Notes

The Effects of World War II on
Economic and Health Outcomes across Europe
IZA DP No. 6296
January 2012
Iris Kesternich
Bettina Siflinger
James P. Smith
Joachim K. Winter
The Effects of World War II on
Economic and Health Outcomes
across Europe
Iris Kesternich
University of Munich
Bettina Siflinger
University of Munich
James P. Smith
RAND and IZA
Joachim K. Winter
University of Munich
Discussion Paper No. 6296
January 2012
IZA
P.O. Box 7240
53072 Bonn
Germany
Phone: +49-228-3894-0
Fax: +49-228-3894-180
E-mail: [email protected]
Any opinions expressed here are those of the author(s) and not those of IZA. Research published in
this series may include views on policy, but the institute itself takes no institutional policy positions.
The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center
and a place of communication between science, politics and business. IZA is an independent nonprofit
organization supported by Deutsche Post Foundation. The center is associated with the University of
Bonn and offers a stimulating research environment through its international network, workshops and
conferences, data service, project support, research visits and doctoral program. IZA engages in (i)
original and internationally competitive research in all fields of labor economics, (ii) development of
policy concepts, and (iii) dissemination of research results and concepts to the interested public.
IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion.
Citation of such a paper should account for its provisional character. A revised version may be
available directly from the author.
IZA Discussion Paper No. 6296
January 2012
ABSTRACT
The Effects of World War II on
Economic and Health Outcomes across Europe*
In this paper, we investigate the long-run effects of World War II on socio-economic status
(SES) and health of older individuals in Europe. Physical and psychological childhood events
are important predictors for labor market and health outcomes in adult life, but studies that
quantify these effects in large samples that cover entire diverse populations are still rare. We
will analyze data from SHARELIFE, a retrospective survey conducted as part of the Survey
on Health, Aging, and Retirement in Europe (SHARE) in 2009. This survey provides detailed
data on events in childhood including those during the war as well as several measures of
exposure to war shocks such as experience of dispossession, persecution, combat in local
areas, and hunger periods for over 20,000 individuals in 13 European countries. We find that
exposure to the war itself, and even more importantly to individual-level shocks caused by
the war such as hunger periods, significantly predict old-age outcomes at older ages.
JEL Classification: I0, H0

Keywords: health, war, SES
Corresponding author:
James P. Smith
RAND
1776 Main Street
P.O. Box 2138
Santa Monica, CA 90407
USA
E-mail: [email protected]

*
We would like to thank Dirk Jenter, Olmo Silva and Till von Wacher as well as participants at
seminars Boston University, Harvard University, and Princeton University. Sarah Lehner and Johanna
Sophie Quis provided excellent research assistance. Kesternich acknowledges financial support from
the DFG through SFB/TR 15. Smith is supported by various grants from NIA. This paper uses data
from SHARELIFE release 1, as of November 24th 2010 or SHARE release 2.5.0, as of May 24th 2011.
The SHARE data collection has been primarily funded by the European Commission through the 5th
framework and 6th framework program. SHARELIFE was supported through the 7th framework
program. Additional funding from the U.S. National Institute on Aging as well as from various national
sources is gratefully acknowledged.
1
Introduction
The Second World War was one of the major transformative events of the 20th century,
with 39 million deaths in Europe alone. Large amounts of physical capital were destroyed as
well through six years of constant ground battles and bombing. Many individuals were forced
to abandon or give up their property without compensation and to move on to new lands.
Periods of hunger become more common even in relatively prosperous Western Europe.
Families were separated for long periods of time especially from their fathers. Many including
young children would personally witness the horrors of War as battles and bombing took place
in the very areas where they lived. Horrendous crimes against humanity were committed. Due
to WW2, political and economic systems in many countries would be permanently altered.
In this paper, we investigate long-run effects of World War II on late-life economic and
health outcomes in Western continental Europe (health, education, labor market and
marriage). We identify several channels of how this war might have affected individual lives,
and document which groups of the population were most affected. Our research relies on
retrospective life data from the European Survey of Health, Aging, and Retirement in Europe
(SHARE) that have recently become available. SHARE covers representative samples of the
population aged 50 and over in 13 European countries, with a total of about 20,000
observations. We also collected external data on casualties, combats, nutrition, population
movements, and male-female population ratios.
This paper is divided into six sections. The next sets the context of the paper by
measuring the scale of WW2 using casualty statistics due to the war that include military,
civilian, and holocaust related deaths by country. Section 3 highlights the main attributes of the
2
SHARE data. SHARE not only measures the salient economic and health outcomes of adults over
age 50 in these countries, but also includes retrospective modules meant to capture salient
parts of early life including those events related to the War. After reviewing the literature on
long-term macro-effects of war, in particular WWII, the forth section summarizes statistical
models that capture the impact of the experience of WW2 on individual adult labor market,
demographic, and health outcomes. Section 5 complements this analysis by modeling the
influence of the war on some of the primary pathways through which it had long lasting
impacts – hunger, dispossession, the absence of a father, and marriage. The final section
highlights our main conclusions.
1. Causalities in WW2
Unlike many earlier wars, which consisted mainly of soldiers meeting and dying in
battlefields, civilians were also heavily affected by warfare with about half of the WW2
European casualties being civilians. Among the civilian deaths, between 9.8 and 10.4 million
civilians were murdered for political or racial reasons by the Nazi regime (Auerbach, 1992; Pohl,
2003). The distributions of deaths due to the War were very unequally distributed across
countries, whether they were military deaths due to combat, civilian deaths or the holocaust.
Figure 1 displays the fraction of the 1939 population who died in a large array of countries.
Among European countries covered by our data, Germany and Poland borne the brunt of these
casualties. In contrast American causalities in the European and Asian theatre combined were a
bit over 400 thousand, the overwhelming majority of whom were soldiers. Similarly, total
deaths in the UK are estimated to be about 450 thousands, 15 percent of whom were civilians.
3
Figure 1. WW2 Causalities as Percent of Population
While we will concentrate for data reasons primarily on the effects of World War II in
continental Western Europe, the War’s impact was just as stark in the Asian theatre. The two
countries most directly affected in terms of number of causalities were Japan and China. About
two million Japanese soldiers died in the war alongside up to a million Japanese civilians - about
4 percent of the pre-war Japanese population. The total number of deaths in China is believed
to range between 10 and 20 million, with more than 70% being civilians. The other country that
stands out are those that would comprise most of the Soviet Union, where one in seven
perished in the war with about 10 million military deaths and 13 million civilian deaths. In 1939,
there were about 2 billion people in the world. The best estimates indicate that between 62
and 78 million of them would die due to the WW2- more than 3% of the world’s population.
Figure 2 displays number of deaths by type in countries in our SHARE data. In these
countries deaths were highly concentrated in Germany and Poland with deaths measured
4
around five million in both countries. In Germany, there were almost as many civilian deaths as
military ones while in Poland civilian deaths including the holocaust are by far the dominant
Figure 2. WW2 Causalities in SHARE Countries
ones. In many of the remaining countries in our data, deaths due to WW2 are measured instead
in the hundreds of thousands, but still often amount to a large fraction of the pre-war
populations in several other countries, particularly Austria and the Netherlands.
2. Data—SHARE
SHARE is a multidisciplinary and cross-national panel interview survey on health, socioeconomic status and social and family networks of individuals aged 50 or over in continental
Europe. The original 2004/2005 SHARE baseline included nationally representative samples
5
originally in 12 European countries (Denmark, Sweden, Austria, France, Germany, Switzerland,
Belgium, Netherlands, Spain, Italy, and Greece) drawn from population registries, or from multistage sampling (http://www.share-project.org/). For these countries, a second wave of data
collection took place in 2006, and the third wave of data collection (4th wave of data collection)
on this panel was completed in 2011.
In addition to a standard set of demographic attributes (age, marital status, education),
Share data include health variables (self-reported health, health conditions, physical
functioning, health behaviors, psychological variables (e.g., depression and well-being)
economic variables (current work activity, sources and composition of current income, and net
wealth (including housing and financial assets).
SHARE’s third wave of data collection, SHARELIFE, collected detailed retrospective lifehistories in thirteen countries (Poland was added) in 2008-09. SHARELIFE was based on life
history calendar (LHC) methods. The interview starts with the names and birth dates of the
respondent’s children (and other information about them), which is followed by the partner
and residential history. This information is used to aide in the dating of events.
The information in the life history includes family composition and type of home
(running water, toilet, etc.) number of books, occupation of father, and school performance all
at age ten. A childhood health history is also included based on the Smith module included in
the PSID and HRS that queries about individual specific diseases and an overall subjective
evaluation of childhood health status. In addition, respondents are asked about childhood
immunizations, and hunger during childhood. Adult health histories and job and income
6
histories were also collected during childhood. Moreover, SHARELIFE provides detailed data on
region of residence and housing during the whole life (childhood and adulthood).
There is of course concern about the quality of recall data, particularly for time periods
decades in the past. However, that concern has been lessened by a realization that recall of
events during childhood is better than for other periods of life, particularly if the events are
very salient as they certainly are in this application. Smith (2009a) investigated several quality
markers and showed that the instrument was successful in matching known secular trends in
childhood illnesses. There was no evidence of backwards attribution of new episodes of adult
health problems into a revaluation of childhood health. Adult respondents whose health
deteriorated between PSID waves were no more likely than before to say their childhood health
was not good or to cite additional childhood health problems. We will provide evidence in this
paper that these recalled events match the historical record.
One aim of the paper is to illustrate how such retrospective life data can further our
understanding of effects of early-life conditions as affected by large external shocks, such as a
War. The existing literature measuring impacts of macro-events mostly used “natural
experiments” such as wars or famines to study effects of early-life conditions at the aggregate
level. However, largely due to data reasons beyond their control, the studies of which we are
aware could not use individual-level measures of whether a particular person was affected by
the war and through which channel. Retrospective life data such as those from SHARE contain
such detailed information and provide the opportunity of opening the black box of effects of
early-life conditions.
7
Analyzing different outcomes is the first step in understanding the channels and
mechanisms by which wars affect people’s lives. Another possibility is using different measures
of war exposure such as the closeness of combat – are individuals from a more severely
bombed city worse off, because its industry has been destroyed, or because of psychological
trauma. In the SHARE data, retrospective questions provide information on several channels of
war exposure hunger, the absence of the father, dispossession, and persecution.
2.1 Other Data Sources
In addition to SHARE data, we also use external data sources to identify aggregate
channels of war-affectedness. Since WWII affected not only countries differentially, but also
regions within countries, we constructed data on combat operations using sources from
military history (Ellis, 1994). We used maps of within country regions and for each year within
the war we documented whether armies engaged in battle. We combine these data with
information about the region in which the respondents lived during each year of WWII and use
it as a measure of individual war exposure.
Since we analyze data over a time period of 50 years, we also have to account for
country-specific economic performance that may have affected childhood circumstances
differently. We therefore use yearly GDP data, which are available for each European country
(Maddison, 2011). For each respondent we identify the GDP value at age 10 and merge it to the
main data.
3. Channels of Influence
This section presents descriptive data on some of the major channels through which
World War II might have affected peoples’ lives well into their older adult years. These
8
channels include periods of hunger, changing sex ratios, the absence of a father, migration,
dispossession and persecution.
3.1 Hunger
One channel by which the Second World War might have affected long-run adult
outcomes is hunger. World War II caused several severe hunger crises which led to many
casualties, and may have had long-term effects on the health of survivors. For example, since
the beginning of the German occupation in Poland, the nutritional situation of the non-German
population was poor. The average caloric intake for the Polish population was about 930
calories in 1941. The situation was worst in the Warsaw Ghetto with where average food
rations were limited to about 186 calories per day in 1941.
Similarly, in the fall and winter 1941/ 1942, Greece was struck by a severe famine with
about 100,000 to 200,000 deaths (Hionidou, 2006). In WWII, Greece was under Bulgarian,
German and Italian occupation. The famine was mainly caused by three factors: First, the
occupiers imposed a naval blockade. Second, prices to farmers were fixed at such low levels
that they were not willing to market their produce any more. Third, mobility between different
regions of the country was largely reduced due to the occupation. The nutritional situation
returned to acceptable levels towards the end of 1942. Neelson and Stratman (2011) use
Cohort Data to show that undernourishment of those children who were 1 or 2 years old at the
time of the famine had a significantly lower probability of being literate or to complete upper
secondary education.
The nutritional situation in the Netherlands under German occupation had been
adequate up to October 1944 (Trienekens, 2000). As retaliation for a railway strike to hamper
9
German troop movements, Germany blocked food supplies to the parts of the Netherlands that
were still occupied. Even though the blockade was lifted in November 1944, the harshness of
the winter made it impossible to transport food to the affected regions. The combination of the
food blockade and the harsh winter led to a severe hunger crisis in winter 1944/ 1945. About
20,000 deaths, mainly among elderly men, are attributed to this famine. The famine ended with
the German occupation in May 1945.
The Dutch famine has been extensively studied because it affected an otherwise wellnourished population at a very specific time and region. The datasets used include military
records, psychiatric registers, and follow-ups of infants from birth records of clinics in the
affected area (Lumey et al., 2007). Individuals exposed to this famine in utero are shown to
suffer from cognitive and mental problems and addiction (Neugebauer et al., 1999; Rooij et al.,
2010), diabetes and coronary heart disease, but they also perform worse regarding
anthropometric and socio-economic indicators (Almond and Curie, 2011).
Germany suffered from hunger between 1945 and 1947 when the food supply from the
occupied countries had ceased. International aid was at this time not given to Germany. The
Office of Military Government for Germany established a goal of 1550 calories per day in 1945.
However, in the first months of occupation, this goal could often not be met. There were
regions where average calories per day were around 700 (Gimbel, 1968). Death rates raised by
the factor 4 for adults, and by the factor 10 for infants during this period. With a good harvest
and the currency reform of 1948, nutritional shortages were overcome (Zink, 1957).
Do these known well-documented hunger periods show up in the retrospective data?
The descriptive data we present in this section show that they do. Using individual-level data is
10
particularly useful since as will become clear, there is heterogeneity in how individuals were
affected by hunger within countries and over time. In contrast, the aggregate measures we
presented above – for instance, whether an individual lived in a region where combat action
took place – mask potentially important heterogeneity.
The existing literature could only use time and perhaps regional information on such
hunger episodes, and researchers merged this information with individual-level data on health
and economic outcomes. In contrast, SHARE respondents provided individual-level information
on whether and when they suffered from hunger in their lives.
Figure 3 demonstrates that hunger episodes during the time period of the war were
much more severe in war countries than in those countries that did not participate in the war.
We also see that there was a great amount of diversity in periods of hunger amongst the war
countries. Hunger appears to also be more common in regions where combat took place within
war countries. Finally and not surprisingly, the experience of hunger was far more common
among those of low socio-economic background as a child. With respect to hunger, our analysis
shows that the individual-level reports in SHARE match the information on the timing and
location of hunger episodes we collected from historic sources well. To illustrate, in Figure 2 the
Greek spike hunger occurred in 1941-1942, the Dutch in 1944-45, and the German in 1946-
1947.
11
Figure 3. Percentage of Respondents Suffering from Hunger in War Countries versus Non-war
Countries
3.2 Sex Ratios and Absence of Father
Largely men died during the war, producing very low male/female ratios in Europe after
the war. Since this male bias in deaths was concentrated among soldiers as the civilian and
holocaust fatalities were largely gender neutral, it is the countries in Figure 2 who experienced
many military deaths that were most affected. With 3 million military deaths, the most affected
12
country in our data was Germany. In Germany, the sex ratio dropped from 0.98 in 1939 to 0.72
men per women in the 15–45 age group right after the war in 1946. Consequently, many
women did not marry, and many children grew up without a father. Even after the war, about 4
of the 11 million German prisoners of war remained in captivity and the last 35,000 German
soldiers returned from the Soviet Union in 1955 which further compounded the problem of
absent fathers.
Table 2 shows the immediate demographic consequences of the war by listing by
country and period of birth the percent of individuals who had a father absent when they were
ten years old. Once again, the largest effects took place in the war ravaged countries of Austria,
Germany, and Poland. In Austria and German, one in four children lived without their fathers
when they were ten during the war. In Germany, this was twice the rate observed both in the
pre-war and post-war (after 1955) periods in Table 2. In Poland, the rate of absent fathers
during the war is lower (about 15%) since many of the deaths are civilians, but this rate was still
twice that of the pre- and post-war periods. We observe war spikes in other countries as well
(Italy, France, Denmark, and Belgium), but the contrasts with the pre- and post-war years are
not as dramatic.
3.3 Dispossession and Persecution
SHARE documents the extent of the experience of dispossession of property linked to
WWII and its aftermaths.1 This dispossession was often associated with persecution and
resulted in significant geographic displacement of populations during and after the war. A
further advantage of SHARE is that we can observe where and when individuals moved during
13
their lifetimes, including the war time period. Figure 4 shows that inflows of people into
countries that largely stemmed from war displacement including dispossession of property.
There were three principal periods during which people were forced to flee their
homelands. During WWII, millions of Jews, but also opponents of the Nazi regime, were
expropriated, and often send to concentration camps and died there. Second, the end of WWII
was associated with dramatic border changes in Eastern Europe. These border changes often
induced millions of individuals to leave their places of residence and flee to other parts of
Europe. The Soviet Union annexed territory from some of its neighboring countries, inter alia
from Czechoslovakia, Germany and Poland. Poland in term received one part of Germany in
compensation. Those Poles having lost their homes in the part occupied by the Soviet Union
were moved to the new part, so Poland and with it millions of people were moved westwards.
Figure 4 shows the inflow of populations towards and after the end of WWII into the
new states in their new borders. Germany lost about one quarter of its territory. About two
million people have been estimated to have died on the flight. After the war, the remaining
territory of Germany was divided into 4 occupational zones. About 4 million people fled from
the approaching Soviet armies to the British and US zone where the occupation was less severe.
In Germany, destroyed cities had to accumulate millions of ethnic Germans from other parts of
Europe. A further wave of dispossessions happened in Eastern countries after WWII when
private property was nationalized in the socialist and communist economies. Even in France,
there was a wave of nationalizations at the end of WWII. Mainly banks, energy and transport
firms were nationalized, but there were also some expropriations which happened as penalty
for cooperation with the Nazi regime.
14
Figure 5 shows the percentage of dispossessed individuals in SHARE. In the Czech
Republic, Germany and Poland more than 5 percent of the respondents experienced
dispossession during their lifetime. For respondents living in Germany and Poland,
dispossession happened during the war period, while they happened later in Czechoslovakia.
Dispossessed individuals in our sample are over proportionally born outside of the current
boarders of their country. Analyzing countries of origin, many of them came from Eastern
Europe, thus they most probably lost their property with the big wave of nationalizations after
WWII:
Figure 4. Total Inflow of Population due to Displacements after WWII
4. Macro-Impacts of WW2 - The long-term effects of WWII
This paper is primarily about long-term micro effects of WWII, but in this section we
briefly discuss possible long-term macro-effects of the War on demographic and economic
trends. While there remain ripple demographic effects of the War particularly on the aging of
15
Figure 5. Dispossession of Population in WW2
populations, if World War 2 were to alter long-run population growth the main mechanisms
would have to be through future fertility rates and migration patterns. Despite the scale of the
damage done by WW2 on the size, age, and gender composition of populations in the countries
involved, these effects will mostly pass in a generation and certainly in two generations as post
WW2 population structures were eventually replaced by new birth cohorts with more standard
age and gender compositions. If WW2 permanently altered fertility rates, it would have to have
operated through the baby boom and/or the subsequent baby bust, but there is little scientific
evidence that long-term future fertility rates of the countries more affected by WW2 were
affected differentially.
Similarly, long-term migration patterns could have been altered, but that would seem to
largely depend on whether the war was to be associated with future conflicts, civil strife (as
often occurs in persistent civil wars), and unpleasant political and social environments. But with
16
a few exceptions (Hungary, East-Germany until the fall of the Berlin wall ), WW2, especially for
continental Western European countries, was transformative partly in the sense that it help
settle long disputed geographic boundaries, re-united ethnic groups within a single geographic
entity, and led to a period of significant economic growth that lasted for decades. This
transformation, however, took place in two distinct phrases- the immediate post war period
and the end of the cold war almost 50 years later which lead to another era of significant
migration from East to West especially to Germany of some remaining German ethnic groups.
In spite of the death of substantial number of civilians in WWII, the existing evidence also
indicates that cities went back to old population growths paths in Western Germany and Japan
(Brackman et al., 2004), while city growth, but also economic growth was permanently
depressed in Eastern Germany and the Soviet Union (Acemoglu et al., 2011; Brackman et al.,
2004). Beyond these important relocations, there is little reason to think that WW2 lead to
large migrations of populations decades removed from the war. Combined, these comments on
fertility and migration suggest that the long-run demographic changes induced by WW2 are not
large.
If wars alter long-term economic growth, it would permanently depress the economic
prospects of future generations. Warfare reduces capital stock through the destruction of
infrastructure, productive capacity and housing through bombing and fighting, and results in a
relocation of food and other production into military production. It destroys human capital –
but will there be catch-up growth, or will the destruction show up many decades later?
Based on Harrison (1998), Table 3 displays GPD per capita of some of the major
countries involved in the War relative to that of the United States at some key illustrate dates.
17
The immediate impact of WW2 was apparently quite destructive for the countries involved
especially so for those on the losing side – Germany, Japan, and Italy – presumably reflecting
their much large losses in both physical and human capital during the war. However, by 1973
and certainly by 1987,with the exception of the USSR relative to the US, the European ‘losers’
actually had higher per capita growth than the European ‘winners’. What turned out to be
essential in the long-term was not whether a country was on the winning or losing side, but
whether or not they transited to democracy and open-market economies.
Unlike WW1 which settled little, WWII contributed to nation building in Europe. The
nations that were created after the forceful redistribution of millions of people and the redrawing of borders in the end consisted of one single ethnic group (with the exception of
Yugoslavia where this redrawing of borders happened much later through yet another war).
After the traumatic experience of WWII, at least Western Europe established permanent
democratic regimes and became very wary of any kind of warfare.
Waldinger (2011) demonstrated one micro channel on human capital, the loss of Jewish
university professors in Germany due to WWII. He shows that the productivity of those
departments in Germany which lost a relatively high share of their professors was permanently
lowered, while shocks to physical capital due to Allied bombing had returned to their old
growth path by the 1960s. More generally, investments in human capital take years through
schooling alone while plants and factories can be repaired and replaced much more quickly.
The large military mobilizations and the fact that virtually all soldiers and most pre-war
workers were men has been recently used to study the impact of the war on women’s labor
supply on the structure of wages in the labor market. Goldin’s (1990) research suggests that
18
half the entry of women into the labor force during the war remained at the war’s end
providing an arguably exogenous increase in female labor supply. Acemoglu, Autor and Lyle
(2004) use WW2 military mobilization by state and show that these war induced exogenous
increases in female labor supply lowered both female and male wages with more of an effect
on the former and increased the wage gap between college and high school male grads.
Somewhat ironically given the scale of the War and the number of ways it
fundamentally changed the world, the existing economic literature using World War 2 as a
natural experiment is surprisingly thin. Moreover, the literature that does exist using WW2 is
relatively recent and more American in context than European. This may reflect the fact that
the popularity of the ‘natural experiments’ framework in economics itself post-dated WW2 by
many decades. Still, it does suggest that excellent research opportunities remain especially
given the wide diversity of European experiences in the Second World War.
4.1 Measures of War Exposure
To analyze the long-term impacts of WWII on individual health and economic outcomes,
we proceed as follows. Our analysis first makes use of the fact that different countries in
Europe were differentially affected by WWII at different points in time, which yields our
aggregate measures of war exposure. These results are presented in section 4.2. We then go
one step further and analyze the various channels outlined above through which these war
effects are most likely to have occurred. These results are presented in section 4.3.
To study the effects of war on adult outcomes, we use the following indicator of being
affected by World War II- that is one lived in a war country during the war period. Thus, our war
dummy is equal to zero for everybody in a non-war country (Denmark, Switzerland, and
19
Sweden), and for everybody born after the war period. Alternatively, it is equal to one for
everybody who was alive in a war country (Austria, Belgium, Czech Republic, France, Germany,
Greece, Netherlands, and Poland) during the war period. The war period ends in 1945 for all
occupied countries, while it includes 1946 and 1948 in Germany and Austria, when they were
under allied occupation. For these countries, the war period ended with the currency reform in
Germany in 1948.2 Individuals could certainly have been affected by the War even if they were
born after the War, but the channels that we emphasize in this paper – combat, hunger,
dispossession, and the absence of a father – were much more likely to have affected those who
lived during the War.
To gain an understanding of what might underlie any effects of WWII on adult
outcomes, we then construct a variable indicating whether there were combats and how many
combats occurred in the region in which the individual lived during WWII. Thus, in the war
countries, we create two dummy variables based on the number of months of exposure the
respondent had to combat in the place they lived during the war – 0 to 2 months of exposure to
combat and 3 or more months of exposure to combat. The purpose of this variable was to test
whether actual exposure to combat was an important mechanism for the War effects that we
estimate below.
4.2 Micro-Regression Results for Adult Health and SES Outcomes Using the Aggregate War
Measure
We next turn to the question of whether individuals’ experiences during and after WWII
predict their health and socio-economic status in their later adult life. Individuals in our analysis
sample are those who are still alive in 2009 so they are a selected sample of the population that
20
experiences the WW2 shock. To the extent that those more affected are less likely to have
survived, our results should underestimate the full effects of the war on long-term health and
SES outcomes.
We consider several adult dependent variables, all of which are measured in 2009, the
year of the SHARELIFE survey. Health outcomes include the prevalence of diabetes and heart
disease, body height in centimeters (an important summary measure of early-life health
conditions), whether an individual is depressed, and self-reported health status. Self-reported
health status is recorded on a scale from one to five with five being the best health status. Our
adult economic outcomes include years of completed education, the log of household net
worth, whether the individual was ever married, and life-satisfaction in 2009.3 For each of these
dependent variables, we present predictive regressions that use only either the aggregate
measures of war exposure or a measure that distinguishes between very limited exposure to
combat (two months or less, including zero) or an more extensive exposure (3 or more
months). Our individual-level regressors include indicators of SES when a child (separated into
three groups- low, middle, and high),4 whether the respondent was male, and Ln GDP per
capita in respondents’ country of residence at age 10.
We begin with predictive regressions that use only the aggregate measure of war
exposure. Table 4 summarizes results obtained for adult health outcomes and Table 5 does the
same for adult SES outcomes. There are no real surprises for variables other than war exposure
which produce standard types of estimates. Compared to those in the highest childhood SES
group, those in the lowest childhood SES group have higher levels of adult diabetes and heart
disease, are smaller in stature as adults, experience higher levels of adult depression and self-
21
rate themselves in worse adult subjective health. The middle childhood SES group consistently
lies between the bottom and top in terms of these adult health outcomes. These results
conform to the general finding in the literature that childhood economic circumstances are very
predictive of later life adult economic and health outcomes (Currie, 2009; Case et al., 2002; and
Smith, 2009b.) Consistent with the epidemiological literature, men have higher levels of adult
diabetes and heart disease, lower levels and depression, and report themselves in better
subjective health than woman do (Banks et al., 2010, Smith, 2007). Individuals who lived in
countries with higher GDP per capita when they were ten years old are taller as adults, have
lower levels of heart disease and better self-reported adult health in 2009 (Deaton, 2007).
Our principal concern involves the estimated impact of the aggregate War variable.
Living in a war country during the period of the Second World War is consistently associated
with poorer adult health for all outcomes in Table 4 except adult height. Individuals exposed to
WW2 have higher levels of adult diabetes and heart disease, are more depressed, and report
their subjective health as worse. In the sixth column of Table 4, we include one additional
model that predicts whether a respondent received immunizations as a child. This model shows
that the experience of war was associated with a lower probability of immunization as a child
which is not surprising given that this was war time. This immunization result may be one
pathway through which adult health would eventually suffer.
Table 5 repeats the same type of models for adult economic outcomes in 2009. Not
surprisingly, higher childhood SES is associated with much higher levels of adult education, net
worth and life satisfaction, another indication of the strong economic transmission across
generations in these outcomes. Similarly, the male advantage in these adult SES outcomes is
22
consistently found. Our measure of country - year War exposure is strongly negatively related
to education and life satisfaction as an adult. War exposure is associated with about 6 tenths of
a year less education. In contrast, ln household net worth is not associated with the war time
experience suggesting that this outcome mainly depends on post war savings behavior and
trends in asset prices. The marriage effect of war is statistically significant for women only
consistent with the sharp decline in the male/female adult sex ratio documented above.
Tables 6 and 7 show results obtained from an parallel set of models with the same adult
health and SES outcomes and the same set of right hand side variables with the exception of
the war exposure measure. The estimated effects of all non-war variables are essentially the
same as in Tables 4 and 5. In Tables 6 and 7, our war exposure measure consists of two
variables which measure the number of months of exposure the respondent had to combat in
the place they lived during the war in war countries - 0-2 months of exposure to combat and 3
or more months of exposure to combat. The purpose of this variable was to test whether the
actual exposure to combat was an important mechanism for the War effects that we estimated
below. With the sole exceptions of adult depression (Table 6) and live satisfaction (Table 7), the
worse adult SES and health outcomes appear to be about the same amongst those with large or
small exposures to combat. This suggests that having combat and battles close by to where you
lived during the war are not mechanism by which these war effects operate. The exceptions are
of interest since it is reasonable that exposure to frequent exposure to combats is associated
with adult depression and lower levels of life satisfaction as the vivid memories of war linger
long into adulthood.
23
4.3 Models of Channels of War
We next turn to our estimates of how the micro pathway channels we highlighted
above- hunger, dispossession, persecution, and the absence of father are related to the
experience of WWII. SHARE respondents were asked ‘Looking back on your life was there a
period of time during which you were hungry?’ If the answer was yes, they were asked when
this occurred. In SHARE, individuals are also asked “whether they or their family were ever
dispossessed of any property as a result of war or persecution”, and if yes the date of that
dispossession. They were also asked whether they had ever been victims of persecution
because of their political beliefs, religion, nationality, ethnicity, sexual orientation or their
background. Finally, the absence of the father is defined as the absence of the biological father
at the age of ten. About 8 percent of our respondents experienced a period of hunger, 9
percent have lived without their father at age 10, and 5 percent suffered from persecution and
dispossession, respectively.
Table 8 shows how these micro channels are related to the experience of war. Having
experienced WWII increases the likelihood of experiencing hunger by 5 percentage points,
persecution by 1 percentage point and the absence of the father by 4 percentage points. These
estimates are large relative increases given the baseline risks. Low childhood SES increases the
probability of suffering from huger, while it decreases the probability to be dispossessed. The
latter is logical, because the prerequisite for dispossession is that there has to be something to
be dispossessed of. The experience of war makes these results even stronger. Males are both
more likely to suffer from hunger, but also to be persecuted. The latter is again what we
expected given that mainly men were politically active during this period of time.
24
In Table 9, our interest lies again in whether it is the experience of combats is the
mechanism that leads to our War effects. The experience of hunger and the absence of the
father is somewhat stronger for our respondents who lived in a region strongly affected by
combats (3-10 months of combats) then for those in regions with none or only mild experience
of combats. However, the differences are not large. In fact, we expect the death of men during
wartime to not necessarily happen in their region of residence. Persecution is expected to war
per se, but not to an increased experience of combats. Thus, combats do come with an
increased likelihood of hunger as for example was the case in the Dutch hunger winter. It can
also be due to other aspects of war, as was the case for the Greek and German experience with
hunger during WWII. Also, combats lead to local deaths of the civilian population, but military
casualties often occur far from the families affected by it.
4.4 The Uneven Consequences of War
In addition to the models summarized in Tables 4-95, we also estimated models that
included interactions of the War variable with SES as a child. These results are contained in
Table 10. In these models, we include both main effects for the war and childhood SES status as
well as interactions of the war with childhood SES. For these SES interactions, among the health
variables, we find the strongest SES war interactions with heart disease, possibly reflecting the
role of lifetime stress with that disease. However, we find very strong interactions of SES and
war for adult education, adult depression, and adult life satisfaction. The negative impact of
war on all three of these outcomes is much larger among those in the low and middle SES
childhood groups. The bottom panel of Table 10 shows that some of the pathways through
25
which War operates are concentrated among the poorest households (ie. hunger and
immunizations) and some are concentrated among the highest SES households (dispossession).
5. Conclusions
In this paper, we present a micro analysis of the effects of World War 2 on some of the
principal SES and health outcomes of those who experienced the war. To conduct this analysis
we use new data-SHARELIFE- that records not only the adult outcomes in 2009, but also the
salient aspects of the war time experiences of these respondents. Our analysis shows that
experiencing war increased the probability of suffering from heart disease, diabetes, and also
depression. Individuals experiencing war or combat have significantly lower self-rated health,
and they are smaller. Experiencing war is also associated with lower years of education and life
satisfaction, and it decreases the probability of ever being married for women, while increasing
it for men. We also analyze the pathways through which these War time effects took place. We
found strong effects for our four pathways we explored - hunger, dispossession, persecution,
and having an absent father.
26
Endnotes
1 The specific question is “There may be cases when individuals and their families are
dispossessed of their property as a result of war or persecution. Were you or your family ever
dispossessed of any property as a result of war or persecution?”
2 We leave out Spain from our regressions, because of the Spanish Civil War of 1936 to 1939.
3 Respondents are asked “On a scale from 0 to 10 where 0 means completely dissatisfied and 10
means completely satisfied, how satisfied are you with your life?” We model this outcome as a
score from 0-10.
4 For childhood SES, we construct an index using information about the socio-economic
background of the respondents when they were ten years old. The variables contributing to our
SES measure are: A dummy indicating whether the father was a white collar worker; the
logarithm of the number of books in the household; the logarithm of the proportion of
individuals in the household who shared a room; information on whether the house had a
bathroom, hot and cold water supply, central heating and an inside toilet. The respective
weights of the different factors are determined through factor analysis. Mazzonna (2011)
shows that the distribution of this index is consistent with historical information on the wealth
distribution within these countries.
27
References
Acemoglu, Daron, David Autor, and David Lyle, “Women, War, and Wages. The Effect of Female
Labor Supply on the Wage Structure at Midcentury,” Journal of Political Economy 112
(2004), 497-551.
Acemoglu, Daron, Tarek Hassan, and James Robinson, “Social Structure and Development: A
Legacy of the Holocaust in Russia” (2011).
Almond, Douglas, and Janet Currie, “Killing Me Softly - The Fetal Origins Hypothesis” (2011).
Auerbach, Hellmuth, “Opfer der nationalsozialistischen Gewaltherrschaft,” in Wolfgang Benz
Legenden, Lügen, Vorurteile. Ein Wörterbuch zur Zeitgeschichte,” dtv (1992).
Banks, James. Alastair Muriel, and James P Smith, “Disease Prevalence, Disease Incidence, and
Mortality in the United States and in England,” Demography 47 (2010), S211- S231.
Brackman, Stephen, Harry Garretsen, and Marc Schramn, “The Strategic Bombing of German
Cities during World War II and its Impact on City Growth,’ Journal of Economic
Geography 4 (2004), 201-218.
Case, Anne, David Lubotsky, and Christina Paxson, “Economic Status and Health in Childhood:
The Origins of the Gradient,” American Economic Review 92 (2002), 1308-1334.
Currie, Janet, “Healthy, Wealthy, and Wise: Socioeconomic Status, Poor Health in Childhood,
and Human Capital Development,” Journal of Economic Literature 47 (2009), 87–122.
Deaton, Angus, “Height, Health, and Development,” PNAS-Proceedings of the National Academy
of Sciences 104 (2007), 13232-13237.
Ellis, John, World War II – A Statistical Survey (London: Aurum Press, 1994).
28
Ernst J. Franzek, Niels Sprangers, A. Cecile, J. W. Janssens, Cornelia M. Van Duijn, and Ben J. M.
Van De Wetering, “Prenatal Exposure to the 1944–45 Dutch ‘Hunger Winter’ and
Addiction Later in Life,” Addiction 103 (2008), 433-438..
Harrison, Mark, “The Economics of World War II: An Overview,” in The Economics of World War
II: Six Great Powers in International Comparison (Cambridge: Cambridge University
Press, 1998, 1-42).
Hionidou, Violetta, Famine and Death in Occupied Greece, 1941-1944 (Cambridge: Cambridge
University Press, 2006).
Gimbel, John, The American Occupation of Germany: Politics and the Military, 1945-1949 (Palo
Alto, CA: Stanford University Press, First Edition, 1968).
Goldin, Claudia, “The Role of World War II in the Rise of Women’s Employment,” American
Economic Review 81 (1990), 741-756.
Lumey, L.H., Aryeh D. Stein, Henry S. Kahn, Karin M. van der Pal-de Bruin, G. J. Blauw, Patricia A
Zybert, and Ezra S. Susser, “Cohort Profile: the Dutch Hunger Winter Families Study,”
International Journal of Epidemiology 36 (2007), 1196-1204.
Maddison, Angus, “Historical Statistics: Statistics on World Population, GDP and per capita
GDP,” 1-2008. http://www.ggdc.net/MADDISON/oriindex.htm (accessed June, 2011).
Mazzonna, Fabrizio, “The Long-Lasting Effects of Family Background: A European Cross-Country
Comparison,” MEA working paper (2011).
Neelsen, Sven, and Thomas Stratman, “Effects of Prenatal and Early Life Malnutrition: Evidence
from the Greek Famine,” Journal of Health Economics 30 (2011), 479-488.
29
Neugebauer, R., H. W. Hoek, and E. Susser, “Prenatal Exposure to Wartime Famine and
Development of Antisocial Personality Disorder in Early Adulthood,” JAMA 282 (1999),
455–462.
Pohl, Dieter, “Verfolgung und Massenmord in der NS-Zeit 1939-1945,” Wissenschaftliche
Buchgesellschaft, Darmstadt (2003).
Rooij, W. H. Sr., J. E. Yonker, R. C. Painter, and T J. Roseboom, “Prenatal Under Nutrition and
Cognitive Function in Late Adulthood,” PNAS, Proceedings of the National Academy of
Sciences USA 107 (2010), 16881-16886.
SHARE Release Guide 2.5.0 for Waves 1 & 2, 2011, http://www.share-project.org/sharelife/
(accessed June 07, 2011).
SHARELIFE Release Guide 1, 2010, http://www.share-project.org/sharelife/ (accessed June 07,
2011).
Smith, James P., “Nature and Causes of Male Diabetes Trends, Undiagnosed Diabetes, and the
SES Health Gradient,” PNAS; Proceedings of National Academy of Sciences 104
(2007),13,225-13,231.
Smith, James P., “The Impact of Childhood Health on Adult Labor Market Outcomes,” The
Review of Economics and Statistics 91 (2009a), 478-489.
Smith, James P., “Re-constructing Childhood Health Histories,” Demography 46(2009b), 387-
403.
Trienekens, G., “The Food Supply in The Netherlands during the Second World War,” in: D. F.
Smith, and J. Phillips (eds), Food, Science, Policy and Regulation in the Twentieth
30
Century: International and Comparative Perspectives (London: Routledge, 2000, 117–
133).
Waldinger, Fabian, “Bombs, Brains, and Science. The Role Human and Physical Capital for the
Creation of Scientific Knowledge” (2011).
Zink, Harold, The United States in Germany1944-1955 (Princeton, NJ: D. Van Nostrand, 1957).
31
Table 1
Number of Observations Available in SHARELIFE
Country Observations
Austria 759
Germany 1,755
Sweden 1,800
Netherlands 2,032
Spain (excluded due to the Civil War) 1,731
Italy 2,354
France 2,252
Denmark 1,990
Greece 2,689
Switzerland 1,176
Belgium 2,655
Czech Republic 1,732
Poland 1,648
Total 24,573
Note: The countries are defined as of 2009 when the SHARELIFE data were collected.
32
Table 2
Percent with Father Absent at age 10 by Time Period
Pre-1939 1939-1955 After 1955
Austria 0.198 0.244 0.151
Germany 0.113 0.279 0.099
Sweden 0.114 0.130 0.123
Netherlands 0.076 0.073 0.045
Spain 0.118 0.089 0.055
Italy 0.045 0.070 0.046
France 0.109 0.131 0.055
Denmark 0.073 0.105 0.077
Greece 0.019 0.053 0.013
Switzerland 0.068 0.039 0.046
Belgium 0.048 0.082 0.055
Czechia 0.081 0.100 0.081
Poland 0.080 0.153 0.082
33
Table 3
GDP per Head Relative to the USA GDP per Head
1938 1950 1973 1987
UK .98 .72 .72 .73
Germany .84 .45 .79 .82
France .72 .55 .78 .78
Italy .53 .36 .63 .70
Japan .38 .20 .66 .77
USSR .35 .30 .36 .33
Source. Harrison 1998. Table 1-10.
34
Table 4
Adult Health Outcomes Associated with World War 2
(1) (2) (3) (4) (5) (6)
VARIABLES Diabetes Heart Height Depression
Self-reported
Health
Childhood
Immunizations
War 0.035*** 0.020*** -0.232 0.028** -0.081*** -0.025***
[0.007] [0.007] [0.147] [0.011] [0.024] [0.006]
Childhood SES low 0.034*** 0.020*** -1.825*** 0.030*** -0.331*** -0.029***
[0.006] [0.006] [0.127] [0.009] [0.020] [0.004]
Childhood SES med 0.019*** 0.015*** -0.923*** 0.008 -0.182*** -0.014***
[0.005] [0.005] [0.107] [0.008] [0.017] [0.003]
Male 0.015*** 0.044*** 11.616*** -0.167*** 0.110*** -0.002
[0.004] [0.004] [0.089] [0.007] [0.014] [0.003]
Ln GDP age 10 -0.008 -0.047*** 1.365*** 0.017 0.169*** 0.022*
[0.015] [0.016] [0.299] [0.022] [0.046] [0.013]
Constant 0.136 0.600*** 150.330*** 0.326* 1.134*** 0.647***
[0.126] [0.138] [2.507] [0.185] [0.384] (0.111)
Observations 19,687 19,687 19,704 19,704 19,695 19,511
R-squared 0.019 0.053 0.515 0.067 0.137 0.075
Models include Country dummies and ten year age bands.
Robust standard errors in brackets.
*** p<0.01, ** p<0.05, * p<0.1.
35
Table 5
Adult Economic Outcomes Associated with World War 2
(1) (2) (3) (4)
VARIABLES Education Ln Net Worth
Ever
Married
Life
Satisfaction
War -0.569*** 0.000 -0.017*** -0.118***
[0.105] [0.035] [0.006] [0.039]
Childhood SES low -4.053*** -0.659*** 0.011** -0.421***
[0.089] [0.031] [0.005] [0.034]
Childhood SES med -2.480*** -0.313*** 0.006 -0.160***
[0.078] [0.025] [0.004] [0.027]
Male 1.060*** 0.223*** -0.017*** 0.189***
[0.063] [0.021] [0.005] [0.023]
Ln GDP age 10 1.019*** 0.158** -0.016 0.052
[0.201] [0.072] [0.010] [0.078]
Currently married 0.884***
[0.041]
War male 0.024*** 0.189***
[0.007] [0.023]
Constant 2.761* 8.548*** 1.083*** 7.044***
[1.670] [0.613] [0.086] [0.657]
Observations 12,676 18,964 19,704 18,127
R-squared 0.363 0.368 0.008 0.127
Models include Country dummies and ten year age bands
Robust standard errors in brackets
*** p<0.01, ** p<0.05, * p<0.1
36
Table 6
Adult Health Outcomes Associated with Months of World War2 Combat Experience
(1) (2) (3) (4) (5) (6)
VARIABLES Diabetes Heart Height Depression Health
Childhood
Immunizations
War Combat 0-2 months 0.038*** 0.018** -0.274* 0.020* -0.079*** -0.025***
[0.007] [0.007] [0.154] [0.012] [0.025] [0.005]
War combat 3 or more months 0.026*** 0.022*** -0.116 0.049*** -0.086*** -0.024***
[0.009] [0.010] [0.184] [0.014] [0.030] [0.006]
Childhood SES low 0.034*** 0.020*** -1.821*** 0.030*** -0.332*** -0.029***
[0.006] [0.006] [0.127] [0.009] [0.020] [0.004]
Childhood SES med 0.019*** 0.015*** -0.916*** 0.008 -0.182*** -0.013***
[0.005] [0.005] [0.107] [0.008] [0.017] [0.003]
Male 0.015*** 0.044*** 11.613*** -0.166*** 0.110*** 0.001
[0.004] [0.004] [0.089] [0.007] [0.014] [0.003]
Ln GDP age 10 -0.009 -0.048*** 1.384*** 0.019 0.170*** 0.022*
[0.015] [0.016] [0.300] [0.022] [0.046] [0.013]
Constant 0.143 0.603*** 150.059*** 0.312* 1.125*** 0.646***
[0.126] [0.138] [2.511] [0.185] [0.384] [0.110]
Observations 19,675 19,675 19,692 19,692 19,683 19,503
R-squared 0.020 0.053 0.515 0.067 0.137 0.075
Models include Country dummies and ten year age bands
Robust standard errors in brackets.
*** p<0.01, ** p<0.05, * p<0.1.
37
Table 7
Adult Economic Outcomes Associated with Months of World War2 Combat Experience
(1) (2) (3) (4)
VARIABLES Education Ln Net worth Ever Married Life Satisfaction
War combat 0-2 months -0.586*** 0.021 -0.015* -0.072*
[0.109] [0.036] [0.007] [0.041]
War combat 3 or more months -0.529*** -0.054 -0.020* -0.234***
[0.130] [0.047] [0.008] [0.050]
War combat 0-2 months * Male 0.022***
[0.007]
War combat 3 or more months *
Male
0.028***
[0.009]
Childhood SES low -4.054*** -0.657*** 0.011** -0.421***
[0.089] [0.031] [0.005] [0.034]
Childhood SES med -2.477*** -0.312*** 0.006 -0.161***
[0.078] [0.025] [0.004] [0.027]
Male 1.059*** 0.222*** -0.017*** 0.188***
[0.063] [0.021] [0.005] [0.023]
Ln GDP age 10 1.019*** 0.157** -0-0.016 0.045
[0.201] [0.072] [0.010] [0.078]
Currently married 0.885***
[0.041]
Constant 2.773* 8.554*** 1.087*** 7.098***
[1.673] [0.613] [0.086] [0.658]
Observations 12,672 18,955 19,692 18,117
R-squared 0.362 0.369 0.008 0.128
Models include Country dummies and ten year age bands.
Robust standard errors in brackets.
*** p<0.01, ** p<0.05, * p<0.1.
38
Table 8
Channels of War Associated with World War2
(1) (2) (3) (4) (5) (6) (7) (8)
VARIABLES Hunger Hunger Dispossession Dispossession persec persec
Dad
Absent Dad Absent
War 0.052*** 0.047*** 0.004 0.014** 0.010** 0.016** 0.037*** 0.032***
[0.005] [0.008] [0.004] [0.007] [0.004] [0.007] [0.006] [0.009]
Childhood SES low 0.013** 0.001 -0.017*** -0.013*** -0.015*** -0.011** -0.003 0.006
[0.005] [0.004] [0.004] [0.004] [0.004] [0.005] [0.005] [0.007]
Childhood SES med -0.004 -0.002 -0.017*** -0.011*** -0.012*** -0.010** 0.005 -0.003
[0.004] [0.003] [0.003] [0.004] [0.003] [0.004] [0.005] [0.006]
War kid SES low 0.025** -0.010 -0.010 -0.016
[0.011] [0.008] [0.008] [0.011]
War kid SES med -0.003 -0.016** -0.007 0.019*
[0.009] [0.007] [0.007] [0.010]
Male 0.015*** 0.015*** 0.003 0.003 0.008*** 0.008*** -0.002 -0.002
[0.003] [0.003] [0.002] [0.002] [0.003] [0.003] [0.004] [0.004]
Ln GDP age 10 -0.018 -0.018 -0.002 -0.002 0.005 0.005 -0.021 -0.020
[0.015] [0.015] [0.009] [0.009] [0.008] [0.008] [0.014] [0.014]
Constant 0.290** 0.293** 0.090 0.085 -0.008 -0.011 0.355*** 0.356***
[0.123] [0.123] [0.073] [0.073] [0.071] [0.071] [0.120] [0.120]
Observations 19,693 19,693 19,693 19,693 19,695 19,695 19,703 19,703
R-squared 0.076 0.077 0.041 0.041 0.016 0.016 0.031 0.031
Models include Country dummies and ten year age bands
Robust standard errors in brackets
*** p<0.01, ** p<0.05, * p<0.1.
39
Table 9
Channels of War Associated with Months of World War2 Combat Experience
(1) (2) (3) (4)
VARIABLES Hunger Dispossession Persecution Dad Absent
War combat 0-2 months 0.046*** 0.004 0.012*** 0.032***
[0.005] [0.004] [0.005] [0.007]
War combat 3 or more months 0.066*** 0.002 0.005 0.049***
[0.007] [0.005] [0.005] [0.009]
Childhood SES low 0.013** -0.016*** -0.015*** -0.003
[0.005] [0.004] [0.004] [0.005]
Childhood SES med -0.004 -0.017*** -0.013*** 0.005
[0.004] [0.003] [0.003] [0.005]
Male 0.015*** 0.003 0.008*** -0.002
[0.003] [0.002] [0.003] [0.004]
Ln GDP at age 10 -0.017 -0.003 0.004 -0.020
[0.015] [0.009] [0.008] [0.014]
Constant 0.277** 0.091 0.000 0.346***
[0.123] [0.073] [0.071] [0.120]
Observations 19,683 19,683 19,685 19,693
R-squared 0.077 0.041 0.016 0.031
Models include Country dummies and ten year age bands.
Robust standard errors in brackets.
*** p<0.01, ** p<0.05, * p<0.1.