Economics and Human Biology 9 (2011) 30–44
Contents lists available at ScienceDirect
Economics and Human Biology
journal homepage: http://www.elsevier.com/locate/ehb
The relationship between height and economic development
in Spain, 1850–1958§
Ramón Marı́a-Dolores a,*, José Miguel Martı́nez-Carrión b
a
Departamento de Fundamentos del Análisis Económico, Facultad de Economı´a y Empresa, Universidad de Murcia,
Campus de Espinardo 30100 Espinardo, Murcia, Spain
b
Departamento de Economı´a Aplicada, Facultad de Economı´a y Empresa, Universidad de Murcia, 30100 Murcia, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
Received 15 December 2009
Received in revised form 20 July 2010
Accepted 20 July 2010
We investigate the relationship between height and some economic-development
indicators in modern Spain by means of a recently constructed times series of data on
conscripts. We estimate a Vector Autoregressive Equilibrium Correction Model (VECqM)
to quantify the height and GDP per capita response to various living-standard indicators.
We observe that conditions that perpetuate an elevated level of sickness and mortality
and that raise the relative price of consumption goods tend to impede human growth, as
reflected in a decline in average adult height, whereas factors that promote the purchase
of health services and that help to open up the economy to international trade and ideas
have tended to have an opposite effect from the 1850s onward. Our results also indicate
that neither the level of per capita GDP nor its growth rate has a unidirectional relation to
various measures of living standards, chiefly adult stature. Instead, our findings suggest
that there may be behavioral factors (e.g., emphasis on health services), political factors
(e.g., degree of openness), and economic factors (e.g., relative consumer costs to GDP
deflator) whose affects may have been influenced the level of GDP, over the sample
period.
Crown Copyright ß 2010 Published by Elsevier B.V. All rights reserved.
JEL classification:
I1
I3
N3
N9
Keywords:
Height
Health
Income
Education
Economic development
Cointegration
1. Introduction
In the last few decades anthropometric history has
enhanced our knowledge of the quality of life of numerous
populations. The use of physical stature and other
anthropometric indicators is widely accepted by economic
historians, and recently by economists as well, for the
§
We are grateful to Máximo Camacho, Rafael Dominguez, Rui Esteves,
Jan Jacobs, John Komlos, Gloria Quiroga, Nikola Koepke, Leandro Prados
de la Escosura, Hipólito Simón, and anonymous referees for their useful
comments. This paper was prepared while the first author was visiting
the Department of Economics at the University of Oxford. This paper has
been written with the support of project SEJ2007-67613 (MICIIN). A first
version of this paper was published as a working paper by the Fundación
de las Cajas de Ahorro (FUNCAS Working Paper Series 556/2010).
* Corresponding author. Tel.: +34868887908.
E-mail address: [email protected] (R. Marı́a-Dolores).
measurement of various aspects of human well-being and
in the examination of the impact of socio-economic
processes on biological welfare and health outcomes
(Fogel, 1994; Komlos, 1985, 1994, 1995a, 1998; Steckel,
1995, 2008; Steckel and Floud, 1997). Some of the main
topics debated are the determinants of height and the
relationship between stature and indicators of economic
development. A high degree of correlation between height
and economic development was observed in developing
countries in the second half of the 20th century (Steckel,
1979, 1983). Later contributions have clarified the existing
relations among height, income per capita, and income
inequality over the long term (Brinkman et al., 1988;
Drukker and Van Meerten, 1995; Coll, 1998; Craig and
Weiss, 1998; Haines, 1998; Jacobs and Tassenaar, 2004;
Moradi, 2010; Peracchi, 2008). In recent years a battery of
living-standard indicators has been used to measure their
1570-677X/$ – see front matter . Crown Copyright ß 2010 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.ehb.2010.07.001
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
relationships to height, revealing important connections
among health, mortality, and economic development
(Easterlin, 2000; Arora, 2001; Deaton, 2003, 2007; Deaton
and Arora, 2009; Fogel, 2004; López-Casasnovas et al.,
2005; Persico et al., 2004; Steckel, 2008).
In cross-sectional analyses, the correlations between
height and levels of economic development are positive
(Komlos, 1995b, 2003). However, in longitudinal analyses,
the correlations are less clear cut. A quarter of a century
ago it was established that time series of income per capita
could explain variations in height changes, and that stature
can be used as a proxy for income in the absence of
information on material well-being (Brinkman et al.,
1988), although it is well known that the two indicators
do not necessarily evolve in parallel. Many scholars have
studied this relationship. Several have found that in
developing countries in Europe, North America, and Japan
in the late-19th and early-20th centuries average height
tended to increase, regardless of the level of industrialization (Sandberg and Steckel, 1997; Weir, 1997; Shay, 1994;
Honda, 1997; Baten, 2000; Federico, 2003; Vecchi and
Coppola, 2006; Arcaleni, 2006). Height declined, however,
in the United States, England, and the Netherlands during
the Industrial Revolution, at the end of the 18th and the
first half of the 19th centuries (Margo and Steckel, 1983;
Floud et al., 1990; Komlos, 1998; Haines, 2004). Nevertheless, nearly everyone agrees that there was an international convergence in biological welfare and other non-income
indicators of the standard of living during the 20th century
(Kenny, 2005; Deaton, 2007, 2008) which was primarily
due to the fact that stature is influenced by (among other
factors) health conditions during childhood and adolescence, maternal education, social-welfare policies, child
labour, location, and cultural values (Komlos and Baten,
1998; Shutkowski, 2008; Steckel, 2009).
More recently, other researchers (Baten et al., 2010)
have pointed out that human stature in adulthood captures
an accumulation of factors that contribute to changes in
the standard of living of a generation. This implies that GDP
per head could also explain height averages.
Evidence of the relationship between height and
economic development has also been found in Spain.
The first studies of anthropometric history there were done
by economic historians who used military data. Thus,
Gómez-Mendoza and Pérez-Moreda examined the relationships among height, educational attainment, and
infant mortality in the early 20th century. They compared
the average height of recruits by province, along with data
on the region’s economic performance and infant-mortality rate (Gómez-Mendoza and Pérez-Moreda, 1995).
Furthermore, Martı́nez-Carrión explored height and income trends between 1850 and 1990 on the basis of local
military-recruitment data (Martı́nez-Carrión, 1994, 2002;
Martı́nez-Carrión and Pérez-Castejón, 2000); and Quiroga
Valle and Coll (2000) discussed height differences as a
proxy for income inequality, and found that changes in
height differences among social groups could indicate
shifts in income inequality. By expanding beyond the
conventional indicators of welfare, the anthropometric
approach to socio-economic history opens up a wide
variety of research possibilities.
31
Recently, using both Spanish height data from the
Encuesta Nacional de Salud (Spanish National Health
Survey) and the dataset of the European Community
Household Panel (Eurostat), economists and demographers presented new evidence concerning the evolution
of adult height and weight (Garcia and QuintanaDomeque, 2007, 2009). The influence of generational
and environmental factors on adult height has been
evaluated, and the mechanisms by which socio-economic
position may influence individual height have been
discussed (Costa-Font and Gil, 2008; Spijker et al.,
2008). Moreover, the issue of the determinants of height
during a period of significant socio-economic transformation (1960–2000) has been explored. Results suggest
that the epidemiological transition before Spain’s entry
into the European Union led to increases in adult height
(Bosch et al., 2009).
Spanish economic growth in the last few years has
traced a parallel path. According to the estimates of Prados
de la Escosura, Spain’s long-term underperformance was
mostly due to its sluggish growth between 1850 and 1950.
In contrast, it was the destruction of human capital during
the Spanish Civil War (1936–1939) and its aftermath that
explains its poor performance during the 1940s and 1950s.
The 1940s constituted a phase of delay in the Spanish
economy – an economy that has been catching up with
those of advanced countries over the last fifty years, in
which 1959–1974 stands out as a period of outstanding
performance (Prados de la Escosura, 2003, 2007).
Comparison between the performance of the Spanish
economy and that of the most advanced nations of
Western Europe reveals that Spain’s GDP was approximately 75% of Europe’s and 50% of North America’s in the
early 1930s. The civil conflict of 1936–1939 caused the
GDP to decline, and although it improved in the 1950s, the
gap between it and the GDPs of developed nations
remained unchanged. According to recent estimates,
Spain’s per capita GDP stood in 1933 and decreased to
35% in the 1950s. The development of a pro-market
attitude, marked by deregulation and a gradual entry into
the international economy, resulted in sustained growth
and thus a closing of the GDP gap between it and the rest of
Western Europe during the second half of the twentieth
century. A dramatic slowdown in growth followed by a
sustained – recovery – the pivotal year being 1986, when
Spain joined the European Union – characterized the last
quarter of the twentieth century (Prados de la Escosura,
2007).
We analyse the relationship between physical stature
and economic development in Spain between 1850 and
1958 on the basis of previously with a new height
database, drawn from military-conscript records. We take
into account several key development indicators: per
capita income, the share of consumption of health series in
total consumption, the ratio of the price of consumer goods
to the GDP deflator, schooling and mortality rates, and the
degree of openness, which could potentially influence
physical stature during the period under consideration. We
study the relationship between various indicators of the
standard of living without presuming that there exists any
interrelationship among them.
32
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
Drawing on these anthropometric data, we address the
following three questions: What are the long-run trends
among height and various indicators of economic development in Spain? How are height and per capita GDP
influenced by certain economic-development indicators?
Is per capita GDP a function of height or vice versa, or is
each a function of the other?
The rest of the paper is organized as follows. In Section 2
we describe our database; in Section 3 we apply a Vector
Auto-Regressive (VAR) framework with a Vector Equilibrium Correction Model in order to identify the determinants
of height changes and other standard of living indicators in
Spain between 1850 and 1978; and in Section 4 we offer
our conclusions.
2. Data description
2.1. Height
In order to analyze secular height changes, a standardized time-cohort series can be constructed for Spain from
the 1850s onwards,1 thanks to Local Military Recruitment
Acts, or LMRA.2 These data sources are preserved in the
Sección de Quintas of the municipal archives. Earlier data are
both scarce and fragmentary; in fact, scarcely any were
recorded until the end of the eighteenth century. The first
military draft dates to 1770. Until the end of that century, the
conscripts were between 16 and 40 years of age. In the first
decades of the 19th century the age range was reduced to
16–25 (Cámara-Hueso, 2006). Not until the 1850s, however,
is recruitment limited to a one-year cohort, which is
modified over the course of the next 120 years: 20 years
(1850–1885), 19 years (1885–1900), 20 years (1901–1906),
and 21 years (1907–1970).
Our height database include records of all the conscripts
in 19 municipalities spanning three geographical regions
(Andalusia, the Region of Murcia, and the Valencian
Community) across from five provinces of Spanish Levant.3
The municipalities are representative of different socioeconomic environments: cities, towns, and villages.
There are reasons to believe that the relationship found
for these provinces of Spanish Levant are representative of
the whole country. According to several studies, when
considered according to total productivity, labour productivity, income per capita, well-being, and inequality the
Levant ranked in the middle (Domı́nguez-Martı́n, 2002;
German et al., 2001; Núñez, 2005). However, when it
comes to height, in several anthropometric studies of
Spanish Levant prove to be nearer to the national average.
The average height of the 1915–1929 and 1965–1980
recruits (year of measurement) are in line with average
1
A state-of-the-art database for stature and other elements of Spanish
anthropometric history; see the Historia Agraria monograph coordinated
by Martı́nez-Carrión (2009).
2
The original records are Actas de clasificación de los mozos y declaración
de soldados, Actas de Reclutamiento y Reemplazo.
3
The municipalities are the next: Vera (province of Almeria); Region of
Murcia: Cartagena, Cieza, Murcia, Mazarrón, Totana, Torre-Pacheco Yecla
(province of Murcia); Alcoi, Elche, Orihuela, Pego, Villena (province of
Alicante); Alzira, Gandı́a, Requena, Sueca (province of Valencia); and
Castellón y Villareal (province of Castellón).
heights nationwide at the time (Gómez-Mendoza and
Pérez-Moreda, 1995, p. 85; Martı́nez-Belmonte, 1983;
Martı́nez-Carrión, 1994, p. 697; Rebato, 1998).4
We determine each cohort’s average height by year of
birth. The height series that we use comprises data on
328,248 conscripts between the ages of 19 and 21 who
were measured between 1857 and 1969 or who were born
between 1837 and 1948. The backgrounds of these series
have been analyzed by Martı́nez-Carrión and PérezCastejón (1998, 2000) and incorporate the recent dataset
of Puche-Gil (2009).5 Our height series is linked with the
national series collected by the Statistics of Recruitment
and Replacement of the Armies from 1955 and estimated
by Quiroga (2003a). In that paper, the height series starts in
1850 as do the economic-development indicators.
The chief problems posed by height data from any
military source are changes in the recruitment age and in
the minimum-height requirement. In Spain, however, all
recruits were measured before being declared fit or unfit
for military service, so there is no selective bias on account
of a truncated distribution. The potential problem posed by
changes in the recruitment age and by the rounding of the
height data, has been resolved by our using standardized
heights at age 21 that take into account estimations of
growth.6 We constructed a new height series by birth
cohorts using five-year moving averages standardized at
the age of 21.7
We present distributions of height data for several
periods in Fig. 1.8 We verified that there is an accumulation
of observations around height finished at five and zero
centimeters that are attributable largely to rounding.
Heaping is mainly observed at 150, 155, 160, 165 and
170 cm, but is very mild. The degree of heaping is more
pronounced in the first periods. Our data are normally
distributed or Gaussian and do not suffer from typical
truncation problems.
2.2. Income and other measures of welfare
To calculate real income, we use the annual series of
Spanish per capita GDP at constant 1995 prices in pesetas
4
Note that Spain is a cultural composite but that all towns in our
sample are in the Southeast where the Mediterranean culture was
predominant.
5
The dataset is composed of 141,861 records on the height of men. See
the doctoral thesis of Puche-Gil (2009), and Martı́nez-Carrión and PucheGil (2010).
6
See in Martı́nez-Carrión and Moreno-Lázaro (2007) the way height
series is standardized at the age of 21. After standardizing the series we
construct our final series as a five moving-average.
7
Our height series is constructed as a five moving-average. This is
frequently done in order to obtain trends rather than short-term
fluctuations. Many authors have provided height series by doing 3moving, 5-moving or even 10-moving year average. Some references
where it is done, just to name a few, are: Komlos (1994, 1995a, 1995b),
Komlos and Baten (1998), and Steckel and Floud (1997). Note also that in
our paper we are estimating a VECqM model and this analysis is
considering long-run relationships among variables. So, it is better to use
a series of this type.
8
Martı́nez-Carrión and Pérez-Castejón (1998) use a sample of the
height of 127,310 conscripts out of a total of 141,911 men (89.7%) called
up for service.
[(Fig._1)TD$IG]
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
Fig. 1. Distribution of heights (1837–1948) several periods by birth year. (a) Distribution of heights by year of birth, 1837–1865. (b) Distribution of heights by year of birth, 1866–1885. (c) Distribution of heights by
year of birth, 1886–1915. (d) Distribution of heights by year of birth, 1916–1994.
33
34
[(Fig._2)TD$IG]
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
provided by Prados de la Escosura (2003). His recent
economic growth estimates show that there have been
three main phases in Spain’s economic development:
1850–1950, 1951–1974, and 1975–2000, a decline occurring during the first phase as a consequence of the Civil
War of 1936–1939. Within this 150-year span, growth
rates fluctuated in reaction to economic policies, access to
international markets, and technological change can be
distinguished century’s average.
During the first phase, the most intense growth of the
period was achieved in the 1920s, coinciding with the
dictatorship of Primo de Rivera (1923–1929). This was a
period of institutional stability that provided a favourable
environment for investment and business. While the Great
Depression (1929–1933) had a moderate impact the Civil
War (1936–1939) had a negative impact on economic
growth, reducing the per capita GDP near 15% of the secular
growth rate (Comı́n, 2002; Prados de la Escosura, 2003).
During the early years of General Franco’s dictatorship – the
so called autarchy period-, the economic growth was weak
and its average growth rate was only 0.6% per annum,
whereas the average for the rest of Europe was 1.4% per
annum. The Spanish economy did not regain its pre-war
GDP levels (in absolute terms), until 1955 (in per capita
terms). As in other countries in the European Periphery, the
main spurt of economic growth in Spain was delayed until
the 1960s. This recovery ushered in an exceptional phase of
rapid growth, lasting until 1974, a period known as the
Golden Age. The years 1959–1974 emerge as a period of
outstanding performance (Prados de la Escosura, 2007).
When we compare the evolution of height and per capita
GDP for the period 1850–1978, we find no similarities until
the end of the 19th century (Fig. 2). Nevertheless, they seem
to be closely correlated during the period 1900–1920. We
obtain a correlation of 0.91 among the series for the whole
sample period.9 Examination of the average height and per
capita GDP by decade, reveals a lack of correlation between
income and height during the initial stages of modern
economic growth (Table 1), which is possibly related to
Kuznets’ inverted-U hypothesis, namely, that income
inequality rises and then falls with the level of economic
development. Researchers have recently observed a longterm rise in the inequality index during the early phase of
globalization that peaked with World War I (Prados de la
Escosura, 2008). Using Spanish-recruit data, Quiroga Valle
and Coll (2000) conclude that there was a long-term
increase in height inequality among socio-professional
groups between the turn of the century and World War I.
Moreover, Martı́nez-Carrión and Moreno-Lázaro (2007)
demonstrate that inequality trends between rural and
urban areas increased during the late-nineteenth century.
Anthropometric measures reveal the interrelations
among various components of human health and the
standard of living. Recent studies have shown that
improvements in housing and nutrition, comprehensive
9
If we split our sample period we observe how the correlation starts
increasing at the end of the 19th century. We also found a positive value
for the relationship between the height series and the detrended per
capita GDP.
Fig. 2. The relationship between height (birth year) in centimeters (left
vertical axis) and GDP per capita in pesetas 1995 (right vertical axis),
1850–1978.
health care, and health education have a positive effect on
height (Komlos and Lauderale, 2007). We use two healthrelated measures of consumer activity: the share of
consumption of health services in total consumption and
the ratio between the price level of consumer goods and the
GDP deflator.10 These series, which end at the end of 1958,
are provided by Prados de la Escosura (2003) at constant
1995 prices.11 Building on previous research into the
relationship between mortality and height (Fogel, 2004),
we include the mortality rate from 1858 to 1980 as an
additional independent variable (Nicolau, 2005, pp. 124–
126; Carreras and Tafunell, 2005).
We plot the relationships among stature, mortality rate,
the share of consumption of health services in total
consumption, and the ratio between the private consumption and the GDP deflators, respectively (Figs. 3–5). The
mortality variable is interpreted broadly as an indicator of
the population’s rate of fatal illness. We infer that a low
rate is associated with an increase in average adult stature
and observe that the relationship becomes especially
pronounced after the Civil War (Table 2). The effects of the
1918 flu epidemic and the Civil War, were immediately
reflected in the mortality series, and influenced the height
of the next generation of conscripts, measured during the
period 1933–1959. As for the share of health services in
total consumption, we observe a high value of the
correlation coefficient for this variable (0.87).
The relationships between stature and schooling
variables (as a proxy for human capital) such as literacy
rates and educational levels have also received the
attention of specialists (Meyer and Selmer, 1999; Schultz,
2003).12 Recently, Case and Paxson (2008) stressed that
10
We are grateful to one of the reviewers for suggesting that we use
these two variables.
11
Our consumption of health services series is taken from Prados de la
Escosura (2003, p. 457), Table A.7.1, column five.
12
Many authors used to employ schooling variables as a proxy for
human capital. However, these variables are inadequate as indicators of
human capital over long sample periods, their roles shifting, for instance,
from regime to regime.
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
35
Table 1
Average height and GDP per head (1850–1978).
1850–1859
1860–1869
1870–1879
1880–1889
1890–1899
1900–1909
1910–1919
1920–1929
1930–1939
1940–1949
1950–1959
1960–1969
1970–1978
Average height
(birth year in
centimeters)
Average GDPpc
(pesetas 1995)
Height variation
(centimeters)
Annual GDP pc
growth rate (%)
161.87
162.08
162.25
163.06
163.38
164.19
165.05
164.38
165.51
166.68
168.82
171.92
174.30
142.25
157.75
188.69
210.67
213.45
236.46
257.55
305.88
290.02
273.23
350.60
582.59
994.50
0.86
0.47
0.40
0.47
0.29
1.07
0.20
0.02
1.37
0.73
3.37
2.64
0.94
1.07
0.45
2.33
0.09
0.78
1.02
0.76
2.32
3.08
0.35
3.37
6.98
4.11
Sources: Prados de la Escosura (2003) and Martı́nez-Carrión and Puche-Gil (2009).
[(Fig._4)TD$IG]
height is positively associated with physical and mental
health and with cognitive ability; we have therefore
included some education variables in our study, such as the
schooling rate (Núñez, 2005). Historians have indicated
that in Spain advances in education and the standard of
living have lagged behind those of other European
countries and that at least until 1960, there was a wide
gap between the gains achieved by men and of women.
Although studies demonstrate that literacy spread and that
average height increased in the course of the first half of
the twentieth century, a height gap separated literate from
illiterate conscripts after the Civil War, the divergence
being due to differences in the degree of their access to
education, and reflected in the varying quality of their
signatures (Martı́nez-Carrión and Puche-Gil, 2009; Quiroga, 2001, 2003b, pp. 616–617).
This reevaluation indicates that most of the human
capital of the Spanish population in the second half of the
twentieth century was due to expanded primary schooling rather than to secondary or university studies.
Likewise, it identifies the Civil War as one of the most
serious
setbacks during two centuries of slow and
[(Fig._3)TD$IG]
Fig. 3. The relationship between height (birth year) in centimeters (left
vertical axis) and CDR crude mortality rate (right vertical axis).
Fig. 4. The relationship between height (birth year) in centimeters (left
vertical axis) and the share of consumption of health services in total
consumption (right vertical axis).
[(Fig._5)TD$IG]
Fig. 5. The relationship between height (birth year) in centimeters (left
vertical axis) and the ratio of private consumption deflator.
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
36
Table 2
Average height and standard of living indicators (1850–1978).
1850–1859
1860–1869
1870–1879
1880–1889
1890–1899
1900–1909
1910–1919
1920–1929
1930–1939
1940–1949
1950–1959
1960–1969
1970–1978
Average height
(birth year in
centimeters)
Mortality
rate (%)
Share of consumption
of health services/total
consumption
Ratio of deflator for
private consumption/
GDP deflator (1995 = 100)
Population
no schooling
rate (%)
Openness (exports +
imports/GDP)
161.87
162.08
162.25
163.06
163.38
164.19
165.05
164.38
165.51
166.68
168.82
171.92
174.30
28.4
30.3
30.9
31.6
30
25.6
23.4
19.7
17.2
13.5
9.6
8.5
8.3
0.43
0.48
0.40
0.65
0.74
1.00
0.73
1.02
1.91
3.12
2.96
–
–
1.35
1.39
1.46
1.56
1.50
1.44
1.40
1.47
1.33
1.77
1.70
5.1
1.62
61.12
56.38
51.14
51.50
53.31
57.23
49.97
47.70
54.62
33.20
33.23
25.12
17.17
0.66
0.90
1.29
2.34
3.02
3.29
4.21
3.42
3.73
1.76
4.02
10.10
26.23
Sources: Height, Table 1; mortality, Nicolau (2005); consumption of health services, deflator for private consumption, GDP deflator and consumption per
head, Prados de la Escosura (2003); population no schooling rates, Núñez (2005, pp. 232–236); openness (Tena, 2005, pp. 628–630).
irregular human-capital accumulation. The early years of
the Franco regime further depleted the stock of human
capital (Núñez, 2005). While we find that there is a
negative correlation between height and level of schooling at 0.77 it is insignificant.13
From our results, it can be observed that the underlying
institutional reasons for schooling in Spain may have been
completely different from the causes underlying increases
in average stature. This observation is based on the fact
that there is no reason to assume that the two variables
would develop in tandem. Thus, it is important to study
adult stature in Spain as distinct from schooling and other
conventional forms of human capital.
Finally, we have also included other important development indicators such as the degree of openness to
international trade, measured as the ratio of export plus
imports over GDP, Prados de la Escosura (2003, p. 188).
Some papers have described a negative association
between the barriers to trade and economic growth
(Ben-David, 1993; Sacks and Warner, 1995; Edwards,
1998; Frankel and Romer, 1999). The interrelationships
among health, the welfare state, and openness have yet to
be sufficiently treated from an anthropometric perspective. This subject is particularly pertinent to Spain because
of the pro-autarky policies, including economic liberalization, that were a feature of Franco’s long dictatorship
(1939–1975). Autarky as practiced in Nazi Germany and
formerly Communist European countries has been shown
to have negative effects on well-being, as a result of market
restrictions and controls such as fixed prices, production
contingents, import restrictions, food rationing, and
limitations on certain types of foods (Baten and Wagner,
2003; Laska-Mierzejewska and Olszewska, 2007). Analysis
of the relationship between human height and openness to
international trade must also include the rise of the welfare
state (Rodrik, 1997, 1999). Norway is an example of a
growing welfare state that was accompanied by a height
increase (Sunder, 2003). The annual series of the degree of
13
Our correlation results are different from Arora (2001), but this
variable was also statistically insignificant in this paper.
openness we used come from recent estimations made by
Tena (2005, pp. 628–31).
The relationship between height and openness is positive
(0.62), especially after 1959 (Fig. 6), Franco’s Stabilization
Plan having ushered in an era of unprecedented height
increases, greater than those that had occurred during two
earlier periods of openness, 1880–1889 and shortly before
the Civil War. The fact that the recruits of the 1940s (i.e., the
1920s cohorts) were on average shorter than previous
cohorts indicates that the autarky had a negative effect on
the stature of adolescents. Recent anthropometric studies
based on data from the Spanish National Health Survey
(2003–2006) that span several generations indicate that the
economic liberalization was accompanied by a height
increase among both men and women (Costa-Font and
Gil, 2008; Spijker et al., 2008).
3. Determinants of changes in Spaniards’ stature
[(Fig._6)TD$IG]
The average height of a given cohort is a function of (in
addition to genetic factors) nutritional intake during
Fig. 6. The relationship between height (birth year) in centimeters (left
vertical axis) and the grade of openness (right vertical axis).
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
infancy and childhood. Attained height reflects the tradeoff between the amount and quality of nutrients available
for growth from childhood to maturity and the demands
imposed by body maintenance, disease, and work. Thus,
the influence of income on height is indirect. Income
affects height because it is one of the determinants of food
consumption and health services consumption, and is also
related to child labour and the disease environment.
Thus, an increase in the average height of individuals
living under a given set of economic conditions could stem
from: an increase in per capita GDP, a decrease in food prices
relative to the GDP deflator, an increase in the share of
consumption of health services, or a decrease in the
mortality or infant-mortality rate. Other factors that
indirectly influence income could also influence height,
such as the degree of openness, and various human-capital
variables, such as the schooling rate, previously mentioned.
Following Jacobs and Tassenaar (2004), we model
human stature as a Vector Autoregressive Model (VAR).
We let Htt be the average height of conscripts at age t of the
cohort measured in year t, which is observed from t = 1, . . .,
T. The attained height at each age is by definition equal to
the increments in stature from the year of birth:
Htt DHtt þ DHtt1 þ þ DHt1 þ DHt0
(1)
where DHtt 1 Htt 1 Htt i1 is the increment in height of
the cohort of conscripts measured in year t between age
t i and t i 1, i = 1,. . ., t and DHt0 (or Ht0 ) is the length at
birth. We assume that the (unobserved) increments in
height depend linearly on income and other explanatory
variables, such as those described in Section 2.
DHtti ¼ ati Y ti þ eti
(2)
where eti is an error term and a is the coefficient
associated with the explanatory variables. Substituting in a
recursive way gives:
Htt ¼
t
X
t
X
i¼0
i¼0
ati Y ti þ
ati eti ¼ aðLÞY t þ et
(3)
where L is the lag operator and et is a moving-average error
expression.
Estimating Eq. (3) poses several problems. Since the
majority of explanatory variables are endogenous, estimating a single-equation for height by Ordinary Least
Squares (OLS) is not appropriate as regressors are
endogenous. We would have an endogeneity problem
because the error term and the regressors would be
correlated.
We therefore estimate a structural VAR model that
permits the existence of a cointegration relationship
among the variables in (3).
For that purpose, we use the Vector Error Correction
Equilibrium (VECqM) VAR model (4):
DY t ¼ c þ
p
X
Fi DY ti lY t1 þ et
(4)
i¼1
where Yt is a vector of endogenous variables, c is a vector of
constants, Fi is the matrix of autoregressive coefficients,
and et the vector of white-noise processes. Identification of
37
the structural shock is achieved by applying the Pesaran
and Shin (1998) technique to the variance-covariance
matrix of the reduced form residuals et.14
The orthogonalized and the generalized impulseresponse functions differ in a number of respects. The
generalized impulse responses are invariant to the
reordering of the variables in the VAR, but this is not
the case with the orthogonalized ones. There are many
alternative reparametrizations that could be employed to
compute orthogonalized-impulse responses, and there is
no indication of which parameterization should be used
(especially in the absence of any a priori theory). In
contrast, the generalized impulse responses are unique
and take into account the historical patterns of correlations
observed amongst the different shocks.
Our baseline VAR consists of a six-variable model. It
includes height in natural logarithms (ln height), mortality
rate (mort), the share of consumption of health services in
total consumption (consume), the ratio between deflator of
private consumption and GDP deflator (relativprice), real
per capita GDP in natural logarithms (ln gdp), and openness
(openness).15
3.1. Preliminary step: unit root, cointegration, and
regime-shift tests
A preliminary step, before estimating Eq. (4), is to
conduct an individual cointegration analysis. Three univariate unit-root tests were done. The results of the augmented
Dickey–Fuller (ADF) test and the Phillips–Perron (P–P)
(1988) test suggest that the majority of the variables are I(1),
and the Ng–Perron (2001) test, which corrects by distortions
size indicates that height continues to be an I(1) variable, but
that the rest of the series are I(2) (Table 3).16
Next, we carry out a Johansen (1990) cointegration test,
in order to examine the possibility of a long-run
cointegration relationship among the variables and to
estimate the cointegration vectors. According to the l-max
statistic, the null hypothesis of no cointegration versus one
cointegration vector and one cointegration vector versus
two was rejected at the 95% confidence level. Two
cointegration vectors were suggested by the unrestricted
cointegration rank test (Maximum Eigenvalue). We now
introduce the two normalized cointegrating coefficients
for the two cointegration equations, since all the variables
are I(1) processes using the Johansen cointegration test for
variables in levels. Table 5 offer the main results of
Johansen cointegration test for variables.
We find that height is positively related to the share of
consumption of health services in total consumption and
14
One of the main advantages of this technique is that it does not
require the orthogonalization of shocks and is therefore invariant to the
ordering of the variables in the VAR, permitting us to obtain generalized
impulse response functions exempt from criticism about the election of
the ordering of the variables in the VAR.
15
The percentage of the population without schooling was also
included as an explanatory variable, but it was insignificant during our
sample period.
16
We present the results of the test in Table 3. The values of the test
statistics are available upon request.
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
38
Table 3
Unit-root tests.
Height (logs)
GDP per capita (logs)
Mortality rate
Consume
Relativprice
Population without
schooling
Openness
Augmented
Dickey–Fuller
Phillips–
Perron
Perron-Ng
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
I(2)
I(2)
I(2)
I(2)
I(2)
I(1)
I(1)
I(2)
the degree of openness and negatively related to the ratio
between the consumer prices and GDP deflators and the
mortality rate. The second equation reveals that the per
capita GDP depends positively on the share of consumption of health services in total consumption and openness,
and negatively on the mortality rate, and the consumer
prices deflator relative to the GDP deflator. All the
explanatory variables were significant in both equations.
The results of the Ng–Perron (2001) test permit us to
apply the Johansen cointegration test for the height series
in levels and the rest of the variables in differences. We
again obtain two cointegration relationships.
In this case height is responding positively to changes in
the share of consumption of health services in total
consumption and the degree of openness, and negatively to
the changes in the ratio between consumer prices of and
the GDP deflator and mortality rates. We also observe that
the per capita GDP is responding positively to changes in
prices and openness and negatively to the mortality rate.
The changes in the share of consumption of health services
in total consumption are insignificant.
Finally, we use the Gregory and Hansen (1996) test to
determine whether a regime shift in the cointegration
relationship is possible. This is designed to test the null of
no cointegration against the alternative of cointegration in
the presence of the regime shift and thus to determine the
place of the shift in the time series. We have run the GAUSS
programs for this test using series at levels for all the
variables and for all the variables in differences, except for
height. We reject the null hypothesis of no cointegration,
and we find a regime shift around 1936 (Table 4).17
3.2. Vector autorregresive error correction model main
results
We estimate the VECqM specification, and we impose
the two long-run cointegration relationships derived there
by means of the Johansen (1990) test. Our baseline VAR
consists of a six-variable model with four lags. It includes
height in natural logarithms (ln height), mortality rate
(mort), the share of consumption of health services in total
consumption (consume), the ratio between the personal
consumption and the GDP deflators (relativprice), real per
capita GDP in natural logarithms (ln gdp), and openness
(openness). We also include one dummy variable for the
17
We are grateful to a reviewer for advising us to use this test.
Civil War period following the regime-shift test results. By
considering four lags in our estimation, we are in effect
including the influence of GDP on height during eight lags,
since our height variable is a five-year moving-average
variable. This implies that we are limiting our analysis to
height changes from adolescence through the age of 21,
and therefore are not capturing the childhood, stage of
growth. It should also be taken into account that there are
statistical difficulties in introducing a very large number of
lags in a VAR. We determine the two normalized derived
cointegration relationships by estimating the VECqM
(Table 6). The results are similar to those obtained by
means of the Johansen (1990) test.
As for the main results of the VAR estimated
coefficients, we center on the error correction results.
The error correction results are not amenable to generalization and indicate more nuanced interaction between the
variables. In particular, we find more instances in which
per capita GDP adjusts to the deviations from its
cointegration relationship with the rest of the economicdevelopment indicators and the cointegration relationship
among height and the same set of indicators. The
adjustment coefficient of height with respect to the
long-run cointegration relationships is negative and
significant for the first cointegration relationship
(0.013) and positive and significant (0.0062) for the
second. Changes in stature depend on changes in the share
of consumption of health services, and changes in the ratio
of consumer prices to the GDP deflator, income, the
mortality rate, and the degree of openness. Using these
variables, we can explain almost 69% (using the value of
the R-square) of the changes in stature during the period
1850–1958.18
With reference to the per capita GDP equation, we can
explain only about 56% of the changes in per capita GDP
during the sample period. Changes in per capita GDP
depend positively on the height variable, the degree of
openness, and the share of consumption of health services,
and negatively on the mortality rate and the ratio between
the consumer-price and the GDP deflators. The two
cointegration relationships are significant, with adjustment coefficients to the deviations from the cointegration
relationships of 0.94 and 0.77, respectively. This result
indicates that the size of the adjustment associated with
the per capita GDP is due to more than the height variable
and it lends support to our argument that the relationship
between per capita GDP and height is not unidirectional.
Although we concentrate our attention on the height
and income variable for the sake of brevity, we will also
comment the mortality equation results. We can explain
about 54% of the changes in mortality rate during the
sample period. Changes in mortality rate depend negatively on the height variable, per capita GDP, the share of
consumption of health services in total consumption and
18
We obtain a high value of the t-ratios for the constants in the long-run
cointegration regressions. This reflects, as we had expected, the elevated
weight of the omitted variables in the model. It is to be noted that we are
studying the relationship among height, per capita GDP, and some
economic-development indicators by means of VECqM models. It is
impossible to explain all the changes because crucial data are unavailable.
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
39
Table 4
Gregory and Hansen (1996)’s cointegration test results.
ADF test
Za-test
Zt-type
Level shift (C)
Level shift and trend (C/T)
Regime shift in trend and constant (C/S)
4.18 (0.363)
12.62a (0.7029)
124.78a (0.7623)
5.06 (0.256)
11.67a (0.7526)
111.63a (0.7623)
4.33 (0.7623)
12.62a (0.7623)
124.26a (0.7623)
Estimated breakpoint between parentheses.
a
Null hypothesis of no cointegration relationship is rejected at 5% level.
Table 5
Johansen cointegration test for variables.
Johansen cointegration test for variables in levels
height ¼ 0:127 consumeð0:025Þ 0:0056 relativ priceð0:008Þ 0:012 mortð0:004Þ þ 0:0092 o pennessð0:001Þ
gdppc ¼ 0:3891 consumeð0:001Þ 0:055 relativ priceð0:033Þ 0:786 mort ð0:03Þ þ 0:1239 o pennessð0:1126Þ
Johansen cointegration test: height in levels, the rest of variables in differences
height ¼ 0:1386 Dconsumeð0:041Þ 0:2638 Drelativ priceð0:117Þ 0:1125 Dmort ð0:014Þ þ 0:2039 Do pennessð0:06Þ
Dgdppc ¼ 0:060 Dconsumeð0:051Þ 0:298 Drelativ priceð0:12Þ 0:583 Dmortð0:152Þ þ 0:452 Do pennessð0:067Þ
Standard errors in parenthesis.
openness, and negatively on the ratio between the
consumer-price and the GDP deflators.
3.3. Generalized impulse-response functions
A convenient feature of this analysis is the possibility of
representing an accumulated impulse response of changes
in stature or other standard of living variables to the rest of
the variables without having to deal with the issue of the
ordering of the variables. This point is rather controversial
because Cholesky’s decomposition is traditionally used to
calculate the impulse response. To apply this methodology
one must first determine the ordering of the variables; this
poses a problem for us, because we have no a priori idea of
what the ordering should be.19 One possible solution is to
apply the Granger causality test to each of the variables but
when we attempted to do so, in a previous version of this
paper, the ordering that we obtained was unrealistic.20
For this reason we decided to calculate generalized
impulse-response functions instead of obtaining them by
means of Cholesky’s orthogonalization. This technique,
proposed by Pesaran and Shin (1998), does not require the
orthogonalization of the shocks and is invariant to the
ordering of the variables.
The accumulated generalized impulse response of
height to one-standard-deviation in shocks (unexpected
changes) in the independent variables is presented in
Fig. 7. These indicate that height responds negatively to a
positive increase in the mortality rate and to the ratio of
the consumer-price and the GDP deflators and positively to
an increase in GDP, the share of consumption of health
19
The Granger causality test provided information about precedence
changes, in which changes in one variable precede changes in the other,
but were unable to derive any realistic interpretations.
20
See Marı́a-Dolores and Martı́nez-Carrión (2009). In that version of the
paper we estimated a first version of the VAR, in which prices were before
per capita GDP (this implies that prices are flexible in the long run); then
we reversed the order, making prices after per capita GDP (this implies
that prices are rigid, and is unrealistic in the long run).
services in total consumption, and the degree of openness.
We conclude from the impulse-response analysis that the
height response to a positive shock in per capita GDP is
positive, but approximately eight years are required to
obtain this positive effect – an indication, perhaps, of the
importance of the standard of living during the adolescence period as a determinant of adult height.
For purposes of illustration, we proceed to interpret
numerically the results of the impulse-response functions.
A one-standard-deviation unexpected increase (a positive
shock) in per capita GDP – equal to 4.20% – will lead to a 1cm height increase (0.0052 in logs) after twenty years. A
one-standard-deviation positive shock to mortality (e.g. an
epidemic) – equal to 5.40% – will result in a decrease in
average stature of 0.019 (in logs) over the same sample
period. This implies that the response of height (in logs) to
a mortality shock is 3.65 times greater than the response to
a per capita GDP shock. A one-standard-deviation positive
shock to the share of consumption of health services in
total consumption – equal to 0.86% – will mean a 0.0055cm (in logs) height increase, and a one-standard-deviation
positive shock to the relative price of consumption – equal
to 5.7% – will lead to a 0.0052-cm (in logs) decrease in
average stature after twenty years. The response of height
to these two latter variables is similar to the response with
respect to per capita GDP.
Table 6
Normalized Long-run cointegration relationships among height and their
determinants by applying Johansen cointegration tests.
Long-run
relationship
Dependent
variable height
1850–1958
Dependent
variables
GDP per head
Consume
Relativprice
Mort
Openness
Constant
Trend
0.222* (2.43)
0.129* (1.89)
0.394* (5.91)
0.1009* (3.90)
5.08* (217.24)
0.00026* (4.21)
0.380* (4.78)
1.147* (5.01)
1.327* (5.93)
0.158* (1.83)
0.036* (26.74)
0.00067* (3.25)
t-Ratio between parentheses.
*
Significant 5% level.
[(Fig._7)TD$IG]
40
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
Fig. 7. Height accumulated generalized impulse-response functions.
The final issue to be considered is that of the
accumulated generalized impulse response of per capita
GDP to one-standard-deviation in shocks (Fig. 8). We
observe how per capita GDP responds positively to positive
height shocks, and the share of consumption of health
services, and openness, and negatively to an increase in the
ratio between the price and GDP deflators and the
mortality rate. The response of per capita GDP to height
changes is small and positive at the start and then
increases after the age of sixteen.
Next, we comment on the numerical results for the per
capita GDP impulse-response functions. A one-standarddeviation positive height shock – equal to 0.044% – will
increase per capita GDP in 1.19 pesetas after twenty years
and one-standard-deviation shock in the share of consumption of health services will increase per capita GDP by
1.46 pesetas (constant prices 1995). In contrast, a one-
standard-deviation shock to the relative price of consumption will decrease per capita GDP by 1.49 pesetas in twenty
years. The response of per capita GDP to the share of
consumption of health services and the relative price of
consumption is larger than its response to a per capita GDP
shock or a height shock.
Our VECqM results – and particularly those that
concern the impulse-response functions also reveal that
per capita GDP is not simply a function of height and that
height is not simply a function of per capita GDP, but rather
that there is a mutual dependency between the two.
3.4. Robustness analysis
By using the Ng–Perron (2001) test, we found that all the
standard of living indicators except the height variable could
be I(2) processes. To assess the extent to which our results
[(Fig._8)TD$IG]
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
41
Fig. 8. GDP per head accumulated generalized impulse-response functions.
are sensitive to these results, we re-estimate the VECqM
specification, deriving the two long-run cointegration
relationships by means of Johansen’s (1990) test (Section
3.1). Our baseline VAR consists of a six-variable model. It
includes height in natural logarithms (ln height), changes in
mortality rate (Dmort), changes in the share of consumption
of health services in total consumption (Dconsume), changes
in the ratio between the private consumption and the GDP
deflators (Drelativprice), changes in real per capita GDP in
natural logarithms (Dln gdp), and changes in the degree of
openness (Dopenness).
The long-term relationships derived from this procedure are of interest (Table 7). Height depends positively on
the increase in the share of consumption of health services
in total consumption and the degree of openness, and it
depends negatively on the changes in the ratio of prices to
the mortality rate. The adjustment coefficients of height
with respect to the two cointegration relationships are
0.013 and 0.00062, respectively. All of the accumulated
generalized impulse-response functions of height to the
different variables are presented in Fig. 9. Height responds
positively to unexpected changes in per capita GDP, the
Table 7
Long-run relationship among Height and their determinants.
Long-run
relationships
Dependent
variable height
Dependent
variable
Changes in
GDP per head
DConsume
DRelativprice
DMort
DOpenness
0.064* (2.49)
0.024* (2.09)
0.092* (4.83)
0.025* (3.79)
5.55* (465.25)
0.00118* (3.69)
0.403* (2.08)
1.112* (2.43)
2.076* (3.10)
0.460* (1.95)
0.030* (2.54)
0.000695* (2.07)
Constant
Trend
t-Ratio between parentheses.
*
Significant 5% level.
[(Fig._9)TD$IG]
42
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
Fig. 9. Height accumulated generalized impulse-response functions.
share of consumption of health services, and the degree of
openness, and negatively to unexpected changes in the
mortality rate and to changes in the ratio between
consumer prices and the GDP deflator.
By observing the behavior of changes in the per capita
GDP, we appreciate how it responds positively to changes
in height, in the share of consumption of health services in
total consumption, and in the ratio between prices, and
negatively to changes in the mortality rate and the degree
of openness.
Finally, we comment on the numerical results for the
height impulse-response functions. A one-standard-deviation positive shock in per capita GDP growth – equal to
4.66% – will lead to a 0.0038-cm (in logs) height increase
after twenty years. A one-standard-deviation positive
shock to the growth rate of share of consumption of
health services – equal to 0.99% will produce a 0.008-cm
(in logs) height increase. In contrast, a one-standard-
deviation positive shock in the growth of the relative price
of consumption – equal to 5.96% – will cause a 0.00525-cm
(in logs) height decrease. We observe how mortality
shocks are again producing greater effects on height.
4. Conclusions
This examination of the relationship between height
and economic development in Spain for the period 1850–
1958 reveals a long-running relationship among height,
income, and other indicators of economic development,
such as the mortality rate, the share of consumption of
health services in total consumption, the ratio between the
consumer-price and the GDP deflators, and the degree of
openness.
Having calculated, by means of a VAR-Vector Equilibrium Correction Model (VECqM), the impact on height of
changes in the explanatory variables, we could account for
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
almost 69% of the changes in stature during the period
1850–1958. The accumulated generalized impulse response of the changes in stature to these variables revealed
that height responds positively to an increase in per capita
GDP, an increase in the share of consumption of health
services in total consumption, and the degree of openness,
and negatively to the mortality rate and the ratio between
the consumer-price and GDP deflators.
To assess to what extent our results are sensitive to the
cointegration order of the variables, we re-estimated the
model, setting height at levels and the rest of the
economic-development indicators in first differences.
Our main results remain unchanged. We therefore
conclude that conditions that provoke an increase in
sickness and mortality rates and that raise the relative
price of consumption tend to impede cellular growth and
adult height, whereas factors responsible for an increase in
the proportion of health services to overall consumption
and for an opening up the economy to trade have tended to
benefit biological growth, which in turn is reflected in an
increase in the average height of Spain’s adult population.
Our VECqM-based results indicate that the dependency
between per capita GDP and height is mutual: that each
can be a function of the other. The impulse-response
function and our main VAR confirm that there is a
synergetic relationship between per capita GDP and height
(at least that of adult males). There is not a unidirectional
relationship. However, this mutual causality is difficult to
determine, because height and per capita GDP trends are
not always synchronized. We can state with confidence
that – setting aside the issue of per capita GDP levels –
height is an important economic-development indicator.
The next step is to apply other methodologies to the
relationship between height and a range of economicdevelopment indicators. For instance, Den Haan (2000) has
suggested a procedure for analyzing the co-movement
between output and prices, by determining the correlations of the corresponding VAR forecast errors at various
forecast horizons; such a procedure might resolve some of
the issues that we addressed in this paper but were unable
to answer.
References
Arcaleni, E., 2006. Secular trend and regional differences in the stature of
Italians, 1854–1980. Economics and Human Biology 4, 24–38.
Arora, S., 2001. Health, human productivity, and long-term economic
growth. Journal of Economic History 61, 699–749.
Baten, J., 2000. Heights and Real Wages in the 18th and 19th centuries: an
international overview. Jahrbuch für Wirtschaftsgeschichte 1, 17–32.
Baten, J., Wagner, A., 2003. Autarchy, market disintegration, and health:
the mortality and nutritional crisis in Nazi Germany, 1933–1937.
Economics and Human Biology 1, 1–28.
Baten, J., Morgan, S., van der Eng, P., 2010. Long Term Economic Growth
and the Standard of Living in Indonesia. Working Paper 514. The
Australian National University.
Ben-David, D., 1993. Equalizing exchange: trade liberalization and income convergence. Quarterly Journal of Economics 108 (3), 653–679.
Bosch, M., Bozzoli, C., Quintana-Climent, C., 2009. Infant Mortality, Income and Adult Stature in Spain, Documento de Trabajo 2009-27.
FEDEA-La Caixa, http://www.fedea.es/pub/papers/2009/dt200927.pdf.
Brinkman, H.J., Drukker, J.W., Slot, B., 1988. Height and income: a new
method for the estimation of historical national account series.
Exploration Economic History 25, 227–264.
43
Cámara-Hueso, A.D., 2006. Fuentes documentales antropométricas en
España: problemas metodológicos para los siglos XVIII y XIX. Historia
Agraria 38, 115–128.
Carreras, A., Tafunell, T. (Eds.), 2005. Estadı́sticas Históricas de España
Siglos XIX–XX. Fundación BBVA, Madrid.
Case, A., Paxson, C., 2008. Height, health, and cognitive function at older
ages. American Economic Review: Papers & Proceeding 98, 463–
467.
Coll, S., 1998. The relationship between human physical stature and GDP
(some experiments with European time series). In: Komlos, J., Baten,
J. (Eds.), The Biological Standard of Living in Comparative Perspective.
Franz Steiner Verlag, Stuttgart, pp. 384–407.
Comı́n, F., 2002. El perı́odo de entreguerras (1914–1936). In: Comin, F.,
Hernández, M., Llopis, E. (Eds.), Historia Económica de España. Siglos
X–XX. Crı́tica, Barcelona, pp. 285–330.
Costa-Font, J., Gil, J., 2008. Generational effects and gender height dimorphism in contemporary Spain. Economics and Human Biology 6,
1–18.
Craig, L.A., Weiss, T., 1998. Nutritional status and agricultural surpluses in
the Antebellum United States. In: Komlos, J., Baten, J. (Eds.), The
Biological Standard of Living in Comparative Perspective. Franz Steiner Verlag, Stuttgart, pp. 190–207.
Deaton, A., 2003. Health, inequality, and economic development. Journal
of Economic Literature 41, 113–158.
Deaton, A, 2007. Height, health, and development. Proceedings of the
National Academy of Sciences 104 (33), 13232–13237.
Deaton, A., 2008. Income, health, and well-being around the world:
evidence from the Gallup world poll. Journal of Economic Perspectives 22, 53–72.
Deaton, A, Arora, R., 2009. Life at the top: the benefits of height. Economics
& Human Biology 7 (July (2)), 133–136.
Den Haan, W.J., 2000. The comovement between output and prices.
Journal of Monetary Economics 46, 3–30.
Domı́nguez-Martı́n, R., 2002. La riqueza de las regiones. Las desigualdades
económicas regionales en España, 1700–2000. Alianza, Madrid.
Drukker, J.W., Van Meerten, M.A., 1995. Beyond Villerme and Quetelet:
the quantitative relation between sex- and age-specific height and
real per capita income. In: Komlos, J. (Ed.), The Biological Standard of
Living on Three Continents: Further Explorations in Anthropometric
History. Westview Press, Boulder, pp. 25–57.
Easterlin, R.A., 2000. The worldwide standard of living since 1800. Journal
of Economic Perspectives 14, 7–26.
Edwards, S., 1998. Openness, productivity and growth: what do we really
know? Economic Journal 108 (447), 383–398.
Federico, G., 2003. Heights, calories and welfare: a new perspective on
Italian industrialization, 1854–1913. Economics and Human Biology
1, 289–308.
Floud, R., Wachter, K., Gregory, A., 1990. Height, Health and History:
Nutritional Status in the United Kingdom. Cambridge University
Press, Cambridge, UK.
Fogel, R.W., 1994. Economic growth, population theory, and physiology:
the bearing of long-term processes on the making of economic policy.
American Economic Review 84 (3), 369–395.
Fogel, R.W., 2004. The Escape from Hunger and Premature Death, 1700–
2100. Cambridge University Press, New York.
Frankel, J., Romer, D., 1999. Does trade cause growth? American Economic Review 89 (3), 379–399.
Garcia, J., Quintana-Domeque, K., 2007. The evolution of adult height in
Europe: a brief note. Economics and Human Biology 5, 340–349.
Garcia, J., Quintana-Domeque, K., 2009. Income and body mass index in
Europe. Economics and Human Biology 7, 73–83.
German, L., Llopis, E., Maluquer, J., Zapata, S. (Eds.), 2001. Historia
Económica regional de España. Siglos XIX y XX. Crı́tica, Barcelona.
Gómez-Mendoza, A., Pérez-Moreda, A., 1995. Heights and Welfare in
Spain 1900–1930. In: Komlos, J. (Ed.), The Biological Standard of
Living on Three Continents. Further Explorations in Anthropometric
History. Westview Press, Boulder, pp. 76–89.
Gregory, A., Hansen, B., 1996. Residual-based tests for cointegration in
models with regime shifts. Journal of Econometrics 70, 99–126.
Haines, M.R., 2004. Growing incomes, shrinking people – can economic
development be hazardous to your health? Historical evidence for the
United States, England, and the Netherlands in the nineteenth
century. Social Science History 28 (2), 249–270.
Haines, M.R., 1998. Health, height, nutrition, and mortality: evidence on
the ‘Antebellum Puzzle’ from Union army recruits for New York State
and the United States. In: Komlos, J., Baten, J. (Eds.), The Biological
Standard of Living in Comparative Perspective. Franz Steiner Verlag,
Stuttgart, pp. 155–175.
Honda, G., 1997. Differential structure, different health: industrialization
in Japan, 1868–1940. In: Steckel, R., Floud, R. (Eds.), Health and
44
R. Marı´a-Dolores, J.M. Martı´nez-Carrión / Economics and Human Biology 9 (2011) 30–44
Welfare During Industrialization. Chicago University Press, Chicago,
pp. 251–284.
Jacobs, J., Tassenaar, V., 2004. Height, income, and nutrition in the Netherlands: the second half of the 19th century. Economics and Human
Biology 2 (2), 181–195.
Johansen, S., 1990. Estimation and hypothesis testing of cointegration
vectors in Gaussian vector autorregresive models. Econometrica 59,
1551–1580.
Kenny, C., 2005. Why are we worried about income? Nearly everything
that matters is converging. World Development 33, 1–19.
Komlos, 1985. Stature and nutrition in the Habsburg monarchy: the
standard of living and economic development. American Historical
Review 90, 1149–1161.
Komlos, J. (Ed.), 1994. Stature, Living Standard, and Economic Development. Essays in Anthropometric History. The University of Chicago
Press, Chicago.
Komlos, J., 1995a.In: The Biological Standard of Living in Europe and
America, 1700–1900: Studies in Anthropometric History. Variorum,
Brookfield, VT.
Komlos, J. (Ed.), 1995b. The Biological Standard of Living on Three
Continents. Further Explorations in Anthropometric History. Westview Press, Boulder.
Komlos, J., 1998. Shrinking in a growing economy: the mystery of physical
stature during the industrial revolution. Journal of Economic History
58 (3), 779–802.
Komlos, J., 2003. An anthropometric history of early-modern France.
European Review of Economic History 7 (2), 159–189.
Komlos, J., Baten, J. (Eds.), 1998. The Biological Standard of Living in
Comparative Perspective. Franz Steiner, Stuttgart.
Komlos, J., Lauderale, B.E., 2007. Underperformance in affluence: the
remarkable relative decline in U.S. heights in the second half of the
20th century. Social Sciences Quarterly 88, 283–305.
Laska-Mierzejewska, T., Olszewska, E., 2007. Anthropological assessment
of changes in living conditions of the rural population in Poland in the
period 1967–2001. Annals of Human Biology 34 (3), 362–376.
López-Casasnovas, G., Ribera, B., Currais, L., 2005. Health and Economic
Growth. Finding and policy implications. The MIT Press, Cambridge.
Margo, R., Steckel, R.H., 1983. Heights of native born whites during the
antebellum period. Journal of Economic History 43, 167–174.
Marı́a-Dolores, R., Martı́nez-Carrión, J.M., 2009. The relationship between
height and economic development in Spain. A historical perspective.
Documentos de Trabajo (DT-AEHE) 0912. Asociación Española de
Historia Económica.
Martı́nez-Belmonte, L., 1983. La talla media de los españoles varones
crece cada año. Estadı́stica Española 101, 121–125.
Martı́nez-Carrión, J.M., 1994. Niveles de vida y desarrollo económico en la
España contemporánea: Una visión antropométrica. Revista de Historia Económica – Journal of Iberian and Latin American Economic
History 12 (3), 685–716.
Martı́nez-Carrión, J.M., 2002. El nivel de vida en la España rural siglos
XVIII–XX. Publicaciones de la Universidad de Alicante, Alicante.
Martı́nez-Carrión, J.M., 2009. Anthropometric history and Ibero-American
historiography. Historia Agraria 47, 11–18.
Martı́nez-Carrión, J.M., Moreno-Lázaro, J., 2007. Was there an urban
height penalty in Spain, 1840–1913? Economics and Human Biology
5, 144–164.
Martı́nez-Carrión, J.M., Pérez-Castejón, J.J., 1998. Height and standard of
living in Spain, 1860–1969: evidence from the southeastern region.
In: Komlos, J.J., Baten, (Eds.), The Biological Standard of Living in
Comparative Perspective. Franz Steiner, Stuttgart, pp. 344–358.
Martı́nez-Carrión, J.M., Pérez-Castejón, J.J., 2000. On the height of Spanish
recruits during the early phases of modern economic growth. Jahrbuch für Wirtschaftsgeschichte 11 (1), 95–112.
Martı́nez-Carrión, J.M., Puche-Gil, J., 2009. Literacy, biological welfare,
and inequality: Region of Valencia (Spain), 1850–1970. Historia
Agraria 47, 167–186.
Martı́nez-Carrión, J.M., Puche-Gil, J., 2010. La evolución de la estatura en
España: nuevos resultados y conjeturas. , http://www.uc3m.es/portal/
page/portal/dpto_historia_economica_inst/workshop.
Meyer, H.E., Selmer, R., 1999. Income, educational level and body height.
Annals of Human Biology 26 (3), 219–227.
Moradi, A., 2010. Nutritional status and economic development in subSaharan Africa, 1950–1980. Economics & Human Biology 8 (1), 16–29.
Ng, S., Perron, P., 2001. Lag length selection and the construction of unit
root test with good size and power. Econometrica 69, 1519–1554.
Nicolau, R. 2005. Población, salud y actividad. In: A. Carreras, X. Tafunell
(Coords.), Estadı́sticas históricas de España, siglos XIX–XX. Fundación
BBVA, Bilbao, pp. 77–154.
Núñez, C.E., 2005. La educación. In: Carreras, A., Tafunell, X. (Eds.), Estadı́sticas Históricas de España. Fundación BBVA, Bilbao, pp. 157–244.
Peracchi, F., 2008. Height and economic development in Italy, 1730–1980.
American Economic Review 98 (1), 475–481.
Persico, N., Postlewaite, A., Silverman, D., 2004. The effect of adolescent
experience on labour market outcomes: the case of height. Journal of
Political Economy 2004 (112), 1019–1053.
Pesaran, H., Shin, Y., 1998. Generalized impulse response analysis in linear
multivariate models. Economic Letters 58, 17–29.
Phillips, P., Perron, P., 1988. Testing for a unit root in time series regression. Biometrika 75, 335–346.
Prados de la Escosura, L, 2003. El Progreso Económico de España (1850–
2000). Fundación BBVA, Madrid.
Prados de la Escosura, L., 2007. Growth and structural change in Spain,
1850–2000: a European perspective. Revista de Historia Económica
(Journal of Iberian and Latin American Economic History).25 (1), 147–
181.
Prados de la Escosura, L., 2008. Inequality, poverty and the Kuznets curve
in Spain, 1850–2000. European Review of Economic History 12 (3),
287–324.
Puche-Gil, J., 2009. La evolución del los niveles de vida biológicos en la
Comunidad Valenciana, 1840–1948. Ph.D. Diss. Universidad Pablo
Olavide de Sevilla.
Quiroga Valle, G., Coll, S., 2000. Income distribution in the mirror of height
differences: the case of Spain 1895–1950. Journal of Income Distribution 9, 107–131.
Quiroga, G., 2001. Estatura, diferencias regionales y sociales y niveles de
vida en España (1893–1954). Revista de Historia Económica XIX (Journal of Iberian and Latin American Economic History).19, 175–200.
Quiroga, G., 2003a. Medidas antropométricas y condiciones de vida en la
España del siglo XX. PhD Diss. Universidad Alcalá de Henares.
Quiroga, G., 2003b. Literacy, education and welfare in Spain. Paedagogica
Historica 39 (5), 599–619.
Rebato, E., 1998. The Studies on secular trend in Spain: a review. In:
Bodzsár, É.B., Susanne, Charles, (Eds.), Secular Growth Changes in
Europe. Eötvös University Press, Budapest, pp. 297–315.
Rodrik, D., 1997. Trade, Social Insurance, and the Limits to Globalization,
NBER Working Paper 5905. .
Rodrik, D., 1999. Making Openness Work: The New Global Economy and
the Developing Countries. Overseas Development Council, Washington, DC.
Sacks, J., Warner, D., 1995. Economic reform and the process of global
integration. Brooking Papers on Economic Activity 1995 (1), 1–118.
Sandberg, L.G., Steckel, R.H., 1997. Is industrialization hazardous to your
health? Not in Sweden. In: Steckel, R.H., Floud, R. (Eds.), Health and
Welfare During Industrialization. Chicago University Press, Chicago,
pp. 127–160.
Schultz, T.P., 2003. Human capital, schooling and health. Economics and
Human Biology 1, 207–221.
Shay, T., 1994. The level of living in Japan, 1885–1938: new evidence. In:
Komlos (Eds.), Stature, Living Standards, and Economic Development:
Essays in Anthropometric History. The University of Chicago Press,
Chicago, pp. 173–201.
Shutkowski, H., 2008. Between Biology and Culture. Cambridge University Press, Cambridge.
Spijker, J., Pérez, J., Cámara, A., 2008. Cambios generacionales de la
estatura en la España del siglo XX a partir de la Encuesta Nacional
de Salud. Revista de Estadı́stica Española 50, 571–604.
Steckel, R.H., 1979. Slave height profiles from coastwise manifests.
Explorations Economics History 16, 363–380.
Steckel, R.H., 1983. Height and per capita income. Historical Methods 16,
1–7.
Steckel, R.H., 1995. Stature and standard of living. Journal of Economic
Literature 33, 1903–1940.
Steckel, R.H., 2008. Biological measures of the standard of living. Journal
of Economic Perspectives 22 (1), 129–152.
Steckel, R.H., 2009. Heights and human welfare: recent developments and
new directions. Explorations in Economic History 46, 1–23.
Steckel, R.H., Floud, R. (Eds.), 1997. Health and Welfare During Industrialization. University of Chicago Press, Chicago.
Sunder, M., 2003. The making of giants in a welfare state: the Norwegian
experience in the 20th century. Economics and Human Biology 1,
267–276.
Tena, A. 2005. El sector exterior. In: A. Carreras, X. Tafunell (Coords.),
Estadı́sticas Históricas de España. FBBVA, Barcelona, pp. 573–644.
Vecchi, G., Coppola, M., 2006. Nutrition and growth in Italy, 1861–1911:
what macroeconomic data hide. Explorations in Economic History 43
(3), 438–464.
Weir, D.R., 1997. Economic welfare and physical well-being in France,
1750–1990. Health and welfare during industrialization. In: Steckel,
R.H., Floud, R. (Eds.), Health and Welfare During Industrialization
Chicago. University of Chicago Press, pp. 161–200.