in
English wages, Lindert ("English Population") similarly
found
that short-termprice movements overwhelmed the population variable in the real-wage equation.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 19/30
310 Ozmucurand Pamuk
trendof 0.3
percentper annum,suggesting
a
modest but
steady
rise in the
demand or labor.We tried o separate he time trend nto two
subperiods
at
1770,hypothesizing hatgrowth nlabordemandmayhave accelerated fter
the Industrial
Revolution. However,
t
in
two
subperiods
did not
perform
differently han a single t for the entireperiod.It is not
easy to interpret
he
performanceof t, because even thoughthere is
considerable
evidence for
increases n per capita income after 1880, the
existing literaturedoes not
suggest
a
long-termgrowth rendfortheOttoman
economy for the period
before1850. Nonetheless,one nowhas to consider he
possibility
thatsome
modest trendfor productivity ncreasesdid
exist in the
easternMediterra-
nean before the IndustrialRevolution.Learningby doing, or the diffusion
of new technologyfromWesternEurope,may
help explain
this
trend.3
COMPARISONSACROSS EUROPE
In
a recentstudyof prices andwages
in
European ities fromthe Middle
Ages
to
WorldWarI,
RobertAllen
has
utilizeda
largebody
of datamost of
which was
compiled
under
he
aegis
of the International
cientificCommit-
tee on Price History.32n order o facilitatecomparisons,he convertedall
price and wage series into gramsof silver, andchose
as
a
base
the index of
average
consumer
prices prevailing
in
Strasbourgduring
the
period
1700-1749.33
Allen has
argued
hateven
thoughwages
in a
single city may
be
accepted
as
a
barometer f
wages
in
the whole economy, nternational
omparisons
must be made between
cities
at similar
evels
in
the urban
hierarchy.Since
his
study
uses data from cities at the
top
of their
respective
urban
hierar-
chies-such asLondon,Antwerp,Amsterdam,Milan,Vienna,Leipzig,and
Warsaw-it
would make sense
to insert
Istanbul,
another
city
at the
top
of
the urban
hierarchy
f its
region,
nto this framework.
t
is
not
very
difficult
to do
so,
since we have
alreadyexpressedprices
and
wages
in
grams
of
silver.
It remains
now to
express
Istanbul
prices
in
terms of the Allen base
(i.e., Strasbourg, 700-1749). Forthispurpose,
Ottoman ommodityprices
for
the interval
1700-1749
were
applied
o Allen's
consumerbasket
with
fixed
weights.
A
second and
equally
useful
method
of
linking
Istanbul's
pricelevel to those of otherEuropean ities in the Allen set was to employ
the detailedannual
commodity-price
eries
gatheredby
Earl Hamiltonfor
Valencia
and
Madridover the
period
1500-1800 and
compare
hemwith the
Istanbul
prices for the same commodities.34 ince Valenciaand
Madrid
prices
were
already
calibrated nto the Allen
set,
it was then
possible
to
31
Persson,
Pre-Industrial Economic Growth.
32
Cole
and
Crandall,
"International
Scientific Committee."
3 Allen, "Great
Divergence."
34
Hamilton, American Treasure and Warand Prices.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 20/30
Ottoman
Living Standards
311
determine
he Istanbul
rice
evel vis-a-vis European
ities for each
nterval.
The price
series
for
flour,
mutton,
olive oil,
cooking oil, onions,
chickpeas,
pepper,sugar,and wood were used in these calculations.The two proce-
dures
produced
very
similarresults.
One
revision
n theAllen seriesconcerns rice
evels during
he 1850-
1899
and
1900-1913 intervals.
n
order o
remain onsistent
hroughout,
llen con-
tinued
o
express
all
prices
and
wages
n
grams
of
silvereven after
1870,
when
silver
prices
declined harply
ndmostEuropean
ountries
bandoned imetal-
licstandards
n
favor
of
gold.
While
echnicallyorrect,
hisprocedure
ives
the
impression
hatmostEuropean
itiesexperiencedigh
rates
of inflationduring
thisperiod,whenin fact nominalpricesandwagesremained uitestable.We
chose
to
express
all
price
andwage
levels
in
gold
after1870
by
deflating
he
Allen series
by
1.3 and
2.1 for 1850-1899
and
1900-1913,respectively.
One
of
Allen's
major
findings
s thatreal
(i.e.,
purchasing-power-parity
[PPP]adjusted)
wages
in
these European
ities divergedstrongly
between
1500 and 1914.
He shows that hedispersion
f
real
wages
was lowest
in
the
earlypart
of
the sixteenthcentury.
This
picture
changed
after 1600
as real
wages
in
northwesternEurope (London,
Antwerp,
and
Amsterdam)
re-
mained teady,whileItalianandSpanishas well as centralandeasternEuro-
pean
real
wages
declined until 1850.
Allen's study
thus suggests
that the
emergence
of a North-Southgap
within
Europe
was not the product
of
developments
afterthe Industrial
Revolution,
but resulted
from
long-term
trends
during
he seventeenth
and
eighteenth
enturies.
Our
ndices show
that
daily
wages
in
Istanbul
andothereastern
Mediter-
ranean ities expressed
n
grams
of silverwere
comparable
o those
in
many
other locations
in northernand southern
Europe
in the earlypart
of the
sixteenthcentury.However,becauseIstanbul riceswerehigherthan hose
of all other
cities
in
Allen's sample,
realwages
in Istanbul
variedbetween
60 and 90
percent
of real
wages
in othercities
during
hat
period
(Figures
2
and
3).
It
is
interesting
hat while
real
wages
continued
o decline
after
1600
in
manyparts
of
Europe, hey
remainedittle
changed
n
Istanbul
until
late
in
the eighteenth
entury.
While
Istanbulwages
approached
hose
else-
where
n
Europe,
a
wage gap
of one-third o
one-half
between
Istanbul
and
the
leading
cities
in
northwesternEurope
continued until
the Industrial
Revolution.
Wage
and
price
indices
for the
western
Netherlands
ecently
constructed
by
Jan
Luitenvan
Zanden
present
another
pportunity
o
investigate
he
real
wage gap
between the eastern
Mediterranean
nd
northwesternEurope
before
he Industrial
Revolution.3s
Measured
n
grams
of
silver,
the nominal
"
The van Zanden indices are available fromwww.iisg.nllhwp/;
for
prices
and
real
wages
in Nether-
lands during the early
modem era, see also de Vries
and van der
Woude, First Modern Economy,
pp. 607-54.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 21/30
312
Ozmucurand Pamuk
-
Istanbul
14
*
Antwerp+Amsterdam
-
*-London
12
M*---Paris
--i-
-Valencia
-Leipzig
10 -- -
-Vienna
-
--Warsaw/
/
8-
4~~~~~~~~~~~~~~~4
2
0
I
I .
I
.
II
.
I .
I
1450- 1500-
1550- 1600- 1650- 1700- 1750- 1800- 1850- 1900-
1499 1549 1599 1649 1699 1749 1799
1849 1899
1913
FIGURE
REAL WAGE OF UNSKILLED
CONSTRUCTION WORKERS
IN
EUROPEAN
CITIES,
1450-1913
(wages [in grams
silver]
divided
by
CPI
[in
grams
silver])
Sources:
Istanbul:
see the text. Other
cities:
Allen,
"Great
Divergence."
wages
of skilled and
unskilledconstruction
workers
were
quite
comparable
in
the two
regions
during
the first half of the sixteenth
century.
However,
ourdirectcomparison f consumerprice evels (basedon thepricesof more
thanhalfa dozen commoditieswhich areavailable orboth
regions)
ndicate
that
n
silvertermsIstanbul
rices
were about60 to
80
percent
higher
during
the same
period.
During
the next two
centuries,
or until
about
1750,prices
andnominal
wages
in
thewestern
Netherlands
measured
n
grams
of
silver
rose
by
about 250
percent,
with the latter
lagging
somewhat behind the
former.
In
Istanbul,
on the other
hand,prices
in
grams
of silver
remained
roughly
unchanged
while
real
wages
declined
slightly.
Our
preliminary
comparisonof the two seriesthussuggeststhat realwages in Istanbulre-
mained
one-third o one-half
below real
wages
in
the
westernNetherlands
until the era
of the Industrial
Revolution.
Finally,
comparing
our Istanbul
data o Hamilton's
wage
and
price
seriesfor Madridand
Seville revealsreal
wage
trends
over the
periods
1500-1550 and
1700-1750
that
are
also simi-
lar
to those outlined
n
Figures
2 and
3.36
36
Commodity-price
and
nominal wage series for
Spain
are
available from Hamilton, American
Treasure and Warand Prices, appendices.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 22/30
OttomanLiving
Standards
313
25
Istanbul
Antwerp+Amsterdam
A-
London
20
-
Paris
- -a-
Valencia
-
--Leipzig
-
--Vienna/
15-
15 - 4-
-Warsaw/
10
5-
0
I
I
.
I
,
I
.
I
. -
.
I
. I
1450-
1500- 1550- 1600-
1650- 1700- 1750-
1800- 1850-
1900-
1499 1549
1599
1649 1699
1749
1799 1849 1899
1913
FIGURE
REAL WAGE OF
SKILLED
CONSTRUCTION
WORKERS
IN
EUROPEAN
CITIES,
1450-1913
(wages [in grams
silver]
divided
by CPI)
Sources:
Istanbul: see
the text. Other cities:
Allen,
"Great
Divergence."
WAGES
AS AN
INDICATOR
OF THE
STANDARD OF LIVING
Istanbul ealwages increasedby about wo-thirds romthe last quarter f
the
eighteenth
centuryuntil World
War
I
(Figure1).
As
a
result,
while
the
gap
in
real
wages
between
Istanbul
ndcities in northwesternEurope Lon-
don,Antwerp,
Amsterdam,
and
Paris)
widened
after he
IndustrialRevolu-
tion,
that
gap
was lesser
than
one
might
haveexpected.On
the eve
of
World
War
,
real
wages
of
unskilled
workers
n Londonwere 2.7
times
as
high
as
in
Istanbul;
he
corresponding
atio for Amsterdam
and
Antwerp
was 1.9,
and
for Paris 1.6
(Figure
2).
While theseresultsarebroadly n linewithourexpectations,a compari-
son with the
PPP-adjusted er
capita
GDP
series
recently
constructed
by
Maddison
evealsdifferences hat
arenot
insignificant.
The Maddison eries
show
that
per
capita
GDP
differencesbetween Turkey
and each
of four
westernEuropean
ountries
the
United
Kingdom,
France,
he
Netherlands,
and
Italy)
were wider
than
the
differentials
n the real
wages
of
unskilled
construction
workers
in the
leading
cities
of
each
pair
of
countries.37
37
Maddison, WorldEconomy.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 23/30
314 Ozmucurand Pamuk
Maddison's PPP-adjusted per capita GDP estimates for 1913 point to a 1:5
gap between Turkey and the United Kingdom, a 1:3.8 gap between Turkey
and the Netherlands, a 1:3 gap between Turkeyand France, and a 1:2.3 gap
between
Turkey
and
Italy.
In
short,
the
gap
in
PPP-adjusted per capita GDP
between Turkey and these European countries
on the eve of World
War
I
appears, on the average, to be twice the real wage gap between Istanbul and
the leading cities in these countries. (Indeed, Istanbul real wages were
slightly higher than those of Florence and Milan.) This is an interesting and
potentially important divergence
that needs
to
be examined further.
Part of this
divergence may
be due to errors of measurement
in
the
avail-
able series. For example, there may be errors in our real wage series for
Istanbul.
We are aware that most
of
our
knowledge
about the
wages
of
construction workers has been based upon state records and records of the
middling
and
larger pious
foundations. These
larger
institutions
could
seek
out the best craftsmen
and pay higher wages
than smaller
employers.
Clearly,
this
bias exists
in
most
wage
series
but
it
might
be
greater
the
in the
case
of
Istanbul
than for
other European cities. Otherwise, we
believe
that
our
wage series are quite reliable, especially since they are based on large
numbers of observations. Infact, the London-Istanbul real wage gap implied
by
our
series
is
actually larger than the wage gap estimated by earlier studies
by Korkut Boratav, GtunduizOk9iin, and Pamuk
on
Turkish urban wages,
and
by
Berov on
Balkan wages. These studies pointed to a gap in nominal
wages
converted
at the prevailing rate of exchange of about 1:2.5.38
Secondly,
there
may
be
errors
in
the
way
our Istanbul
prices
have
been
inserted into the
Allen
framework. We doubt, however, that these can be
significant,
since our
price
series was linked to the Allen
base by comparing
a numberof independent price series. In addition, since our real wage series
for Istanbul
show
a 20
to 30 percent overall increase over the
four centuries
to
1914, raising
the nominal
price
level
and
lowering
real
wages
for
1900-1913 would not only bring Istanbul prices uncomfortably close to
those of London at the
endpoint,
but it would also
raise
Istanbul
prices
across time
and
thus widen the
Istanbul-London
real
wage gap
to
more than
1:2 for
the
early part
of the sixteenth
century.
Thirdly,
there
may
be room for
improvement
in the
way
the
Maddison
series measure the differences in per capita GDP between high- and low-
income
countries.
In a
recent
study,
Leandro Prados de
la
Escosura has
pro-
vided an
alternative set of
GDP series for
the
United States
and
Europe since
1820. While
the
Maddison series were derived
by working
backwards
from
a recent benchmark
year
on
the
basis of
growth rates,
Prados de
la
Escosura
used
a short-cut method to
obtain
current
price estimates,
also
adjusted
for
PPP. His
estimates indicate
that
the use of current
prices
reduces
considerably
38
Boratav, Ok9iun, and Pamuk, "Ottoman Wages"; and Berov, "Wages in the Balkans."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 24/30
Ottoman
Living
Standards
315
the gap
in
per
capita
ncome
between the
high-
and low-income
countries:
specifically,
he gap
betweenTurkey
nd heUnited
Kingdom
n 1913would
be reduced rom1:5 (as estimated y Maddison)o approximately:3.
Despite
these
potential
ourcesof error,we
still
doubt hat
mprovements
in the available
seriescan
eliminate
altogether
he
divergence
between
the
urban
eal
wage
and
percapita
GDP
series. We
think
thereare
a
number
of
other
actors
which
may
have also contributed
o this
divergence.
First,
here
had emerged
by
the nineteenth
century mportant
differences
between
the
nominal
wagesof
construction
workers
n Istanbul
ndthose
of other
Otto-
man cities.
Between 1850
and 1914
nominal
wages
were,on
average,
40 to
50percenthigher n Istanbul han n othercities withinthe bordersof mod-
em Turkey.
n contrast,
officialdata
ndicate
hat in 1913/14
the
pricesof
essential
commodities
purchased
by
consumers
appeared
o be
lower in
Istanbul
han
the average
of the 20
leading
cities
within
Turkey.40
Secondly,
here s evidence
fora
growing
scarcity
of labor n urban
areas
from the mid-nineteenth
entury
untilWorld
War
I.
Anatolia
was
a
labor-
and capital-scarce,
and-abundant
egion
until
the
middle of
the twentieth
century;
hese factorproportions
upported
easant
ownership
andproduc-
tion.The availabilityof landand heprevalenceof smallpeasantproduction
may
have slowed
rural-urbanmigration
and
contributed o
laborshortages
in the
urban
areas.
In
a recent
study
of
nineteenth-century
natolia,
Bogag
Ergene
has
providedevidence
for
the growing
scarcity
of labor
n the urban
areasand
the
increasing
gap
between
urban
andruralwages.
He alsoargues
that ncreases
n urban
real
wages
must
have exceeded
increases
n
per
ca-
pita
GDP
during
he nineteenth entury.41
These urban-rural
differences
combined
with the
regional
differences
mentionedabove to createlargedifferencesbetween per capitanominal
income
levels
in the Istanbul
region
and the rest of
the
country.
Vedat
Eldem's estimates
ndicatethat on
the
eve of
World
War
I, per
capita
in-
come levels
in the Istanbul egion,
which
was mostly
urban,
were
twice
as
high
as the
average
or the
Ottoman
Empire
as a whole.
All
this
suggests
hat he
daily
wages
of urban
onstruction
workers
n the
Ottoman
Empire
may
have
been
high
in
relation
o the
underlyingper
capita
income.
In
a recent
study,Jaime
Reis encountered
imilar
difficulties
n
his
attempt o useurbanwagesasaproxyfor percapitaGDP.Whilecomparing
two
groups
of countries
n the
European
eriphery
during
he firsthalf of
the
nineteenth
entury,
he
observed
hat
daily
wages
of construction
workers
n
the
capital
cities of
Spain
and
Italy
were
at
the
same
level as
comparable
wages
in
Scandinavia;
but all other
indicatorssuggested
that
per
capita
39
Prados de la Escosura,
"InternationalComparisons."
40
Turkey, State
Institute of Statistics,
Ihsai
Yilblk.
41
Ergene, "Wages."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 25/30
316
Ozmucur
and
Pamuk
income
levels
were
considerablyhigher
n
the northern ountries.42
t
the
otherend
of thespectrum,he real wages
of construction
workers n Milan
andFlorenceduring henineteenth enturyappearunusuallyow in relation
to
the availableseries
for
per capitaGDP
in
Italy
during
he same
period.43
CONCLUSIONS
One
of the more
important
uestionsregarding
he world
economy
in
the
earlymodemera concerns
he emergence
and evolutionof
the
gap
in levels
of real ncome
between
today's
developed
and
developing
countries.Utiliz-
ingalargevolumeof archivaldocuments, hisstudyhas established or the
first
ime
the
long-term
rends
n
wagesof
skilled
and
unskilled
construction
workers n IstanbulandotherOttoman
ities
in
southeastern
Europeand
the
Near
East from the second
half
of
the fifteenth
century
until
WorldWarI.
These
price
and
wage
series were then
inserted nto
a
larger
ramework
of
price
andwage
trends n
European
ities
during
he same
period.One
needs
to be
cautiousabout
using
the
daily
wagesof urban
workersas indicators
of
the
standards f
living
for an
entirecountry.
Nonetheless,
n the absence
of
reliableinformationaboutproductionand income, real wage series still
serve as
the best indicator
available
for long-termtrends
in standardsof
living.
The
Ottoman
ndices indicate hat
he
realwages
of unskilled
urbancon-
structionworkersdeclinedby 30 to
40 percent
during he sixteenthcentury.
After
remaining oughlyunchanged
until
the
middle of
the
eighteenth
cen-
tury, hey
increased
by
about20 to
30 percent
rom
the
late
eighteenth
until
mid-nineteenth
enturies,
and
then
by
another40
percent
during
the late
nineteenth ndearly wentieth enturies.On the eve of WorldWar , the real
wages
of unskilledconstruction
workerswere about
10 to 20 percentabove
their evels
in
1500.
Because
the
skill
premium
rose
during
he
nineteenth
century,
real
wages
of skilled workers
n
1914
were 50
percent
or
more
above their
evels
in
1500.
Comparing
hese resultswith
Allen's
findings
for other
European
ities
reveals three
basic results.First,real
wages in Istanbul
and otherOttoman
cities remained lose to
urban
wages
in
many
other
parts
of
Europe
untilthe
mid-eighteenth entury.Real wages in Ottomancities were lower in 1750
than
n
1500,
but the decreasewas
probably
not as
large
as
in
many
other
parts
of
Europe.
Secondly,
our
comparisonsuggest
hat
real
wages
in
north-
western
Europe(Amsterdam,
Antwerp,
andLondon)were
alreadyhigher
than n
Istanbulat the beginningof the
sixteenthcentury,and
that this gap
remained
oughly
unchangedup to the Industrial
Revolution.
It would be
42
Reis, "How Poor?"
4
Allen, "Great Divergence."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 26/30
OttomanLiving Standards
317
useful to
check this result
with additional
esearch.
Thirdly,
he trendwas
upwards
or realurban
wages
in most
regions
of Europe n the
wake of
the
IndustrialRevolution,especiallyafter1850;real wages in western Europe
increased
more than
in the
eastern
Mediterranean
uring
his
period.
Our
real
wage
series also
point
to the existence of a modestbut
statisti-
cally
significantupward
rend n
Ottoman
urbanwages dating
back to the
sixteenthcentury,
uggesting
a
modest
but
steady
rise in demand or
labor.
It is not
easy to interpret
his result,because
theexisting literature
doesnot
point
to
the
existence
of
long-term
growthtrend,
especially
for theperiod
before 1850.
In
light
of the
new
evidence,
we
now need
to
consider
the
possibility of a slow and modest rise in productivityaroundthe eastern
Mediterraneann
the erabefore he
IndustrialRevolution.
Learning ydoing
or the diffusion
of
new technology
fromWesternEurope
may help explain
this
trend.
Our
overall results thus amountto a
guardedlyoptimistic
revisionism
regarding
Ottoman
tandards f living,
both
in
theearly
modem era
and
in
the wake
of the Industrial
Revolution.This
picture
s
consistent
with
the
recentarguments
n Ottoman
historiographyo the
effectthat the Ottoman
empiredidnot simplyandirreversibly eclineafter 1600. Recentliterature
has
emphasized
hat
Ottoman
tateand
society
showed
considerable
ability
to
reorganize
as
a
way
of
adapting
o changingcircumstances
rom
the
sev-
enteenthhrough
he nineteenth enturies.With
pragmatism
nd
flexibility,
the
Ottoman
tate
adapted
not
only
the
military echnology
but
also its fiscal,
financial,
andmonetary nstitutions
uring hisperiod.
Thispragmatism
nd
flexibilitymay
help explain
the surprisingongevity
of the
so-called "sick
man of
Europe."
Appendix
1: Long-Term
Trends n Indices
Based on trends
n
money,prices,population,
andrealwages,
it is
best
to examinethese
four-and-a-half
enturies
nfive
distinct
subperiods.Separate
egressions
are used to
esti-
mate trend
ines
for
each one."
The first
period,
1469-1585,
is characterized y stable
money, rising population,
and
rising prices
expressed
n
grams
of
silver.
Thetrend ines indicate
that the silver content
of theakVedeclinedatanannualrate of 0.52 percent,pricesrose atan annualrate of 1.13
percent,
and real
wages
declined
at
an
annualrate of 0.64
percentduring
his
period.
During
the second
period,
1585-1690,
there
was a
high
degree
of
monetary
and
price
instability ogether
with
decliningpopulation.
The silver content
of
the currency
luctuated
wildly
around
a
trend
of
1.05 percent
per annum,prices
rose
at an annual rate of 0.67
percent,
and real
wages
rose at annualrate of 0.29 percent.
The third
period,
1690-1768,
was once
again
a
period
of
monetary
and
price
stability.
The silver contentof the
currency
declinedat
an
annual
ateof
0.30 percent,prices
rose at
an annual
rate
of 0.44
percent,
and real
wages
declined
at
an
annual
rate of
0.15
percent.
4 Poirier, Econometrics of Structural Change; and
Greene, Econometric
Analysis.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 27/30
318
Ozmucurand Pamuk
The fourthperiod, 1769-1843,
is characterized y
the most rapidrates of debasement
and inflation n Ottomanhistory,and
by rising real wages. The silver content
of the
cur-
rency declined
at an annualrate of
3.49 percent,pricesrose at an annual
rate of 3.81 per-
cent, and real wages rose at an annualrate of 0.49 percent.Frequentwarsmay have con-
tributed o the rising trend
n
real
wages.
The fifthperiod,1844-1914, witnessed
stablemoneyunderbimetallism,
apidly xpand-
ing international
rade,stable prices,
and rising real wages. The silver content
of the cur-
rency remained
unchanged,pricesincreasedat an annual
rate of 0.67 percent, and real
wages
rose at 0.21 percentper
annum.Most of the
price
increases
occurredearly
in
the
period and
were the lagged result
of debasements n the previousperiod.
Appendix2: Determinantsof Real Wages
This
appendix
presents
he detailsof our ime-seriesanalysisregarding
he determinants
of
realwages.
In the shortandmedium erms,
he most important eterminants
f Ottoman
real wages were changes in prices
and the manner n
which nominal wages adjusted o
them.
Because nominalwages were slow to adjust,price
shocks led to fluctuations
n
real
wages. The
most importantdeterminant f the price evel,
in turn,was debasements.
Otto-
man
prices
almost always rose
in the aftermath f debasementsbecause
a debasement
typically increased
he nominal value of coinage in circulation. n order
to study the lag
structure
with which
prices
and
wages adjusted
o debasements,
we have
developed
a
distributedag
model
that
relates
nominalwages (W)andthe consumerprice
index (P) to
present and laggedvalues
of
silver
contentof coinage (S). All variables
ncluded in the
equation
are
expressed
n
logarithms.
Using the general-to-specific
methodology,45qua-
tions
are
estimatedwith lags up to 10 years; only lags
with significant coefficients are
included
n the final
equation.
This approach as the advantage f not placing
any restric-
tions
on
the
lag structure,
ut
it
has
the disadvantage
f
possible multicollinearity
mong
explanatory
ariables.The
model
is of
the following
form46
ty
=
1AS,_
+
ut,
i
=
1,2,...k
(Al)
where
k
is the
number
of
lags.
The mean lag indicates the average
numberof years
needed to see the effect of
the
independent
ariableon the dependent ariable,
while
the
median ag gives
the numberof
years during
which half of
the effect has occurred.
The mean
lag
can be calculatedas
k
ifi
i=O
Ai
and
the
median
ag
can
be calculatedwith
the
formula
i
i=O
Ai
The median
lag
is
the first k* value
that
gives
the sum
0.5.
45
Hendry, Dynamic
Econometrics.
46
Greene, Econometric Analysis.
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 28/30
OttomanLiving
Standards
319
Regressionresults ndicatethat the
impactof a change n
the silver contentof the cur-
rency was quicker
on prices than on wages. When prices are
used as the dependentvari-
able, the lags
that are significantlydifferent rom zero are:
zero, one, four, six,
and eight.
The total effect is -1.088 (sum of coefficients).At equilibrium, 1-percentdecrease n the
silver
content of coinage causes
a
1.088-percentncrease
in
the consumerprice index.
Furthermore,
lmost half of the total
effect is felt in that year and the year
following.
Ceterisparibus,
a
1-percentdecrease
n the silver contentof
coinage
thus causes a
0.236-
percent
ncrease n the price level
in
that
year,and a 0.261-percent ncrease
the following
year.
The mean lag is 3.59 years, and the median ag
is
4
years.
P,= 0.113
-
0.236 S,- 0.261
St,
-
0.207
St-4-
0.129
St6-
0.255
St8
(0.86)
(-3.60) (-4.03) (-3.33)
(-1.87) (-3.92)
R2=
0.993, SEE
=
0.1327, F= 6814.8, Durbin-Watson 2.30, AR(1)
=
0.874
The impact on wages
of a
change
in the silver content
of the currencyoccurredmore
slowly.
When
nominal
wages
are used
as
the
dependentvariable, zero-, one-,
two-,
and
three-
period ags
arenot significantly
different romzero.However,coefficients
associated
withfour-,six-,
seven-,
and
eight-periodags
are
significantly
different rom
zero. The total
effect
is
-1.323 (sum
of
coefficients).
At
equilibrium,
a
1-percentdecrease
n
the silver
content
of
coinage
causes
a
1.323-percent
ncrease
n
wages.
A
1
percentdecrease
n
silver
content
of coinage causes a 0.232 percent ncrease n wages
after four years,
and 0.402
percent
aftersix
years.
The mean
lag
is
6.43,
and
median ag
is
seven
years.
Wt
=
0.973
-
0.232
St4
-
0.402
St6
-
0.339
St7
-
0.350
St8
(4.19) (-2.43) (-2.99) (-2.57)
(-2.47)
R2=
0.995, SEE
=
0.1378, F=
3630.3, Durbin-Watson 2.73, AR(1)
=
0.871
Our
distributed-lag
model thus
indicates
hat n
response
o
debasements,
both
prices
and
nominal
wages adjusted.However,because
he
adjustment
fnominalwages
was slower han
thatof
prices,
debasementsed to
declines
n
real
wages
in the shortand
even
medium erm.
This nteraction etweenprices
andrealwagescanbe examined urtherwithin
hebroader
framework f a real-wageequation.We have employedtwo models, one with adaptive
expectations
n
prices,
andanother
ne withpartial djustment
n real
wages.47
n
bothmodels
real
wages
are
expressed
as a
functionof inflation
and
real wages
in
the
previous
year.By
appropriateubstitution,
t is not difficult
o
see
that
realwages
aredetermined
y
the
present
and
past
nflationrates.
In
this
respect
ourmodel
is similar o thatused
by
PeterLindert.48
In our
regressions,
owever, nly
the nflation
f
thepresent
nd
past
four
years
appear
o
be
statistically ignificant.
The estimated
quation's egression
oefficientsand determination
coefficient
are
statistically ignificant;
he Durbin-Watsontatistic
ndicates
hat
there
s no
serialcorrelation. his
was achieved
by usinggeneralized
east
squares,
which
s
indicated
y
theAR(1)coefficient. t is noteworthyhat he time-trend ariable,usedto capturehe effect
of other ariables uchas
population
nd
echnologicalhange,
s
statisticallyignificant.
Other
thingsbeing
constant,
eal
wages
ncreased
t 0.34
percent
eryear
n this
equation:
RW,
=
-4.496
-
0.900
P,
-
0.773
Pt-,
-
0.649 P-2
-
0.513
Pt3
-
0.451
Pt-4
+
0.0034 t
(-2.51) (-6.45) (-4.42)
(-3.20)
(-2.41) (-2.65) (8.23)
R2=
0.808,
SEE
=
0.140, F=
40.95, Durbin-Watson
2.31, AR(1)
=
0.777
4 Greene, Econometric Analysis.
48Lindert,
"English Population."
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 29/30
8/10/2019 Real Wages and Standards of Living in the Ottoman Empire
http://slidepdf.com/reader/full/real-wages-and-standards-of-living-in-the-ottoman-empire 30/30
Ottoman
Living Standards
321
Europe
since 1500." This JOURNAL
62, no. 2 (2002): 322-55.
Inalcik, Halil, and Donald Quataert, eds. An Economic and Social History of the Ottoman
Empire. Cambridge: Cambridge University Press, 1994.
Lindert, Peter. "English Population, Wages and Prices, 1541-1913." Journal
ofInterdisci-
plinary History 15 (Winter 1985): 609-34.
Maddison, Angus.
"A
Comparison
of
Levels of GDP Per Capita
in
Developed and Devel-
oping Countries, 1700-1980." This JOURNAL43, no. 1 (1983): 27-41.
. Monitoring the WorldEconomy, 1820-1992. Paris: OECD Development Centre
Studies,
1995.
. The WorldEconomy,
A
Millennial Perspective. Paris: OECD Development Centre
Studies,
2001.
Mokyr,
Joel. The Lever
of
Riches:
Technological Creativity
and
Economic Progress.
Oxford: Oxford University Press, 1990.
Pamuk, Aevket.
A
Monetary History of the Ottoman Empire. Cambridge: Cambridge Uni-
versity Press, 2000.
. Istanbul ve Diger Kentlerde
500
Yillik Fiyatlar ve Ucretler, 1469-1914 [500
Years of Prices
and
Wages
in
Istanbul
and
Other Cities].
Ankara:
State Institute
of
Statistics of
Turkey,
2000.
.
"The Price Revolution in the
Ottoman Empire Reconsidered." International
Journal
of
Middle East Studies
33, no.
1
(2001):
69-89.
Persson, K. G. Pre-Industrial Economic Growth, Social Organization and Technological
Progress in Europe. Oxford: Basil Blackwell, 1988.
Phelps Brown, E. H., and Sheila V. Hopkins. "Seven Centuries of the Prices of Con-
sumables
Compared
with Builders'
Wage
Rates." Economica
23,
nos. 89-92
(1956):
296-314.
.
A
Perspective of Wages
and
Prices. London: Methuen, 1981.
Poirier,
Dale
J.
TheEconometrics ofStructural Change. Amsterdam: North-Holland,
1974.
Prados de la
Escosura, Leandro. "International Comparisons of Real Product, 1820-1990:
An
Alternative Data Set." Explorations in Economic History 37, no. 1 (2000): 1-41.
Raymond,
Andre.
Artisans et
commerVants
au Caire au XVIIIe siecle.
2
vols. Damas-
cus: Institut
Francais
de
Damas,
1973-74.
Reis,
Jaime. "How Poor Was the
European Periphery before 1850?
The
Mediterranean
vs. Scandinavia."
In
The Mediterranean Response to Globalization before 1850,
edited
by $.
Pamuk and
J.
Williamson,
17-44. London:
Routledge,
2000.
Sahillioglu,
Halil.
"Yeniceri
?uhasi ve II.
Bayezid'in
Son Yillannda
Yeniceri
?uha
Muhasebesi."
Guiney-DoguAvrupa Aratrmalar Dergisi 2-3 (1973-74): 415-65.
Turkey.
State Institute of Statistics. Merkezi Istatistik
MuiduiriyetiUmumiyesi.
Ihsai
Yillhk.Ankara:
n.p., 1928.
Van
Zanden,
Jan
Luiten.
"Wages
and
the
Standards of
Living
in
Europe,
1500-1800."
European Review of Economic History 3, no. 2 (1999): 175-98.
. "The Great Convergence from a European Perspective." Itinerario, European
Journal of Overseas History 24, no. 3/4 (2000):
9-28.
Williamson, Jeffrey.
"Real
Wages and
Relative Factor Prices Around the Mediterranean
1500-1940." In
The Mediterranean
Response
to Globalization
before
1850,
edited
by $.
Pamuk
and J.
Williamson,
45-75. London:
Routledge,
2000.