The biological standard of living (and its convergence) in Colombia, 1870-2003. A tropical success...

The biological standard of living (and its

convergence) in Colombia, 1870–2003

A tropical success story

Adolfo Meisel *, Margarita Vega

Banco de la Republica (Central Bank of Colombia), Centro de Estudios Economicos Regionales CEER,

Centro Calle 33 #3-123, Cartagena, Colombia

Received 10 October 2006; accepted 11 October 2006

Abstract

During the 20th century, the evolution of the biological standard of living in Colombia was a tropical

success story from the point of view of the secular increase in height as well as the reduction of inequality.

During the period 1905–1985 the average height of females and males increased by nearly 9 cm on the basis

of 9 million records examined from National Identification Cards. We also study the evolution of height of

Colombians on the basis of passport records. The elite group of passport holders was much taller than

average, and remained stable for the birth cohorts of 1870–1919. In the early 20th century the height of

passport recipients was 168.7 cm (men) and 158 cm (women) compared with 162 cm and 150 cm for

heights in the national ID cards. The results also show that Colombians experienced significant regional and

intrapersonal convergence in height.

# 2006 Elsevier B.V. All rights reserved.

JEL classification : I12; I31; N16; N36

Keywords: Anthropometric history; Welfare; Health; Biological standard of living; Colombia; Height; Physical stature;

Convergence; Inequality; Latin America

1. Introduction

Since the late 1970s there has been a growing interest among economic historians in the study

of the development of height over time, as an alternate perspective on the standard of living

http://www.elsevier.com/locate/ehb

Economics and Human Biology 5 (2007) 100–122

* Corresponding author. Tel.: +57 5 6600761; fax: +57 5 6645885.

E-mail addresses: [email protected] (A. Meisel), [email protected] (M. Vega).

1570-677X/$ – see front matter # 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.ehb.2006.10.004

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.ehb.2006.10.004

(Komlos, 1989).1 Adult height reflects the net nutritional status of a person during the years of

physical growth, which is influenced by food intake, health, and work effort (Steckel, 1995;

Tanner, 1994).2 Thus, it provides a measure of the biological standard of living, a dimension of

the overall standard of living.

For Latin American populations, the studies of height from recruitment files show for Mexico

a stagnation from the end of the 19th century until the 1930s or 1940s. It is similar to the results

found for the Mexican passports records, although the latter group was 6.0 cm taller (Lopez-

Alonso and Porras, 2003). From 1880 until 1910 the average height of men was 170 cm for the

passports records and 164 cm for military data. In the case of Argentina, army data show that

between 1900 and 1914 height was stagnant, although this was a period of rapid economic

growth. After 1914, when the ‘‘Argentina golden age’’ had come to an end, the height of recruits

increased by about 2.1 cm, from 165.1 cm to 167.2 cm (Salvatore, 2004a,b).

This paper studies the evolution of the height of Colombians born between 1870 and 1985. For

those born in 1985 height was recorded in 2003, when they received their citizenship card (cedula

de ciudadanıa) at age 18.3 With 9,321,776 observations, this is the largest database in the world

on anthropometric history to date.

In economic terms Colombia was a success story during the 20th century. The rate of growth

of per capita GDP from 1905 to 2000 was 2.3% per annum, one of the highest in Latin America

(GRECO, 2002). This economic success was reflected in several dimensions of the standard of

living, one of which is the height of its population. Colombian men and women born in 1985 were

8.9 cm taller than those born in 1905–1909, an enormous improvement in physical well being

achieved in only three generations.

The paper is divided into five sections. First, the trend in the height for a group of Colombians

born in the period 1870–1919 is presented. The information was obtained from passports, thus the

sample has a clear bias due to the social composition of the group. Next, a database constructed with

information from the national citizenship card for Colombians born in the period 1905–1985 is

discussed. Then, the evolution in their height is analyzed. Section 4 compares the behavior in the

average height by departments, the main sub-national territorial units, and a convergence analysis is

performed for the period as a whole. The last section concludes.

2. Stature of the Colombian elite4, born 1870–1919

For the pre-industrial era in Colombia, that is to say for the 19th century, information on height

is available only from 1870. We have obtained about 16,000 observations from the records of

A. Meisel, M. Vega / Economics and Human Biology 5 (2007) 100–122 101

1 The research of Robert Fogel and his collaborators in the 1970s contributed to the increasing interest in anthro-

pometrics on the part of economic historians.2 The first three years of life are the very important for a person’s final height, although growth continues until around

18 years of age.3 Registradurıa Nacional del Estado Civil (RNEC). After 1975 the citizenship card began to be issued at that age of 18.

Before that it was issued at the age of 21. Since the citizenship card is needed for almost all commercial and legal

transactions the overall of population obtain it one or two years after age 18. However there is not upper age when a

person can receive this document. The database was obtained by the Central Bank of Colombia (Banco de la Republica)

through a formal request. The Registradurıa Nacional del Estado Civil (RNEC) provided a database with 9,321,776

observations. For more information see: http://www.registraduria.gov.co/, Address: Av. El Dorado # 46-20, Bogota,

Colombia.4 The Colombian elite was basically comprised of rich landowners, merchants, high officials, and professionals of

European ancestry.

http://www.archivogeneral.gov.co/

passports issued to Colombian citizens both men and women.5 This group belonged mostly to the

elite born between 1870 and 1919.6 The database was constructed by the authors from the records

of passports issued to citizens by the Colombian Ministry of Foreign Relations.7 This information

has not been used before by social scientists, although it is in excellent condition. There are no

other height records before 1870. This paper uses passports records only until 1919 and compares

them to heights in the national ID cards during a 10-year period (1910–1919). The passports

A. Meisel, M. Vega / Economics and Human Biology 5 (2007) 100–122102

Table 1

Average height (cm) of passport recipients born in 1870–1919

Date of birth Average Increase

(%)

Number of

observations

Standard

deviation

Coefficient of

variation

Increase

(cm)

Men

1870–1874 167.4 195 8.44 0.0497

1975–1879 168.2 0.5 393 7.01 0.0415 0.79

1880–1884 168.7 0.3 594 6.90 0.0407 0.51

1885–1889 168.8 0.1 881 7.23 0.0428 0.18

1890–1894 168.5 �0.2 1229 7.29 0.0432 �0.31

1895–1899 168.5 0.0 1591 7.13 0.0423 �0.01

1900–1904 168.3 �0.1 1739 6.71 0.0398 �0.24

1905–1909 168.7 0.3 1871 6.77 0.0401 0.44

1910–1914 168.3 �0.3 1246 7.14 0.0424 �0.43

1915–1919 168.6 0.2 651 6.89 0.0407 0.33

Total 168.39 0.7 10390 1.24

Women

1870–1874 158.1 120 6.20 0.0393

1975–1879 157.5 �0.4 197 6.21 0.0395 �0.62

1880–1884 158.2 0.5 331 7.64 0.0483 0.73

1885–1889 158.9 0.4 390 7.72 0.0486 0.68

1890–1894 158.7 �0.1 512 7.56 0.0477 �0.22

1895–1899 158.3 �0.3 688 7.72 0.0488 �0.42

1900–1904 158.2 0.0 887 7.46 0.0472 �0.06

1905–1909 158.1 �0.1 1006 7.01 0.0444 �0.09

1910–1914 158.6 0.3 835 6.86 0.0433 0.46

1915–1919 158.7 0.1 555 6.58 0.0415 0.14

Total 158.33 0.4 5521 0.59

Source: Archivo General de la Nacion and archives of the Foreign Ministry (hereafter: AGN_FM). Note: Includes only

persons whose age was between 18 and 60 years.

5 Colombia began to issue passports in 1824 when the recently created Republic passed a law authorizing the granting

of passports to those Colombians who found themselves abroad (Salamanca, 1959).6 These passports were issued from 1918 to 1940. A growing number of observations per year were obtained until 1906,

when 420 records for men and 219 for women were found. Then, the sample drops to 83 observations per year for both

men and women.7 Although there is some information for passports issued in the period 1859–1869 the number of observations is so

limited that the results are not significant. For this reason we have used only the data that begins in 1870. For the period

before 1859 no information on passports is available at the archives of the Ministry of Foreign Relations and Archivo

General de la Nacion. This latter archive contains the public records for all government institutions except for the army

which maintains separate archives. Passport data are available after 1919 until the present, and could be the subject of

future research. The authors did not obtain access to the army archives. Therefore, the military data still remains to be

explored. For more information see: http://www.archivogeneral.gov.co/, Address: K. 6a. # 6-91, Bogota, Colombia.

http://www.archivogeneral.gov.co/

included information on name, city where it was issued, date of issue, age, height, color of eyes

and hair, destination, reasons for trip abroad, and occupation. It also included a photograph. The

place of birth was not reported. All the above information is available in the database we have

constructed, except the eye and hair color, which we did not consider useful for our purposes.

The quality of these data was tested by applying the Lilliefors and Jarque-Bera normality tests

to the observations of each year. In both tests, and for almost every year, the null hypothesis of

normality was not rejected at the 3% level of significance.8

The increase in height of men during the whole period considered was 1.24 cm—only about

0.25 cm per decade (Table 1). Among women the rate of increase was even lower, 0.12 cm per

decade. Furthermore, the trend for the heights of men and women in the period 1870–1919 is not

statistically significant at 5% level of confidence, as the t-statistics on the time trends were 1.6

and 1.8, respectively (Appendix A, Tables A.1–A.4).9

The dispersion of the heights of Colombian passports holders as measured by the coefficient

of variation (CV) of height is an important indicator of interpersonal inequality of height. The CV

of height was relatively constant, fluctuating around 0.04, for both men and women (Fig. 1). The

trend of the coefficient of variation is not statistically significant.10

That the sample obtained from passports is not representative of the Colombian population is

evident when comparing the average height obtained from passports and ID cards. The height of

women obtained from passports was 7.8 cm above the height obtained from ID cards. In the case

of men the difference is smaller but still large: 4.8–5.0 cm.


Fig. 1. Coefficient of variation of height of Colombian passport recipients born 1870–1919.

8 The t-statistics for men under the Lilliefors test was between 0.05 and 0.14.9 Since there could be possible changes in sample composition in the period 1870–1919, for example more people

started to travel, we performed a correction for this possible bias using a weighted and restricted least-squares regression.

Using this method we estimated average height series for men and women, creating dummies for year, gender, and place

of issue. Since the sample has some biases, for example sample composition and rounded data, we complete the

estimation using weighted least-squares regression. The results show that average height (crude average) does not vary

from the average estimated as a mean of height. Thus, average height for men and women can be used for this analysis.10 The t-statistic for the trend of the coefficient of variation is �1.09 for men and 0.53 for women, both are no

statistically significant at the 5% of confidence, correcting by changes in sample composition (Appendix A).

Another important difference between the heights observed in the passport records and those

in the ID cards is that in the former the long term trend is constant, while in the latter there is a

clear positive trend in height (Fig. 2). It took a long time for the rest of the population to achieve

the heights reached by the elites already in the early 20th century. The height of passport

recipients born 1919 was reached by average Colombian male ID card holders only in 1960–

1964, and in the case of women only around 1980–1984.

3. National citizenship files

The first systematic treatment of the long run behavior of average height in a Latin American

country was published in 1992 (Ordonez and Polania, 1992; Lopez-Alonso and Porras, 2003).11

Ordonez and Polania (1992) used a sample of 14,103 observations for the period 1900–1972,

obtained from the national citizenship card files (5,839 women and 8,264 men). That study

concluded that there was a positive secular trend in height both for men and women. There was a

gain of 7.0 cm form men and 8.7 cm for women. However, Ordonez and Polania did not relate the

evolution of height with other social and economic variables, limiting themselves to a descriptive

presentation of the data. The height data used in our study come from the citizenship cards

(cedula de ciudadanıa), the same source that had been used by Ordonez and Polania (1992).

However, unlike them we do not use a sample but the total of the database available from the

electoral authority, the Registraduria Nacional del Estado Civil (RNEC).

The existence of continuous information on the height of Colombians during the 20th century

is not accidental. The citizenship card originated in the need to have a document to identify

voters. Colombia has one of the longest traditions of democratic elections in Latin America and

has never had long periods of authoritarian government.12 The first law to establish a voter


Fig. 2. Height of Colombian men and women by Source, 1870–1985.

11 For Mexico there is an anthropometric study for the period 1870–1950 which uses data from military recruitment and

passport records, Lopez-Alonso and Porras (2003). Those authors found that Mexicans born in 1950 were not taller than

those born in 1870.12 Over the last few years the electoral history of Colombia since independence has attracted the attention of several

historians: Bushnell (1971, 1984, 1994), Deas (2002) and Posada (1994, 1997).

identification document, although very deficient, was issued in 1853 (RNEC, 1988, p. 35). Only

in 1934 was a citizenship card (cedula de ciudadanıa) established for men over 21 years of age.

The information included in that document recorded such physical characteristics as the color of

skin, hair, and eyes, as well as height. Due to numerous problems with the quality of the records

of the initial citizenship document, in 1952 a new one was issued. That new document included a

picture of the citizen, his fingerprint, height, skin color, date of birth, and place of birth.13 For

women the citizenship document was first issued in 1956.14 In 1954 women had been granted the

right to vote, although it was only in December of 1957 that they participated in an election

(RNEC, 1988, p. 51). The database used in the present study corresponds to the records of this

second citizenship card, which was first issued in 1952.

The percentage of those born in specific year captured by the ID documents increases from

3.7% in 1938 to 94.9% in 1980 (Table 2). Thus, the results obtained in this paper are highly

representative of the overall Colombian population (Komlos, 2004).15 An important aspect of

this database is that it also includes information for women, which is normally absent from

military records. Additionally, it enables us to analyse heights at the regional level (and thus of

ethnic characteristics) and even for the main cities and towns. Finally, it extends for almost a

century of Colombian history, since it begins with those born in 1905 and extends to those born in

1985, and who reached their adult height around 2003.

The quality of these data was further corroborated with a non-parametric Kolmogorov–

Smirnov test, which compares the distribution between two samples. The results show that the

average Colombian height had a significant increase during the whole period.16


Table 2

Number of observations on the height of Colombians (selected years)

Census year Total population

of Colombia

Number of

births

Number of

observations on height

Observations as a percentage

of those born in that year (%)

1938 9,072,894 388,592 14,836 3.82

1951 12,411,101 547,330 27,939 5.10

1964 17,484,510 795,021 66,808 8.40

1973 20,666,920 848,790 158,893 18.72

1980 24,225,517 790,721 750,225 94.88

1984 27,136,688 885,742 684,578 77.29

1985 29,265,499 766,248 393,537 48.88

Note: Number of births and total population for 1980 and 1984 have been estimated by the authors. Source: Registraduria

Nacional del Estado Civil (hereafter RNEC).

13 Through the years this last document has had a few minor changes concerning the information it includes and its

presentation. In 1975, when Colombians over 18 years were granted the right to vote, the citizenship card began to be

issued at that age. Since some people grow in height past their 18th birthday, especially in the early stages of development,

this could introduce a structural break in the height series. For this reason a simple test proposed by Ben-David and Papell

(1997) was performed, but no structural break in the series was found, either for men or women.14 Records of women’s heights begin before 1956, due to the fact that women who were born before that date were also

given a citizenship card.15 Height studies often run into difficulties created by deficient samples. For example, as a result of minimum height

requirements such as those used in the military, truncation of the data is very common. Also, some samples may present

regional or socio-economic status biases, which lead to results that may not be representative of the total population.16 The test was performed for the period 1905–1985, every five years. In each case at the 5% level of significance the

alternative hypothesis was accepted (Null hypothesis = D no significant; Alternative hypothesis = D significant at the 5%

of significance).

4. Growing taller: continuous increases in height

The average height of Colombians increased throughout the 20th century in every decade, and

for both sexes, regions, and social classes. This growth was neither hampered by the cycles of

political and criminal violence, nor by those of concentration of personal income that occurred in

several periods of that century.

On average, male Colombians born between 1905 and 1909 grew to an adult height of

162.05 cm. In contrast, those born in 1985 achieved an average height of 171.01 cm, an increase

of 8.96 cm, which corresponds to a 5.5% growth in stature. In the same period women increased

their height from 150.02 cm to 158.97 cm. That is an increase of 8.95 cm or 5.9% above the

initial height. In both cases, the data show an increase of about 1 cm per decade, a remarkable

achievement by international standards (Honda, 1997).17

What are the main reasons behind the almost continuous increases in the height of

Colombians? The results of research in the field of anthropometric history identify at least four

main factors for the rise of average height in a population over time: reduction in work stress,

advances in overall health conditions, changes in the ethnic composition, and improvements in

nutrition (Komlos, 1989, Chapter 1). Among these possible influences on stature only the

changes in ethnic composition do not seem to play a role in the Colombian case, since during the

20th century the country experienced a negligible amount of foreign immigration.

At the beginning of the 20th century Colombia was predominantly an agricultural country

with most of its population working in physically demanding jobs in that sector. A majority of the

children were also involved in the labor force. For example, in a population census carried out in

the late 19th century most of the children who were older than eight years were part of the labor

force (Munoz, 1996). Additionally, with economic development there was an increase in the

number of jobs in the industrial and service sectors, where jobs were often less physically

demanding. Also, there was a significant reduction in child labor which was aided by the

expansion of education and by legislation directed at prohibiting work deemed detrimental to the

health of children (Munoz, 1996, p. 100).

Throughout the 20th century there was a continuous reduction in the mortality rate in

Colombia (Florez, 2000). One of the main reasons for the decline was the improvement in overall

health conditions as a result of advances in personal hygiene, public sanitation and medical

science. Especially important was the control or elimination of many tropical diseases. For

example, the last reported urban epidemic of yellow fever occurred in 1929 in Socorro, Santander

(Abel, 1996). Public campaigns to eradicate this disease, which were undertaken since the early

1920s, with the aid of the Rockefeller Foundation, were responsible for this result. Additionally,

since 1937 vaccination against yellow fever was undertaken by the government. These

improvements in disease prevention must have contributed to the observed gains in the average

height of Colombians.

The diet of Colombian laborers at the beginning of the 20th was completely inadequate: it was

deficient in calorie and protein intake, and lacked essential nutrients. In 1893, Manuel Cotes, a

physician, studied in great detail total food consumption by rural laborers around Bogota, where

the best agricultural lands in Colombia are located (Cotes, 1893). The daily diet was composed of


17 Between the end of the nineteenth century and the beginning of the World War II, Japan had one of the fastest growing

economies in the world. From 1892 to 1938 the average height among recruits increased from 156.1 cm to 160.3 cm, an

average of 0.91 cm per decade.

3500 g of chicha (a fermented maize drink), 600 g of mazamorra (maize porridge), 360 g of

bread, and 40 g of chocolate.

In the 1930s the Colombian government undertook several studies on the living conditions of

urban workers in the main cities of the country. According to these studies most of the income of

the working class families was spent on food. In Medellın these expenditures represented 63.5%

of the total budget, while in Bogota it was 65.6% (Abrisqueta, 1940, p. 44). The results indicated

that the main reason for the generalized malnutrition of the working class was its low income

(Abrisqueta, 1940, p. 50).

A study of the nutritional status of Colombian children by FAO found that as recently as 1965

around 31.9% were so malnourished that their achieved height was below their potential.

However, there was a continuous improvement in nutrition that led to a decrease in the number of

children below their height potential, to 22.4% in 1977, 16.6% in 1986, 15.0% in 1995, and

13.5% in 2000 (FAO, 2001).

The improvements in the amount of food consumed by the Colombian population in the 20th

century were, to a large extent, a result of the increase in GDP per capita, which expanded at an

average annual rate of 2.3% in the period 1905–2000 (GRECO, 2002). Additionally,

technological advances reduced the real price of key components of the diet. For example,

beginning in the 1970s the poultry sector was industrialized and as a result the real price of

chicken and eggs fell drastically, leading to an enormous increase in their consumption. While in

1950 Colombians ate an average of 2.7 kg of chicken per capita annually, by 1998 it had

increased to 15.3 kg (Galvis, 2000).

Men were taller than women by at least 11 cm. However, that differential went through

distinct cycles (Fig. 3). From 1905 to circa 1960 the height of women converged towards that of

men and the difference between them fell from more than 13 cm to 11 cm. This convergence

could have been the result of the overall improvement of the status of women in Colombian

society.

On the other hand, the divergence in the average height of men and women after about 1960 is

puzzling since the improvement in the relative situation of women has continued after that date,

as evidenced by their gains in education and access to the labor market. Perhaps urbanization has

favorably affected the relative work effort of men in comparison to what was prevalent when

Colombia was a more rural society, and thus demanded more physical labor.


Fig. 3. Difference between the average height of Colombian men and women by the date of birth.

In contrast, the difference between the height of men and women in the passport sample born

in the period 1870–1919 remained almost unchanged, with men’s height exceeding that of

women by about 10 cm. This is a smaller difference than was observed for the height between

men and women using the national ID card information. This indicates that there was greater

equality between genders in the nutrition status among the elite.

The upward trend in the height of Colombians throughout the 20th century was only

interrupted during two short periods, 1957–1960 and 1970–1975, when average height actually

decreased (Table 3). Though relatively small, the setback during the 1970–1974 quinquennium is

somewhat puzzling since during those years per capita income and agricultural output both

increased. For example, in the five years from 1970 to 1974 the average growth rate of

agricultural GDP was 5.4%. Additionally, during the 1970s personal income distribution was

improving. Estimates of the Gini coefficient show that it declined from 0.53 in 1971 to 0.48 in

1978 (Londono, 1995). Two interrelated economic events seem to explain the 1970–1974

reduction in average height. First, beginning in the 1970s the relative price of food began a long

period of increase which lasted until the end of the 1980s. Thus the relative prices of food

increased 71.1% with respect to the overall price index. Although the decline in average height

occurred only for those born in 1970–1974 and the relative price increase extended over two

decades, it has to be kept in mind that those persons born in that five-year period grew up during

1970s and 1980s.

An additional factor which could explain the reduction in height was pointed out by Urrutia

(1984), who has done some of the most important research on income distribution in Colombia.

Urrutia argues that when inflation accelerated in the early 1970s the real income of urban laborers

in the formal sector lagged behind. The reason was that at the time labor unions bargained for


Table 3

Average height (cm) of Colombians by date of birth (ages between 18 and 21)

Year of birth Average height

(cm)

Increase (%) Increase

(cm)

Standard

deviation

Number of

observations

Men Women Men Women Men Women Men Women

1905–1909 162.05 150.02 7.62 8.49 1,163

1910–1914 163.48 150.78 0.88 0.51 1.43 0.77 7.15 8.01 3,948

1915–1919 163.61 151.49 0.08 0.47 0.13 0.71 7.22 7.68 9,575

1920–1924 164.16 152.38 0.34 0.59 0.55 0.89 7.11 7.34 18,865

1925–1929 164.70 153.06 0.33 0.44 0.54 0.67 6.90 7.22 31,958

1930–1934 165.17 153.48 0.29 0.27 0.48 0.42 7.01 7.03 46,838

1935–1939 165.76 154.21 0.35 0.48 0.59 0.74 6.89 6.97 68,457

1940–1944 166.26 154.69 0.30 0.31 0.50 0.48 6.82 6.86 85,809

1945–1949 167.10 155.59 0.51 0.58 0.84 0.90 6.80 6.72 113,887

1950–1954 167.84 156.40 0.44 0.52 0.74 0.80 6.87 6.73 154,698

1955–1959 168.07 156.81 0.13 0.26 0.22 0.41 7.00 6.66 212,891

1960–1964 168.47 157.17 0.24 0.23 0.41 0.36 6.98 6.64 300,004

1965–1969 169.00 157.34 0.32 0.11 0.53 0.18 7.02 6.73 398,488

1970–1974 168.91 157.21 �0.06 �0.08 �0.09 �0.13 7.08 6.64 725,568

1975–1979 169.66 157.81 0.44 0.38 0.75 0.59 6.77 6.29 3,286,377

1980–1984 170.54 158.56 0.52 0.48 0.88 0.76 6.65 6.22 3,469,713

1985 171.01 158.97 0.28 0.25 0.47 0.40 6.64 6.17 393,537

Total 5.53 5.97 8.96 8.95 9,321,776

Source: RNEC.

salary increases for a two-year time period (Lopez, 2003).18 This situation changed around 1975,

when salary increases began to be made on an annual basis.

What happened in the late 1950s is more complex. Between late 1953 and early 1957 the

Colombian economy experienced a boom in the foreign sector due to the high price of coffee, the

main export crop at the time. However, in 1957 the real price of coffee fell in the international

market. By 1960 the real price of Colombian coffee was 42% below its 1956 level (Lopez, 1990).

There was a growing fiscal deficit and the exchange rate became overvalued. A sharp devaluation

in 1957, plus a growing fiscal deficit, increased the level of inflation, from 8.4% in 1956 to 24.3%

in 1957. As a result, rural wages declined in real terms in 1957 and 1958 (Londono, 1995, p. 225).

Per capita GDP also declined, by �0.92% and �0.76% in 1957 and 1958, respectively.

Moreover, this was a period in which an economic and a political crisis coincided. The

increase in the rate of homicides in the late 1950s coincides with a reduction in the average adult

height of men born between 1957 and 1960 (the same occurred with women) (Fig. 4). During the

late 1950s violence in Colombia rose to the highest level in the country’s history up to that time.19

Since the late 1940s the intensity of political competition between the two traditional parties,

Liberal and Conservative, led to increasing levels of violence (Wilde, 1978). The result was a

breakdown of democracy followed by the military dictatorship of General Rojas Pinilla (1953–

1957). On May 10, 1957, Rojas was overthrown and democracy reestablished in 1958 through a

coalition of the two traditional parties. However, violence continued at historically high levels

between 1958 and the early 1960s as a result of rural banditry and the formation of guerrilla

groups (Ofstein, 2003). Since average height also fell in regions with very little political violence,


Fig. 4. Height of Colombian males and homicide rate per 100,000 inhabitants.

18 The estimates of a modified version of Okun’s Index, which also includes deviations in the rate of growth of GDP, by

Jose Ignacio Lopez for the Colombian presidents since 1958, shows that the administration of Misael Pastrana (1970–

1974) obtained the worst results according to this index.19 In 1958 the number of homicides per 100,000 inhabitants rose to 49.

such as the Caribbean Coast departments, the reason for the reduction in average height during

this period was most probably the economic crisis prevailing at the time and not the political

violence.

In sum, the average adult height of Colombians increased during successive generations

throughout the 20th century. Equally important, however, is that the dispersion in height fell

significantly. This reflects an improvement in the inter-personal distribution of biological welfare

(Fig. 5). In the early 1900s the coefficient of variation of the height of Colombian women born at

the time was 0.054; by the 1980s it was 0.038. For males, the coefficient of variation fell from

0.045 in the early 1900s to 0.038 in the early 1980s. However, in the case of males the coefficient

of variation increased from the 1950s to the early 1970s. During this period the coefficient of

variation for women stagnated and then decreased after 1972 to the level of the men. As has been

mentioned, this was also a period during which the distribution of income deteriorated.

Thus, the dispersion in height among the passport applicants (0.04 for both men and women)

was less than the value observed with the national ID card heights, around 0.044 for men and

0.050 for women. Only towards 1980 did the dispersion of heights obtained from the ID data

approaches a level similar to that observed for passports in the period 1870–1919. Thus, it can be

inferred that the group in the passport sample was relatively homogeneous in social and

economic terms.

5. Columbia: a country of many regions

Colombia is one of the most mountainous countries in the world. It is divided by the three

mountain ranges in which the Andes opens itself in northern South America (Luke et al., 2003).20

As a result there are clearly differentiated regions in terms of their culture, economic

development, and even ethnic composition. Since 1886 the main internal political divisions have

been the departments. Currently there are 32 departments. Bogota, the capital, is a special


Fig. 5. Coefficient of variation for the height of Colombians (by year of birth).

20 In a study by the Inter-American Development Bank on the effects of geography on economic development a

geographical fragmentation index, which varies between 1 and 0, was calculated for Latin America and the major regions

of the world. Colombia and Ecuador, with an index above 0.8, had the highest fragmentation among Latin American

countries and their index was above the one found for each of the major regions of the world.


Fig. 6. Colombian departments and territorial units.

territorial unit. In this paper all of the departments created in 1991 have been aggregated, because

they are sparsely populated and there is no disaggregated GDP data for them.21 Also the special

district of Bogota is included. Thus, 24 territorial units are used in the analysis (Fig. 6).

5.1. Increases in height in the regions

As a result of different levels of economic development, geographic endowment, and ethnic

composition, the average height of the inhabitants of the Colombian departments differ

significantly. In Figs. 7 and 8 the average height by department of birth is shown for those


Fig. 8. Average height of women born in Colombia by departments in 1985.

Fig. 7. Average height of men born in Colombia by departments in 1985.

21 Departments created in 1991 have been aggregated in one region called ‘‘New Departments’’.

Colombians born in 1985. The tallest Colombians are the inhabitants of San Andres, both men

and women at a height of 175.2 cm and 162.6 cm, respectively. The department of San Andres is

composed of several small islands in the Caribbean Sea, 180 kilometers off the coast of

Nicaragua. Up to the early 1950s these islands were populated by an Afro-Caribbean, Protestant,

and English-speaking population (Meisel, 2003). Thus, its ethnic characteristics were different in

many ways from that of mainland Colombia. In contrast, the people of the department of Narino,

located in the southern border with Ecuador, were the shortest in 1985, with women 156.1 cm tall

and men 168.2 cm tall (Montenegro, 2002).22 Thus, there was a difference between the height of

the inhabitants of these two departments of seven cm, both women and men.

Throughout the 20th century the economic development of the Colombian departments has

followed different paths because of differences in their relative export success and in their

capacity for industrial growth. Thus, the evolution of the biological standard of living has varied

among them. In terms of height Bogota was most successful. Between 1910 and 1914 and 1985

the average adult height of women born in Bogota increased more than 7% (11 cm), the largest

gain for any territorial unit (Fig. 10). In the case of men the increase was close to 5% (7.95 cm)

(Fig. 9). The rapid economic growth of Bogota in the last few decades has transformed it into

Colombia’s most important growth pole. From 1960 to 2001 its share in total GDP increased

from 14% to 22%. Currently, Bogota is the territorial unit with the highest GDP per capita.23 It

would seem that this economic prosperity is the main cause for the increase in height. Bogota has

received a large influx of immigrants from other parts of Colombia. The majority of immigrants

(86%) came from the departments of Cundinamarca, Boyaca, Tolima and Santander, which have


Fig. 9. Increase in the average height of men born in the departments of Colombia between 1910–1914 and 1985.

22 The inhabitants of Pasto, the capital of Narino, have been the object of deriding jokes by other Colombians. Armando

Montenegro suggests that the reason pastusos have been singled out is that they have a very particular way of speaking

and that the level of social and economic development of that region has lagged behind the rest of Colombia. An

additional reason could be that they are the shortest Colombians.23 GDP per capita of Bogota is the highest one if the new departments are not taken into consideration. New department

only contain 5% of national population.


Fig. 10. Increase in the average height of women born in the departments of Colombia between 1910–1914 and 1985.

Table 4

Annual growth rate of height by departments

Men Women

Department Annual growth rate

(1910–1985)

Department Annual growth rate

(1910–1985)

Bogota 0.078 Bogota 0.087

Santander 0.072 Santander 0.086

Tolima 0.068 Narino 0.085

Cundinamarca 0.067 Cundinamarca 0.081

Norte Santander 0.066 Boyaca 0.077

Boyaca 0.064 Tolima 0.073

Narino 0.063 Huila 0.067

Risaralda 0.058 Norte Santander 0.063

Quindıo 0.056 Cesar 0.060

Valle 0.054 Antioquia 0.059

Antioquia 0.054 Risaralda 0.058

Huila 0.054 Valle 0.056

Sucre 0.053 Bolıvar 0.051

Caldas 0.052 Cordoba 0.050

Bolıvar 0.049 Sucre 0.049

Choco0 0.049 Caldas 0.047

Atlantico 0.047 Cauca 0.047

Cordoba 0.045 Quindıo 0.046

San Andres 0.043 Choco 0.044

Cesar 0.041 Magdalena 0.039

Magdalena 0.039 Atlantico 0.031

Cauca 0.035 La Guajira 0.019

La Guajira 0.024 San Andres �0.013

a similar ethnic composition, so that there were no changes in the ethnic makeup of the city

(DANE, 1993). Hence, immigration could have affected average heights only if the taller than

average persons migrated to the cities from these provinces.

The experience of San Andres in the period 1910–1985 is almost opposite to that of Bogota.

The annual growth rate in average height for men of San Andres (0.04%) was one of the lowest in

the country, while that of men from Bogota was the highest (0.08%) (Table 4). In the case of

women a similar pattern is observed, except that women from San Andres had the only negative

growth rate among the departments (Table 4). The most likely explanation for the behavior of

average height in San Andres is the change in the ethnic composition of the island that began in

the 1950s, when it was declared a free port by the Colombian government. The possibility of

buying duty free foreign goods attracted a large influx of tourists and immigrants to San Andres

from continental Colombia. By 1964 immigrants represented 51% of the population. Currently

the Afro-Caribbean English-speaking population, the so called raizales, constitutes a minority of

San Andres’s inhabitants. Most of these immigrants came from the continental Colombian

departments of Atlantico and Bolivar, and their average height was below that of the raizales.

Thus, the observed reduction in the average height of people born in the island in the last four

decades does not imply that the average height of raizales has diminished.

The departments of the Caribbean Coast (Cordoba, Sucre, Bolivar, Atlantico, Magdalena,

Cesar, and Guajira) are among those that had the smallest gains in average adult height in the

period 1910–1985 (Figs. 9 and 10). Perhaps it is not accidental that as a result of the consolidation

of coffee exports and industrialization in other areas of the country, throughout most of the 20th


Table 5

Estimates of b-convergence in the average height of Colombians at the department level, period 1910–1985

Period: 1910–1985 Beta (b) Standard error t-Statistic Probability R2 Correlation coefficient

Men 0.00707 0.00088 8.0702 0.00 0.69 �0.92

Women 0.00864 0.00078 10.9607 0.00 0.83 �0.97

Fig. 11. Average height of Colombians born in the 10 largest cities and those born in the rest of Colombia.


Fig. 14. Sigma convergence in the average height of Colombian males, by departments.

Fig. 12. Beta convergence in the average height of Colombian males by departments.

Fig. 13. Beta convergence in the average height of Colombian females by departments.

century the economy of the Caribbean region lagged behind that of the rest of the country and is

currently the poorest of Colombia (Meisel, 1999).

However, at the beginning of the 20th century the inhabitants of the Caribbean Coast were

among the tallest. The reason for this was the large percentage of the population of African

ancestry, as well as the fact that the nutritional status was comparatively good, especially because

of the consumption of meat and fish. The region was the main cattle producing area of the country

and was scarcely populated, so meat consumption was among the highest in Colombia (Vergara-

Velasco, 1901). Additionally, the inhabitants of the Coast had access to fish from the sea and the

numerous rivers and lakes of the region.

Unlike what happened during the process of industrialization and urbanization in some of the

currently developed countries, where the height of rural inhabitants tended to be above that of

urban dwellers, the Colombian pattern is the opposite (Floud et al., 1990; Komlos, 1989).

Throughout the 1910–1985 period, the average height of those born in the 10 largest Colombian

cities, which currently represent 36% of the total population, has been consistently around 2 cm

above that of the rest of the country (Fig. 11). This coincides with the dualism that prevails in the

Colombian economy in which the cities are more prosperous and have a higher standard of living

than rural areas, as common in many developing countries.

5.2. Convergence analysis

The analysis of how regional disparities in height evolved in Colombia during the 20th

century follows the methodology proposed by Barro and Sala-i-Martin (1991) for the study of


Table A.1

Trend in the height of men in the passport sample

Note: N, normal; P, percentile; BC, bias-corrected.

convergence in per capita GDP. Beta (b) convergence occurs when the unit that was farthest

behind grows faster. For Columbian men and women there was regional b-convergence in height

in the period 1910–1985 (Table 5). That is, those departments which were the tallest in 1910

grew more slowly than those that were the shortest. For example, Norte de Santander had the

shortest men in 1910, and it experienced the fastest rate of growth in height (Fig. 12). In contrast,

Magdalena, which had the tallest men in 1910, had the lowest rate of growth in the period 1910–

1985. b-convergence is similar among women (Table 5 and Fig. 13). The department with the

shortest women in 1910, Narino, experienced one of the fastest rates of growth in height in the

period 1910–1985. At the opposite end, the women of San Andres, the tallest in 1910, grew the

least between 1910 and 1985.

The presence of b-convergence is a necessary but not a sufficient condition for the

reduction in the dispersion of a variable. In the case of the average height of men and

women in the Colombian departments in the period 1910–1985 there was also a process of

reduction in the observed dispersion, sigma convergence (s). Fig. 14 shows the evolution

of s-convergence, measured as the standard deviation of the logarithm of departmental

height, for both men and women. In both cases there is a clear secular trend towards a reduction

of the dispersion of average height. In the case of men, s declines from around 0.02 at

the beginning of the period to 0.007 in the early 1980s, while for women the reduction is

from 0.03 to 0.008. Particularly significant reductions occurred in the periods 1910–1930 and

1972–1980.


Table A.2

Trend in the height of women in the passport sample


6. Conclusion

For Colombians born between 1870 and 1919 and who obtained a passport, mostly members

of the elite, the average height was stagnant throughout that 49-year period. However, men were

5 cm taller than the average Colombian and women by 7 cm (Tables 1 and 3). To be sure, by

present Colombian standards even the elite was short. While the average height of men in that

elite group in 1900 was 168.2 cm, Colombians born in 1985 had an average height of 170.6 cm.

Even if the Colombian elite was well nourished at the beginning of the 20th century, the health

conditions under which it lived were probably retarding its physical growth.

The study of the evolution of the average height of Colombians in the period 1905–2003 based

on identification card records reveals a remarkable achievement. On average, Colombian men

and women born in 1985 reached an adult stature that was 8.9 cm above those born in 1905/1909.

These advances were almost uninterrupted through the 20th century and throughout the country.

Average height declined only twice, 1957–1960 and 1970–1974, and there was an immediate

return to the long run trend of increasing height.

The evolution of the biological standard of living in Colombia in the 20th century was also a

success story from the point of view of the reduction of inequalities that existed initially. There

was a reduction in the dispersion of height of Colombians at the individual level, as measured by


Table A.3

Trend in the coefficient of variation male height in the passport sample


the standard deviation of the logarithm of this variable. Thus, in this dimension of the overall

standard of living, Colombia became a more equal society. There was also convergence in

average height at the departmental level, as reflected by both b- and s-convergence.

The main reason for the increase in height was the continuous improvement in the nutrition of

Colombians throughout the century due to sustained economic growth. Additionally, advances in

health and reductions in work effort, for reasons such as the reduction of child labor, contributed

to these gains.

Acknowledgements

Archivo General de la Nacion and Ministry of Foreign Relations provided records of

passports. Registradurıa Nacional del Estado Civil provided Digital Database of Colombian

citizenship card.

Authors also acknowledge the collaboration of Miguel Urrutia, David Zabala, Francisco

Rivas, Diana Mejıa, and Camilo Morales, and comments by Javier Perez, Haroldo Calvo, and

Julio Romero.

Appendix A. Results for parametric regression to show significance of trend

.


Table A.4

Trend in the coefficient of variation male height in the passport sample


References

Abel, C., 1996. Ensayos sobre historia de la salud en Colombia, 1990–1996. CEREC, Bogota.

Abrisqueta, F., 1940. Las condiciones y el costo de vida de la clase obrera en Medellın. Anales de Economıa y Estadıstica

3 (6 Oct), 36–62.

Barro, R., Sala-i-Martin, X., 1991. Convergence across states and regions. Brookings Papers on Economic Activity, vol. 1,

pp. 107–182.

Ben-David, D., Papell, D., 1997. Slowdowns and meltdowns: post-war growth evidence from 74 countries. NBER

Working Paper, vol. 6266, pp. 1–34.

Bushnell, D., 1971. Voter participation in the Colombian election of 1856. Hispanic Am. Hist. Rev. 51 (2), 327–349.

Bushnell, D., 1984. Las elecciones presidenciales, 1863–1883, vol. 18. Revista de la Universidad Nacional, Medellın, pp.

219–314.

Bushnell, D., 1994. La elecciones en Colombia: Siglo XIX. Credencial Historia, Bogota.

Cotes, M., 1893. Regimen alimenticio de la Sabana de Bogota. Imprenta La Luz, Bogota.

DANE, 1993. XVI Censo Nacional de Poblacion y V de Vivienda, Censo 1993. Bogota.

Deas, M., 2002. El papel de la Iglesia, el ejercito y la policıa en las elecciones colombianas entre 1850 y 1930. Boletın

cultural y bibliografico, Biblioteca Luis Angel Arango, Bogota, 39 (60), 350–386.

FAO, 2001. Perfiles nutricionales por paıses: Colombia. Roma, Mimeo.

Florez, C., 2000. Las transformaciones sociodemograficas en Colombia durante el siglo XX. Tercer Mundo Editores,

Bogota.

Floud, R., Gregory, A., Wachter, K., 1990. Height, Health and History: Nutritional Status in the United Kingdom 1750–

1980. Cambridge University Press, Great Britain.

Galvis, L., 2000. La demanda de carnes en Colombia: Un analisis econometrico. Documentos de trabajo sobre economıa

regional 13. Centro de Estudios Economicos Regionales, Banco de la Republica, Cartagena.

GRECO, 2002. Grupo de estudios del crecimiento economico colombiano. El Crecimiento economico colombiano en el

siglo XX. Banco de la Republica, Fondo de Cultura Economica, Bogota.

Honda, G., 1997. Differential structure differential health: industrialization in Japan, 1868–1940. In: Steckel, R.H., Floud,

R. (Eds.), Health and Welfare During Industrialization. NBER, University of Chicago Press, USA, pp. 251–284.

Komlos, J., 1989. Nutrition and Economic Development in the Eighteenth-Century Habsburg Monarchy, an Anthro-

pometric History. Princeton University Press, USA.

Komlos, J., 2004. How to (and how not to) analyze deficient height samples: an introduction. Hist. Methods 37 ((4) Fall),

160–173.

Londono, J.L., 1995. Distribucion del ingreso y desarrollo economico, Colombia en el siglo XX. Tercer Mundo Editores,

Bogota.

Lopez, A., 1990. El Banco de la Republica y la polıtica macroeconomica, 1955–1962. In: Banco de la Republica,

Antecedentes Evolucion y Estructura, Banco de la Republica, Bogota, pp. 463–503.

Lopez, J., 2003. Calificando a los presidentes: Una aplicacion para Colombia del ındice de Okun. Carta Financiera 125,

12–25.

Lopez-Alonso, M., Porras, R., 2003. The ups and downs of mexican economic growth: the biological standard of living

and inequality 1870–1950. Econ. Hum. Biol. 1, 169–186.

Luke, J., Alejandro, G., Eduardo, L., 2003. America Latina: Condenada por su geografıa. Banco Interamericano de

Desarrollo, Washington, DC.

Meisel, A., 1999. Por que perdio la Costa Caribe el siglo XX? In: Calvo, H., Meisel, A. (Eds.), El rezago de la Costa

Caribe Colombiana. Banco de la Republica-Fundesarrollo-Universidad del Norte-Universidad Jorge Tadeo Lozano,

Seccional del Caribe, Cartagena, pp. 69–105.

Meisel, A., 2003. La continentalizacion de la Isla de San Andres: Panyas, raizales, y turismo, 1953–2003. Aguaita, vol. 9.

Observatorio del Caribe, Cartagena, pp. 7–22.

Montenegro, A., 2002. Una historia en contravia: Pasto y Colombia. El Malpensante, Bogota, pp. 1–244.

Munoz, C., 1996. The Working Child in Colombia since 1880, Child Labor in Historical Perspective 1800–1985.

UNICEF, Italy.

Ofstein, N., 2003. An Historical Review and Analysis of Colombian Guerrilla Movements. Documentos Cede 21,

Universidad de los Andes.

Ordonez, A., Polania, D., 1992. Cambios de estatura en Colombia durante el presente siglo. Coyuntura Social 6 (June).

Posada, E., 1994. Elections and Civil Wars in 19th Century Colombia: The 1875 Presidential Campaign. J. Latin Am.

Stud. 26 (3), 621–649.


Posada, E., 1997. Limits of power: elections under the conservative hegemony in Colombia, 1886–1930. Hispanic Am.

Hist. Rev. 77 (2), 245–279.

Registradurıa Nacional del Estado Civil, 1988. Historia electoral Colombiana, Bogota.

Salamanca, L., 1959. Manual Para el Servicio Exterior de Colombia. Ministerio de Relaciones Exteriores, Bogota.

Salvatore, R., 2004a. Stature nutrition and regional convergence. The Argentine northwest in the First Half of the 20th

Century. Soc. Sci. History 28 (2), 297–324.

Salvatore, R., 2004b. Stature decline and recovery in a food-rich export economy: Argentina 1900–1934. Explorations in

Economic History 41 (July), 233–255.

Steckel, R., 1995. Stature and the standard of living. J. Econ. Lit. 33 (4), 1903–1940.

Tanner, J., 1994. Growth in height as a mirror of the standard of living. In: Komlos, J. (Ed.), Stature, Living Standards and

Economic Development: Essays in Anthropometric History. West View, USA, pp. 1–6.

Urrutia, M., 1984. Los de arriba y los de abajo, La distribucion del ingreso en Colombia en las ultimas decadas. CEREC,

Bogota.

Vergara-Velasco, F., 1901. Nueva geografıa de Colombia. Tomo 1. Imprenta de Vapor, Bogota.

Wilde, A., 1978. Conversations among Gentlemen: Oligarchical Democracy in Colombia. In: Linz, J.J., Stepan, A.

(Eds.), The Breakdown of Democratic Regimes, Latin America. The Johns Hopkins University Press, USA,

pp. 28–81.


The biological standard of living (and its convergence) in Colombia, 1870-2003. A tropical success...

Documents

Transcript of The biological standard of living (and its convergence) in Colombia, 1870-2003. A tropical success...