Mojgan Stegl and Joerg Baten University of Tuebingen

1
Mojgan Stegl and Joerg Baten (University of
Tuebingen)

• Tall and Shrinking Muslims, Short and Growing
  Europeans an Anthropometric History of the
  Middle East, 1850-1980

2
Height versus GDP and Real Wages

• Biological Standard of Living
• GDP per capital advantages
• Comprehensive account of purchasing power
• Comparability over time
• GDP per capita disadvantages
• Subsistence farming and production and
  consumption within households
• Not capturing of black markets
• Huge Data requirement for GDP estimation
• Real Wages as standard of living indicator
• better data qualtiy as prices and nominal wages
  were recorded by contemporaries
• but less standardzied goods are not included

3
Basic question of the study

• How did the standard of living develop in the
  Middle East over the past two centuries?
• When did the Western World become clearly
  better-off in terms of income and health?
• Did height development follow GDP, or perhaps
  rather real wages (-gtinequality?)
• Motivation interest in this world region has
  substantially grown over the recent years

4
Coverage and main sources

• rural and farmer populations
• Countries Turkey, Iraq, Iran (West), Egypt,
  Syria, Lebanon, Palestine/Israel, and Yemen
• test the hypothesis that some Middle Eastern
  populations were taller than Europeans in the
  mid-19th century, using British, German, Czech,
  and Italian samples
• primary sources
• anthropological measurements for the birth
  cohorts of the 1850s to the 1910s
• for the recent period 1920-1980 Baten (2006),
  based on Demographic and Health Surveys

5
Number of height observations by country, birth
decade, and individual versus grouped data
Notes Grouped data are grey-shaded. Source and
Abbreviations Table 2. Abbreviation EG-Egypt,
IQ-Iraq, IR-Iran, IS-PS-Israel/Palestine,
LB-Lebanon, SY-Syria, TR-Turkey, YE-Yemen
6
Representative data?

• To what degree is our data set representative?
• For our individual samples
• among recorded occupations, 71 were engaged in
  the agricultural sector
• Issawis (1966) estimated four fifths of the
  Mideastern population were in agriculture
• our sample might include a slightly higher share
  of the rural population than given in the overall
  population
• broadly representative

7
Main problem of anthropological surveys birth
decade estimates

• many anthropologists of the late 19th and early
  20th century assumed no change of height over
  time
• Hence when were most of the measured individuals
  probably born?
• time trend based on estimated birth cohorts has
  some properties of moving averages
• Turkish sample one is by far the largest - almost
  29,000 observations between the 1880s and 1910s.
• most accurate information
• drawback of this sample is the oldest cohort of
  Turkish males aged 50-59 years are included
• Underestimate of height by about one centimeter

8
Turkish males 1850-1910

• large sample 33,447 observations,
  representatively drawn
• height underestimated here (incl. very old
  young)
• but regional difference still informative.
  West-Centre-diff
• frequent change of borders, hence we use modern
  borders

9
Height by region type and country

• Note Height averages are weighted by the number
  of observations. Only cases with Ngt30 shown.
  Source and Abbreviations Table 2.

10
Estimation issues

• we estimated with WLS multiple regression
  analysis and controlled for sample composition by
  region and age
• plus robustness check with other classification
  (desert, coast, urban, other agricultural)
• we controlled for region, age, religion,
  migration
• only ages 20-50, heights 120 200 cm, cases 30
  underlying observations
• we report dummy variables for each country and
  birth cohort (Table 7, column 1)
• we checked the distribution for normality
• the desert region variable is positive and
  significant but not very large

11
Heights in the Middle East
12
Heights in Turkey, Iraq, and Iran

• Tallest populations during the 1870s
• Iraq 1910s might be an outlier (not too many
  cases)
• strong nomadic element in central Anatolia and
  Iraqi deserts (but controlling
• downward trend over the following half century
• (Western) Iran is around the ME average (or
  slightly below)

13
Heights in Palestine/Israel, Lebanon

• Lebanon one of commercial centers
• wealthy mercantile strata
• but only average heights
• demanded high quality food from the surrounding
  countryside
• invested in the silk and later fruit exports
• Palestine high anthropometric values - 1850s -
  1860s

14
Heights in Yemen

• Yemen very low anthropometric values
• high population density
• the economy specialized on spices, coffee and
  other cash crop production
• but conflicts did not allow to sustain
  irrigation systems
• around 1900 many small tribal units
• political situation instable tribal attacks
  against Ottomans (1872), and famines

15
Developments in Egypt

• subsistence to export-oriented economy
• rapid population growth
• reform policy of Muhammad Ali (1805-49) in the
  1820s to 1840s -gt GDP was rising
• modest improvement of educational situation
• but situation of the peasant population did not
  improve in the early 19th C
• heavy burden of taxation
• slight increase in 1860s (afterwards decrease)
• Crimean war boom
• early 1860s the cotton boom

16
Heights in the Middle East in general

• people tended to be fairly tall, with the
  exception of Yemen
• not much trend, especially no upward trend around
  1900 as in industrial countries (see below)
• perhaps some decrease of heights in the 1880s

17
The combined picture Middle East vs. Industrial
Countries (ICs)
Sources for 1850-1910 see Table 2 for
1940-1980 Baten (2006), based on Demographic and
Health surveys (DHS) Janghorbani et al. (2007).
The 1920s and 1930s for Middle East are
interpolated.
18
Was the difference to ICs before the 1880s
statistically significant?
Reported is always the coefficient for Middle
East in pooled samples of ME and Germany
(Bavarian prisoners) Source J. Baten (2006), ME
and Italian conscripts and criminals Sources B.
AHearn (2003) Meineke (2008), ME UK Army
Floug, Wachter, Gregory (1990)
19
Was the difference to ICs before the 1880s
statistically significant?
Reported is always the coefficient for Middle
East in pooled samples of ME and Germany
(Bavarian prisoners) Source J. Baten (2006), ME
and Italian conscripts and criminals Sources B.
AHearn (2003) Meineke (2008), ME UK Army
Floug, Wachter, Gregory (1990)
20
GDP per capita in 1990 PPP Dollars

• GDP/c development
• much worse
• situation for the
• Middle East
• already in the
• 19th C
• GDP divergence
• increase as
• well

Source S. Pamuk (2006) Weighted by population
size
21
Real wages Istanbul vs. industrialized countries

• heights and real
• wages provide
• a consistent
• picture
• relatively
• favorable living
• standards in ME
• during the mid-
• 19th century

Sources Istanbul S. Özmucur, S. Pamuk (2002),
Industrial Allen (2001), in grams of silver, but
adjusted for cost of living weighted by
population size
22
Conclusion I

• Middle East had better anthropometric values than
  the Industrial Countries before the 1880s
• afterwards, the ICs first reached equal levels,
  and then overtook
• real wages tend confirm this picture (if
  London/Amsterdam , the top 10, are removed
  Middle East fared better)
• GDP per capita was higher in Europe already in
  the 19th century

23
Conclusion II

• explanation Europe suffered from urban penalty
  and non-transportable protein
• During the 1880s-1900s period, the Industrial
  countries started to overtake the Middle East in
  terms of net nutritional status.
• Afterwards, a strong divergence was observable.
  In this study we have been able for the first
  time to identify the point in time when the
  Middle East fell back relative to industrial
  countries.