Hum Nat

DOI 10.1007/s12110-010-9092-2

Examining the Relationship between Life Expectancy,

Reproduction, and Educational Attainment
A Cross-Country Analysis
Nicola L. Bulled & Richard Sosis

# Springer Science+Business Media, LLC 2010

Abstract Life history theory aims to explain the relationship between life events,
recognizing that the fertility and growth schedules of organisms are dependent on
environmental conditions and an organisms ability to extract resources from its
environment. Using models from life history theory, we predict life expectancy to be
positively correlated with educational investments and negatively correlated with
adolescent reproduction and total fertility rates. Analyses of UN data from 193 countries
support these predictions and demonstrate that, although variation is evident across
world regions, strong interactions exist among life expectancy, reproductive investments, and educational attainment, and these relationships occur independently of
economic pressures and disease burdens. The interactions are strongest, however, in
countries with a life expectancy of 60 years as these countries tend to have stable
economies and a limited HIV/AIDS burden. These findings suggest that policies aimed
at influencing education and reproductive decisions should consider environmental
characteristics that drive peoples expectations about their longevity.
Keywords Demographic factors . Educational status . Fertility . Life cycle .
Mortality . Reproductive behaviors
According to the World Health Organizations World Health Report (2008), people
are healthier, wealthier, and living longer today than 30 years ago. The average
global life expectancy at birth is estimated to increase by 7 years from 1998 to 2025,
with 26 countries having a life expectancy at birth above 80 years. Increases in
global life expectancy are attributed to improvements in sanitation and access to
clean water; medical advances, including childhood vaccines; and massive increases
N. L. Bulled (*) : R. Sosis
Department of Anthropology, U-2176, University of Connecticut, Storrs, CT 06269-2176, USA
e-mail: Nicola.Bulled@uconn.edu
R. Sosis
Department of Anthropology, University of Connecticut, Storrs, CT 06269-2176, USA

Hum Nat

in agricultural production as a result of the Green Revolution. Rapid population

growth resulting from decreased mortality began in 1950. However, the global
population growth rate has decreased by almost half since reaching its peak of 2.2%
in 19651970 (United Nations 1999) as a result of many countries going through
demographic transitions characterized by increasing education and declining fertility.
In this article we examine how changes in life expectancy influence the timing and
extent of reproduction and investments in education.
Life history theory recognizes that the timing of life events (birth, adolescence,
reproductive onset, reproductive termination, and death) is dependent on environmental conditions and an organisms ability to extract resources from its
environment (Roff 2002; Stearns 1992). Natural selection is assumed to have
designed organisms to balance the inherent trade-offs between investments in
reproduction and growth (often considered as trade-offs between current and future
reproduction). Fluctuating environmental conditions that increase extrinsic mortality
favor increased reproductive effort and short-lived organisms (i.e., little investment
in somatic maintenance; Schaffer 1974). When the risks of mortality are high,
organisms are expected to reproduce frequently, to increase the probability of some
offspring surviving to maturity, and early, to ensure reproduction before death
(Koons et al. 2008). However, in stable environments with low risks of extrinsic
mortality, selection can favor delays in reproduction to invest in somatic growth,
which not only impacts survivorship but also increases an organisms ability to
extract resources from the environment. When competition for resources is high in
stable environments, selection favors greater parental investment and a reduced
number of offspring (Promislow and Harvey 1990; Stearns 1992; Wilson 1975).
Life history theory is concerned with investments in physical growth, but organisms
especially humansalso make significant ontogenetic investments in knowledge and
skills (Hill and Kaplan 1999; Kaplan et al. 2009). These educational investments can
be analyzed as part of the basic life history trade-offs between growth and
reproduction (Eisenberg 1981; Gould 1977; Hill and Hurtado 1996; Kaplan et al.
2000; Kaplan and Lancaster 2003; Lerner 1984). Investments in education delay
reproduction but increase future income as well as survivorship, a relationship that has
been observed worldwide (Schultz 1993). For example, in Brazil, increasing years of
education were found to correlate with incremental decreases in total fertility and
increases in wages (Lam and Duryea 1999). In a Thai population experiencing
declining fertility rates, Knodel et al. (1990) found that children from smaller families
were more likely to continue education to higher levels.
However, if extrinsic mortality or future unpredictability is high, delays in
reproduction will not be favored, nor will investments in education whose benefits
cannot be realized until later in the lifespan. Various studies have found that adolescents
who anticipate having a shorter lifespan reproduce at an earlier age than adolescents who
expect to have a longer lifespan (Brumbach et al. 2009; Geronimus 1996a, b, 2001,
2004; Hill et al. 1997; Wilson and Daly 1997). In addition, in the United States
between 1980 and 2001, Meara et al. found very little change in life expectancy
among less-educated black and white non-Hispanics and very substantial increases in
life expectancy among the more educated (2008:356). Similar relationships between
education and life expectancy have been observed elsewhere, including Brazil

Hum Nat

(Camargos et al. 2007), Finland, Sweden, Norway, Denmark (Silventoinen and

Lahelma 2002; Valkonen et al. 1997) and Europe (Mackenbach et al. 1997).
In this paper we build on Low et al.s (2008) life history analysis of cross-cultural
patterns of female reproduction. Their analysis of more than 170 nations reveals a
strong relationship between life expectancy and age at first birth; however, they found
a threshold life expectancy at 60 years that affects the relationship between life
expectancy and age at first birth across countries. Life expectancy and age at first birth
are positively correlated in countries in which life expectancy at birth is 60 years, but
for countries with life expectancies shorter than 60 years, no clear relationship exists.
The authors attribute the lack of a relationship in countries with life expectancies less
than 60 years to potential states of non-equilibrium resulting from volatile socioecological conditions.
Life expectancy at birth is a demographic indicator describing, as defined by the
United Nations Human Development Report, the number of years a newborn infant
would live if prevailing patterns of age-specific mortality rates at the time of birth
were to stay the same throughout the childs life (2007:368). Low et al. (2008)
analyze the relationship between life expectancy and reproduction by stratifying
country-level data into development status using a composite indicatorthe Human
Development Index (HDI). First calculated in 1990, HDI serves as a frame of
reference for both social and economic development. HDI measures the average
achievements of a country in a single statistic by combining indicators of health,
knowledge, and standard of living. Health is measured by life expectancy at birth;
knowledge is measured by a combination of the adult literacy rate (%) and the
combined gross primary, secondary, and tertiary school enrolment ratios; and
standard of living is measured by Gross Domestic Product per capita (Purchasing
Power Parity US$).
Analyses of the relationships among life expectancy, reproduction, and education
relying on stratifications by HDI are unable to take into account the composite
indicators direct relationship with life expectancy and education. In addition, analyses
cannot control for, or isolate the effects of, economic development, as these are
included in the composite HDI value. Low et al. conclude: If we can isolate particular
components that strongly affect life expectancy and AFB [womens age at first birth],
we [will] have a clearer idea of what relatively easily influenced variables might be the
focus of policy interventions to increase life expectancy and AFB (2008:215).
This paper aims to take on that challenge by isolating the component measures of
HDIlife expectancy at birth, adult literacy rate, gross enrolment in school, and GDP
to examine the interaction among life expectancy, reproduction, and education. The
relationships between life expectancy, reproductive behavior, and educational investments consist of a feedback loop in which improvements in education are expected to
increase life expectancies and favor declines in reproduction, which further increases
investments in education (Hill and Kaplan 1999). The causal links among life
expectancy, reproduction, and educational investments are dependent on the stage of
the demographic transition a population is experiencing. Countries vary in the rates at
which they pass through stages in the demographic transition. Some countries, such as
China, Brazil, and Thailand, have moved through the stages of demographic transition
rapidly as a result of economic and social changes; other countries, particularly in

Hum Nat

Africa, have stalled owing to economic stagnation and the impact of AIDS. Therefore,
when examining the relationships among life expectancy, reproduction, and education
it is important to control for economic factors (e.g., Dorling et al. 2006; Lamptey et al.
2006; Preston 1975; Rodgers 1979; Sen 1993) and disease indicators (e.g., Anderson
2010; Quinlan 2007). Indeed, these factors may account for the lack of any correlation
between life expectancy and age at first birth among populations with low life
expectancy (<60 years) as observed by Low et al. (2008).
Here, building on Low et al.s (2008) results and recommendations, we examine
the components of the HDI that affect life expectancy across societies in three steps:
(1) correlating life expectancy with variables for reproduction and education, and
factors known to influence life expectancy, such as GDP and HIV/AIDS deaths; (2)
determining if a threshold life expectancy exists for measures of reproduction and
education used in our analyses, consistent with the findings of Low et al. (2008); and
(3) controlling for the influence of covariates and regional variation on the
relationships among life expectancy, reproduction, and education, and determining
if differences in the relationships exist between counties with high and low life
expectancy. Based on the above discussion, we predict life expectancy to be
negatively correlated with adolescent reproductive rate and total fertility and
positively correlated with educational investments, whilst controlling for covariates
and regional variation, and that these relationships may not exist in countries with
low life expectancy.

Methods
We acquired the data used in our analyses from specialized agencies of the United
Nations system, including health (World Health Organization, WHO), education
(United Nations Education Scientific and Cultural Organization, UNESCO), economics
(World Bank), and life expectancy (World Population Prospects). National datasets must
be regarded with some level of caution as data gaps and issues of inconsistency and
incoherence remain owing to differences in the effectiveness of infrastructure, political
agendas, and additional factors, such as internal conflicts. However, the UN agencies
rely on an extensive peer review process conducted through leading regional and
national statistical offices as well as international organizations to ensure the highest
level of data consistency and accuracy. For this analysis we used data from 2006.
Missing data were replaced with 2005 data when available.
Our analysis includes all components of the HDI: life expectancy; adult literacy
rate; primary, secondary, tertiary, and combined gross school enrolment ratios; and
GDP. School enrolment ratios are the numbers of students enroled in primary
(elementary and middle school), secondary (high school), and tertiary (college or
university) levels of education, regardless of age, expressed as a percentage of the
population of theoretical age corresponding with those levels (United Nations 2007).
Gross Domestic Product (PPP US$) (GDP) is the sum of the net output of all
producers in an economy, plus product taxes not included in the valuation of output,
converted to US dollars, and accounting for price differences across countries such
that one US dollar has the same purchasing power in the domestic economy as it
does in the United States (United Nations 2007).

Hum Nat

In addition to the variables used in the calculation of the composite HDI index,
our analyses examine adolescent birth rate and total female fertility. Adolescent birth
rate is defined as the annual number of births to women aged 1519 years per 1,000
women in that age group (WHO 2009). Total female fertility is defined as the
number of children that would be born to each woman if she were to live to the end
of her childbearing years and bear children at each age in accordance with agespecific fertility rates in the region and in a given year (United Nations 2007). Our
analyses also include a disease indicatordeaths from HIV/AIDS, defined as the
estimated number of adults and children that have died because of HIV/AIDS in a
specific year, expressed per 100,000 population (WHO 2009).
Bivariate correlations (Pearsons r) were conducted to identify relationships among
life expectancy, reproduction, educational investment, and covariates. Correlation
analyses do not require the grouping of variables or imply causal directionality. Given
the evidence of existing variation in womens education, marriage, and fertility
prospects across nations (Low 2005), the 193 nation-states used in this analysis were
categorized into eight UNESCO world regions to further isolate indicators that may
have greater impact in some regions relative to others (see Tables 1 and 7 for world
regions). Variations between countries within the UNESCO regional categories likely
exist; however, these regions do offer some categorization of countries by level of
development, ethnicity, religion, culture, and disease burden. In addition, we did not
weight data by the relative population size of each country because relationships in
countries with large populations would override those in countries with smaller
populations, thus negating the impact of sociocultural factors.
To analyze both trends and potential thresholds in the relationships between life
expectancy and investments in reproduction or education, we divided countries into
five groups based on life expectancy and calculated median values for indicators of
reproduction (adolescent birth rate and total female fertility) and education (adult
literacy rate and primary, secondary, tertiary, and combined gross school enrolment
ratios). Median values were compared across life expectancy stratifications to
identify threshold life expectancies.
Finally, we used multivariate regression analysis to control for factors known to
impact life expectancy (GDP and HIV/AIDS deaths) and regional differences
(dummy variables). We elected this method rather than using adjusted life
expectancy estimates that exclude deaths from HIV/AIDS so as to observe the
impact of both economic and disease indicators independently. We also conducted
separate multivariate regression analyses for countries with life expectancies above
and below 60 years to determine if the potential thresholds hold when controlling for
covariates. All analyses were conducted using SAS version 9.2 for Windows (SAS
Institute, Cary, NC, 2002).

Results
Table 1 shows the bivariate correlations among life expectancies, indicators for
reproduction, educational attainment, and possible covariates. Increases in life
expectancy correspond to significant decreases in adolescent birth rate and total
female fertility. All variables measuring educational attainment, excluding primary

Arab States
(n)

.691** (9)

.103 (182)

.810*** (174)

.676*** (145)

Primary school
enrolment ratio

Secondary school
enrolment ratio

Tertiary school
enrolment ratio

.217 (19)
.269 (19)

.893*** (20) .164 (8)

.043 (8)

.556*** (20)
.938*** (9)

.293 (21)

.389* (21)

.313 (20)

.644*** (20) .299 (8)

.046 (9)

.137 (21)

.800*** (20) .170 (9)

.604** (16)

.057 (21)

.191 (9)

.680*** (20)

.613*** (182) .737*** (20)

*p<0.10, **p<0.05, ***p<0.01

HIV/AIDS deaths
per 100,000
population

Gross Domestic
Product (GDP)

.595*** (193)

.753*** (190)

Overall school
enrolment ratio

Covariates

.699*** (186)

Adult literacy rate

(%)

Education

.021 (21)

.469** (21)

Central Asia Central &

(n)
Eastern
Europe (n)

Total female fertility .805*** (188) .674*** (20) .725** (9)

Adolescent birth rate .705*** (181) .505** (19)

Female fertility

Overall life
expectancy (n)

World region

.487** (25)

.776*** (29)

.673** (23)

.743*** (24)

.196 (25)

.686*** (28)

.783*** (23)

.660 (29)

.596*** (25)

.269 (31)

.396** (35)

.392* (24)

.142 (34)

.031 (34)

.488***(35)

.502*** (35)

.605*** (32)

.192 (31)

East Asia &

Latin America
the Pacific (n) & the
Caribbean

.044 (23)

.035 (25)

.128 (23)

.127 (24)

.349* (24)

.271 (24)

.267 (25)

.114 (23)

.381* (23)

.278* (44)

Sub-Saharan
Africa (n)

.262 (9)

.611* (9)

.634* (7)

.706** (9)

.124 (9)

.520 (9)

.820*** (9)

.556*** (44)

.226 (45)

.337* (33)

.464*** (36)

.062 (42)

.324** (44)

.136 (44)

.953*** (9) .435*** (45)

.750** (9)

Northern America & Southwest

Western Europe (n) Asia (n)

Table 1 Pearsons correlations between life expectancy and indices for reproduction, education, GDP, and HIV/AIDS by world region

Hum Nat

Hum Nat

school enrolment, were highly correlated with overall life expectancy, suggesting that
there are corresponding increases in life expectancy and investment in education.
The lack of an association between life expectancy and gross primary school
enrolment (r=0.103, p=0.166) may be a result of overall high enrolment rates, and
therefore a lack of variation both globally and locally relative to secondary and
tertiary school enrolment [mean (SD): 102.4 (15.1), 75.5 (28.6), and 31.66 (26.4),
respectively]. These findings also may suggest that a threshold level of education is
necessary for improvements in life expectancy beyond a certain number of years
(Lam and Duryea 1999) or, conversely, that a threshold life expectancy is needed for
educational attainment beyond a certain level.
We examined correlations between GDP and life expectancy at the global and regional
levels. Consistent with other research (Dorling et al. 2006; Lamptey et al. 2006; Preston
1975; Rodgers 1979; Sen 1993) we found significant bivariate correlations between life
expectancy and GDP on a global scale (Table 1). However, there was considerable
variation in this relationship by region, with three of the eight regions exhibiting no
significant relationship between standard of living (GDP) and life expectancy (Table 1).
Examination of the impact of disease, using HIV/AIDS deaths per 100,000 population,
found a significant, negative correlation with life expectancy. As one would expect,
increases in HIV/AIDS deaths correspond to declines in life expectancy. Regional
differences exist as countries with high HIV/AIDS prevalence experience a drop in life
expectancy (Lamptey et al. 2006). Variations in the relationship between life expectancy
and HIV/AIDS may be a result of the rapid and dramatic influence HIV/AIDS has had
on life expectancy in certain countries (Low et al. 2008). Dramatic shifts in life
expectancy over relatively short periods of time may confound the relationship between
life expectancy and reproduction. However, Gant et al. (2009) found that not all
countries with a heavy HIV/AIDS burden have experienced drastic declines in life
expectancy. As these authors recognize, [s]adly, for many people in Sub-Saharan
Africa, AIDS is just another way of dying (2009:43).
Table 2 displays the median adolescent birth rate, total female fertility, adult
literacy rate, and school enrolment ratios when countries are separated into five
strata of life expectancy. As anticipated, these data show decreasing trends in
adolescent birth rates and total female fertility and increasing trends in the education
measures (with the exception of gross primary school enrolment ratio) as life
expectancy increases. Median gross primary school enrolment ratios remain fairly
stable across life expectancy strata, consistent with correlations in Table 1. Figure 1a
shows the variation, but also consistent negative trends, across regions when total
female fertility is plotted against life expectancy. Figure 1b shows the positive
relationship between adult literacy rates and life expectancy in regions where there is
sufficient variation in adult literacy rates across countries to detect such a trend.
Trends in reproduction and education variables, however, are not linear. Table 2
shows that the greatest differences (approximately twofold) in median reproduction
rates (adolescent birth rate and total female fertility) occurred between the second
(51.061.2 years) and third (61.371.2 years) life expectancy strata, indicating the
same clear threshold at approximately 60 years that Low et al. (2008) found for age
at first birth. The greatest differences in median adult literacy rate and overall,
secondary, and tertiary enrolment ratios also occurred between the second and third
life expectancy strata, ranging from approximately 1.5-fold to a fourfold increase.

Hum Nat
Table 2 Summary statistics for indicators of education and reproduction by five life expectancy groups
Variable

Life expectancy
40.2
50.0 years

51.0
61.2 years

61.3
71.2 years

71.3
75.2 years

75.3
82.4 years

Median, SD
(n)

Median, SD
(n)

Median, SD
(n)

Median, SD
(n)

Median, SD
(n)

Adolescent birth rate

136.00, 59.49 105.50, 53.18 47.00, 33.02

(24)
(24)
(43)

30.00, 32.12
(42)

14.00, 23.58
(48)

Total female fertility

5.55, 1.34
(24)

4.83, 1.25
(24)

2.55, 1.05
(47)

2.20, 0.82
(44)

1.75, .49
(49)

Adult literacy rate (%)

67.40, 18.35
(23)

60.95, 20.64
(24)

90.55, 17.81
(42)

92.40, 6.36
(46)

99.00, 4.64
(51)

Overall school enrolment 52.50, 13.42

ratio
(23)

52.65, 13.30
(24)

71.90, 13.25
(47)

77.35, 7.52
(46)

89.00, 11.56
(50)

Primary school enrolment 105.40, 20.76 93.87, 27.79

ratio
(21)
(22)

102.41, 11.06 103.40, 10.78 102.10, 6.97

(46)
(44)
(49)

Secondary school
enrolment ratio

34.90, 14.01
(19)

80.76, 23.60
(46)

87.12, 11.57
(44)

98.84, 13.06
(48)

3.73, 3.38
(17)

15.41, 20.74
(33)

33.12, 18.53
(32)

56.16, 25.10
(47)

30.37, 22.35
(17)

Tertiary school enrolment 3.55, 3.85

ratio
(16)

Correlations between life expectancy, adolescent birth rate, and total female fertility
calculated for countries with a life expectancy of <60 years were not significant
(adolescent birth rate: r=0.253, p=0.094, n=45; total female fertility: r=0.241, p=
0.112, n=45). Further analyses reveal that countries with a life expectancy of 60 years
show strong positive correlations between life expectancy and both adolescent birth
rate and total female fertility (adolescent birth rate: r=0.511, p<0.001, n=136; total
female fertility: r=0.588, p<0.001, n=143).
Correlations between life expectancy and education calculated for countries with a
life expectancy of <60 years were not significant (primary: r=0.239, p=0.127, n=44;
secondary: r=0.092, p=0.597, n=42; tertiary: r=0.145, p=0.421, n=35; Fig. 2). In
contrast, countries with a life expectancy of 60 years show strong positive
correlations between life expectancy and secondary and tertiary school enrolment
(primary: r=0.002, p=0.986, n=140; secondary: r=0.677, p<0.001, n=139 ;
tertiary: r=0.519, p<0.001, n=112; Fig. 2). These findings suggest either that high
secondary school enrolment ratios are important for improvements in life expectancy
beyond 60 years or that a life expectancy beyond 60 years is fundamental for
education at or beyond the high school level. These relationships are examined further
using regression analysis to control for covariates.
Given the variation that exists in the standard of living and disease burden
indicators, additional analyses presented here examine the relationship among life
expectancy, education, and reproduction controlling for GDP, HIV/AIDS deaths, and
regional variation. Results obtained from multivariate linear regression analyses are
shown in Tables 3, 4, 5, and 6.

Hum Nat

Fig. 1 Patterns of life expectancy, total female fertility, and adult literacy rate by world region: x-axis on
all graphs is life expectancy (years) and y-axis is a total female fertility and b adult literacy rate (%)

All variables in our analyses are highly intercorrelated, so we initially examine

them separately before including them in models with reproduction, education, and
life expectancy. We also examine regional differences and provide models of two
representative regions: northern America and western Europe, which hold the
highest life expectancy estimates, and sub-Saharan Africa, which includes countries
with the lowest life expectancy estimates. For relevant models, we will indicate
regions that are similar to our two sample regions but will not present all of the
regression models here (they are available upon request). Finally, to identify
differences based on threshold values, we present models separately for countries
with life expectancies above and below 60 years of age.
Tables 3 and 4 reveal significant associations between variables measuring
reproduction (adolescent birth rate and total female fertility) and life expectancy. The
relationships remain highly significant when GDP, HIV/AIDS deaths, and dummy
variables for both regions are added to the models. Although multivariate models
that include life expectancy suggest a negative relationship between HIV/AIDS

Hum Nat
Fig. 2 Patterns of life expectancy and gross primary, secondary,
and tertiary education enrolment
ratios by world region

Hum Nat
Table 3 Multivariate linear regression models showing the association of standardized regression
coefficients for life expectancy and covariates (GDP and HIV/AIDS deaths per 100,000 people, dummy
variables for the regions Northern America & Western Europe and Sub-Saharan Africa) with adolescent
birth rate
Adolescent birth rate
.775***

.777***

.489***

Gross Domestic Product (GDP)

.078

.091

.138**

HIV/AIDS deaths per 100,000 people

.206**

.208***

.233***

Life expectancy

.694***

Northern America & Western Europe

.064

Sub-Saharan Africa
Adjusted r-square (n)

.803***
.494 (181)

.524 (181)

.522 (181)

.565 (181)

p<0.001 for all models: *p<0.10, **p<0.05, ***p<0.01

deaths and variables measuring reproduction, bivariate correlations show that HIV/
AIDS is positively correlated with both adolescent birth rate (r=0.290, p<0.001,
n=181) and total fertility (r=0.265, p<0.001, n=182). The inclusion of a dummy
variable for the northern America and western Europe region shows that variation in
adolescent birth rate and total female fertility within this region is attributable to life
expectancy and HIV/AIDS, with GDP having no significant impact on the
relationship. This is also observed in East Asia and the Pacific, and in Latin
America and the Caribbean for total female fertility only. However, the inclusion of a
dummy variable for sub-Saharan Africa shows that other factors specific to this
region account for variation in adolescent birth rate and total female fertility. This
result is also observed in the Arab states, Central Asia, central and eastern Europe,
Southwest Asia, and Latin America and the Caribbean for adolescent birth rate only.
Factors that may account for regional variation include cultural values placed on
childbearing and children, the need for children to work the land or in the household,
religious values, womens access to the labor market, and lack of access to or
utilization of contraception.
In Table 4, variables for education (adult literacy rate, overall and secondary school
enrolment ratios) were added to the models. Education variables contribute significantly
to the relationship between life expectancy and total female fertility. With the addition of
regional dummy variables, we see that variation in total female fertility when controlling
for adult literacy rate is attributable to region-specific factors in sub-Saharan Africa,
central and eastern Europe, Latin America and the Caribbean, and Southwest Asia.
Region-specific factors also account for variation in total female fertility when
controlling for overall and secondary school enrolment ratios in sub-Saharan Africa,
the Arab states, Central Asia, central and eastern Europe, and Southwest Asia.
Adult literacy rate and overall and secondary school enrolment ratios are all
associated with life expectancy, even when controlling for GDP, HIV/AIDS, and the two
regions (Table 5). Although multivariate models including life expectancy suggest a
positive relationship between HIV/AIDS deaths and variables measuring education,
bivariate correlations indicate that HIV/AIDS is negatively correlated with education
variables (adult literacy rate: r=0.171, p=0.022, n=179; overall enrolment: r=
0.269, p<0.001, n=180; secondary school enrolment: r=0.320, p<0.001, n=164).

.737 (180)

.750 (164)

p<0.001 for all models: *p<0.10, **p<0.05, ***p<0.01

.731 (182)

.235

.754 (182)

.787 (179)

.409***

.410***

.034

.853***

Adjusted rsquare (n)

.729 (182)

.107***

.392***

.035

.281***

.123**

1.090***

.696***

.644***

.646 (188)

.383***

.334***

.101*

.915***

Sub-Saharan
Africa

Northern
America &
Western
Europe

Secondary
school
enrolment
ratio

Overall school
enrolment
ratio

Adult literacy
rate (%)

.226***

.389***

HIV/AIDS
deaths per
100,000
people

.741***

.074

1.081***

.082

.803***

Gross
Domestic
Product
(GDP)

Life
expectancy

Total female fertility

.788 (179)

.154

.378***

.231***

.043

.723***

.800 (179)

.459***

.345***

.258***

.042

.588***

.742 (180)

.296*

.203***

.344***

.044

.912***

.760 (180)

.600***

.162**

.362***

.054

.727***

.752 (164)

.195

.404***

.285***

.081

.708***

.764 (164)

.473***

.366***

.308***

.089*

.574***

Table 4 Multivariate linear regression models showing the association of standardized regression coefficients for life expectancy, education, and covariates (GDP and HIV/
AIDS deaths per 100,000 people, dummy variables for the regions Northern America & Western Europe and Sub-Saharan Africa) with total female fertility

Hum Nat

.486 (186)

.693***

.964***

.586 (179)

.428***

.013

.973***

.586 (179)

.220

.431***

.030

p<0.001 for all models: *p<0.10, **p<0.05, ***p<0.01

Adjusted r-square
(n)

Sub-Saharan
Africa

Northern America
& Western Europe

HIV/AIDS deaths
per 100,000
people

Gross Domestic
Product (GDP)

Life expectancy

Adult literacy rate (%)

.778***

.757***

.883***

.304***

.132**

.871***

.291*

.299***

.074

.602 (179) .565 (190) .655 (180) .659 (180)

.531***

.446***

.024

Overall school enrolment ratio

.806***

.857***

.981***

.283***

.113*

.656 (180) .654 (174) .730 (164)

.219

.312***

.148**

.729 (164)

.083

.285***

.130*

.985***

Secondary school enrolment ratio

.737 (164)

.374**

.295***

.136**

.852***

Table 5 Multivariate linear regression models showing the association of standardized regression coefficients for life expectancy and covariates (GDP and HIV/AIDS deaths per
100,000 people, dummy variables for the regions Northern America & Western Europe and Sub-Saharan Africa) with adult literacy, and overall and secondary school enrolment
ratios

Hum Nat

p<0.001 for all models: *p<0.10, **p<0.05, ***p<0.01

.259 (44)

.639 (44)

.472 (35)

.671 (44)

Adjusted r-square
(n)

.416 (44)

.849***

.082

Secondary school
enrolment ratio

.496***

.183**

Adult literacy rate .474***

(%)

Overall school
enrolment ratio

.394***

.347***

HIV/AIDS deaths
per 100,000
people

.063

.696 (35)

.355*

.306***

.534

.754***

.501 (137)

.492***

.496 (135)

.490***

.201

.008

.358***

.372 (142)

.191**

.561***

.351 (136)

.206**

.284

.020

.571***

.356***

.994***
.316

.326*

.183

.947***

.407**

Gross Domestic
Product (GDP)

Life expectancy

Total female fertility where life expectancy 60 years

Total female fertility where life expectancy <60 years

.436 (135)

.355***

.428***

.421 (129)

.383***

.227

.047

.445***

Table 6 Multivariate linear regression models showing the association of standardized regression coefficients for life expectancy, education, and covariates (GDP and HIV/
AIDS deaths per 100,000 people) with total female fertility separated by life expectancy thresholds (greater or less than 60 years of age)

Hum Nat

Hum Nat

The inclusion of dummy variables for all regions shows that variation in indicators
for education (adult literacy rate, overall and secondary school enrolment) is
attributable to life expectancy and HIV/AIDS. However, the inclusion of a dummy
variable for sub-Saharan Africa shows that in addition to life expectancy and HIV/
AIDS, other factors specific to the region account for variation in adult literacy.
This result is also observed in the Arab States, Central Asia, central and eastern
Europe, and Southwest Asia. In addition, other, unknown factors account for
variation in overall school enrolment in the Arab states and Central Asia, and for
variation in secondary school enrolment in sub-Saharan Africa, Central Asia, and
central and eastern Europe.
In models that include a dummy variable for the sub-Saharan Africa region, the
relationships among reproduction, education, and life expectancy (excluding gross
secondary school enrolment), HIV/AIDS deaths, and other unmeasured factors
contribute significantly whilst GDP is only marginally important. (For further details
on the influence of HIV/AIDS in the sub-Saharan Africa region, see Anderson
2010). In contrast, models containing a dummy variable for northern America and
western Europe are representative of the combined data set, displaying the
significant impact of life expectancy on the relationships but limited if any impact
of GDP and/or other unmeasured factors.
Table 6 displays models where life expectancy, adult literacy, overall and
secondary school enrolment ratios, and covariates are added as independent
variables to determine the combined effect on total female fertility separately for
countries with life expectancies below or above 60 years of age. For countries with
life expectancies below 60 years, life expectancy in combination with HIV/AIDS is
highly predictive of total female fertility. The education variables, overall and
secondary school enrolment ratio, were significant in the models only when the
impact of HIV/AIDS was not accounted for. This suggests that HIV/AIDS has a
greater impact than education on the relationship between life expectancy and total
female fertility in countries with lower life expectancies. In countries with life
expectancy 60 years, variables for education remain significant even with the
addition of GDP and HIV/AIDS, which were not significant in these models. These
results support the findings in Table 2 that a threshold level of life expectancy may
be necessary for changes in total female fertility and/or education.

Discussion
The results presented here support our hypotheses that life expectancy is associated
with educational investments and the timing and extent of reproduction. Increases in
life expectancy correspond to decreases in reproductive rate and total number of
offspring, and increases in educational attainment. From this we surmise that when
life expectancy is high, educational attainment is also high, reproductive timing is
delayed, and overall reproduction reduced. As adolescent birth rate increases, adult
literacy and the overall school enrolment ratio decrease (r=0.717, p<0.001 and r=
0.624, p<0.001, respectively). The same strong inverse relationship exists between
total female fertility and education (adult literacy: r=0.812, p<0.001; overall
school enrolment ratio: r=0.744, p<0.001).

Hum Nat

Variation in reproduction and education measures across regions suggests that

although the relationship between short life expectancies and early and high fertility
holds at multiple scales, differences exist at local levels (Wilson and Daly 1997;
Geronimus 1996a, b, 1997, 2001). These variations may be the result of sociocultural
differences, the influence of economic changes, or, as suggested by Low et al. (2008),
the result of sudden changes in life expectancy (resulting from diseases, such as HIV,
or warfare). In addition, stress created by a heavy disease burden or by conflict may
also be a predictor of earlier age at menarche and at first birth (e.g., Chisholm et al.
2005). The multivariate models presented above show that when controlling for the
influences of economic factors (GDP) and disease stressors (HIV/AIDS deaths), the
relationships among life expectancy, reproduction, and education hold. GDP and HIV/
AIDS explain variation in life expectancy in some regions (the Arab states, East Asia
and the Pacific, and sub-Saharan Africa), but not all.
Uganda offers one example of the impact of disease on life expectancy and
investments in education and reproduction. Uganda was one of 16 countries
experiencing a decline in life expectancy between 1975 and 2000, mostly a result of
high death rates from HIV/AIDS (it was ranked fifteenth of the 193 UN nation-states).
In 2006, life expectancy was 27.5 years lower than that observed in the US; adolescent
fertility rate was more than 250% higher than in the US, with only 24% of women using
some form of contraception (2004 estimate, United Nations). However, the overall
school enrolment ratio reached 62.3%, with 72.6% of the adult population being literate.
In contrast, Afghanistan has experienced no change in life expectancy but has
experienced a persistent stressful environment owing to prolonged conflict. Indeed,
Afghanistans indicators for life expectancy, reproduction, and education are outliers
compared with the other nine countries in the UNESCOs South and West Asia
region (Table 7 and Fig. 1). Although social indicators were low even before the
Soviet invasion in 1979, the prolonged conflict has left Afghanistan severely
impoverished. Gross enrolment in education differs greatly by gender. The primary
school enrolment ratio for males was 1.6 times greater than for females, and the
secondary school enrolment ratio was 2.8 times greater. Half of the men and 85% of
women are illiterate. Lack of proper housing, adequate schools, rural roads,
communication, electricity, water, and sanitary conditions impedes improvements
in health and contributes to low life expectancies.
Countries with high life expectancy generally experience greater political and
economic stability, have greater economic success as measured by GDP, and are
impacted less by diseases affecting adult survival, such as HIV/AIDS. Offspring
success in such countries is dependent on acquired skills needed to compete
economically, requiring considerable parental time and financial investment (Kaplan
et al. 2000). As a result, more opportunities are available that reward educationally
acquired skills and knowledge (Handwerker 1986). This compels a greater proportion
of women to devote additional years to educational attainment, marry later, work
before and after marriage, and increase their resource base (e.g., Low 2000a, b, 2005).
As our data show, countries with higher life expectancies (60 years) have higher
school enrolment ratios at every level of education, higher adult literacy rates, and
lower adolescent birth rates and total female fertility. For these countries, relationships
among indicators for education, reproduction, and life expectancy are significant.

Hum Nat

In contrast, countries with low life expectancy have weaker government

infrastructures and economies that are unable to respond to the needs of their
populations, making them susceptible to civil warfare, natural disasters, and disease,
and creating stressful environments. Although greater educational attainment
improves health outcomes (Appiah and McMahon 2002), unstable environments
support fewer opportunities that reward educationally acquired skills, thereby
fostering low educational investments and early and frequent reproduction. As
stated in the Human Development Report, increased exposure to environmental
stress is holding back the worlds poor to build a better life for themselves and their
children (United Nations 2007:1). Our data show that in countries with a life
expectancy below 60 years, school enrolment ratios and adult literacy rates are
lower, and adolescent birth rates and total female fertility are higher than in countries
with life expectancies above 60 years. However, in contrast to the latter countries, in
countries where life expectancy is below 60 years relationships among education
indicators, adolescent fertility rate, and life expectancy are not statistically
significant.
Within-country variation undoubtedly exists that we cannot account for given the
national-level data used. In addition, other factors influencing control over female
fertility, including access to and acceptance and utilization of modern contraceptives,
and religious and cultural beliefs, likely account for variation between and within
countries. However, and consistent with other research, global and regional
comparisons are highly suggestive of the predictive value of life expectancy on
educational attainment (Camargos et al. 2007; Mackenbach et al. 1997; Meara et al.
2008; Silventoinen and Lahelma 2002; Valkonen et al. 1997) and reproduction
(Chisholm et al. 2005; Low et al. 2008).
By isolating the component measures of HDI and using country-level data we
have demonstrated that although cross-country variation exists, there are strong
interactions among life expectancy, reproductive investments, and educational
attainment, and these interactions occur independently of economic pressures and
disease burdens. As argued by others (Caldwell 1980; Gant et al. 2009; Mamdani
1972; Maternowska 2006), our results suggest that increases in overall knowledge
improve quality of life, access to resources and opportunities, and average life
expectancy, and decrease and delay energy invested in reproduction. However, basic
mass education at the primary level alone may not influence fertility transitions. As
the feedback loop proposed by Hill and Kaplan (1999) suggests, changes in
opportunity structure resulting from improvements in life expectancy and basic
education further reward educationally acquired skills and perspectives, promoting
secondary and tertiary education, which further increases life expectancy and lowers
reproduction. Therefore, given their interactive effects, policies aimed at influencing
education and reproductive decisions should consider environmental characteristics
that drive peoples expectations about their longevity.

Acknowledgments We thank Ruthanne Marcus, Merrill Singer, Katherine Hsu, Anna Layton, and Paul
Swartwout for valuable comments on earlier drafts of this manuscript. A special thanks to Sowmya Rao
for all her statistical guidance and continued support.

Moldova

Romania

Russian Federation

Serbia

Slovakia

Slovenia

Turkey

Ukraine

Oman

Palestinian Autonomous
Territories

Qatar

Saudi Arabia Sudan

Syria

Tunisia

United Arab Emirates

Yemen

Dominican Republic

Dominica

Cuba

Costa Rica

Colombia

Chile

Brazil

Bolivia

Bermuda

Belize

Barbados

Bahamas

Argentina

Antigua and Barbuda

Palau

New Zealand

Nauru

Myanmar

Guyana

Guatemala

Grenada

El Salvador

Micronesia (Fed. States Ecuador

of)

Marshall Islands

Malaysia

Laos

Korea (South)

Poland

Japan
Kiribati

Lithuania Montenegro Korea (North)

Hungary

Indonesia

Hong Kong SAR

Morocco

Uzbekistan

Lebanon

Estonia

Czech Rep.

Mauritania

Turkmenistan

Kuwait

Fiji

China

Cambodia

Brunei Darussalam

Australia

East Asia & the Pacific Latin America & the

Caribbean

Latvia

Tajikistan

Jordan

Croatia

Bulgaria

Bosnia &
Herzegovina

Belarus

Albania

Central & Eastern

Europe

Libya

Kazakhstan

Georgia

Djibouti

Kyrgyzstan
Mongolia

Azerbaijan

Bahrain

Iraq

Armenia

Algeria

Egypt

Central Asia

Arab States

Table 7 United Nations 193 nation-states included in the analyses by UNESCO world region

Appendix

Spain

Portugal

Netherlands Norway

Monaco

Malta

Luxembourg

Italy

Israel

Ireland

Iceland

Greece

Germany

France

Finland

Denmark

Cyprus

Canada

Belgium

Austria

Northern America & Western

Europe

Sri Lanka

Pakistan

Nepal

Maldives

Iran

India

Bhutan

Bangladesh

Afghanistan

Southwest
Asia

Kenya

Guinea-Bissau

Guinea

Ghana

Gambia

Gabon

Eritrea, Ethiopia

Equatorial Guinea

DRC

Cte dIvoire

Congo

Comoros

CAR, Chad

Cape Verde

Cameroon

Burundi

Burkina Faso

Botswana

Angola, Benin

Sub-Saharan
Africa

Hum Nat

Arab States

Table 7 (continued)

Central Asia

Sao Tome &

Principe
Senegal
Seychelles
Sierra Leone
Somalia

St. Vincent & the

Grenadines
Suriname
Trinidad & Tobago
Uruguay
Venezuela

Zimbabwe

Zambia

Tanzania

Uganda

Togo

Swaziland

South Africa

Nigeria

Niger

Namibia

Mozambique

Mauritius

Mali

Malawi

Madagascar

Liberia

Lesotho

Sub-Saharan
Africa

Rwanda

Viet Nam

Southwest
Asia

St. Lucia

Peru

Vanuatu

United States

United Kingdom

Switzerland

Sweden

Northern America & Western

Europe

St. Kitts & Nevis

Panama

Nicaragua

Thailand
Paraguay

Netherlands Antilles

Solomon Islands

Tonga

Mexico

Singapore

Timor-Leste

Honduras
Jamaica

Samoa

Haiti

Papua New Guinea

Yugoslav Rep.
(former)
Philippines

East Asia & the Pacific Latin America & the

Caribbean

Central & Eastern

Europe

Hum Nat

Hum Nat

References
Anderson, K. G. (2010). Life expectancy and the timing of life history events in developing countries.
Human Nature, 21, 103123.
Appiah, E. N., & McMahon, W. W. (2002). The social outcomes of education and feedbacks on growth in
Africa. Journal of Development Studies, 38(4), 2768.
Brumbach, B. H., Figueredo, A. J., & Ellis, B. J. (2009). Effects of harsh and unpredictable environments
in adolescence on the development of life history strategies: a longitudinal test of an environmental
model. Human Nature, 20, 204268.
Caldwell, J. C. (1980). Mass education as a determinant of the timing of fertility decline. Population and
Development Review, 6, 225255.
Camargos, M. C., Machado, C. J., & do Nascimento Rodrigues, R. (2007). Disability life expectancy for
the elderly, city of Sao Paulo, Brazil, 2000: gender and educational differences. Journal of Biosocial
Sciences, 39(3), 455463.
Chisholm, J. S., Quinlivan, J. A., Petersen, R. W., & Coall, D. A. (2005). Early stress predicts age at
menarche and first birth, adult attachment, and expected lifespan. Human Nature, 16, 233265.
Dorling, D., Shaw, M., & Smith, G. D. (2006). Global inequality of life expectancy due to AIDS. British
Medical Journal, 332, 662664.
Eisenberg, J. F. (1981). The mammalian radiations. Chicago: University of Chicago Press.
Gant, L., Heath, K. M., & Ejikeme, G. G. (2009). Early motherhood, high mortality, and rates of HIV/
AIDS in sub-Saharan Africa. Social Work in Public Health, 24(1/2), 3946.
Geronimus, A. T. (1996a). Black/White differences in the relationship of maternal age to birthweight: a
population-based test of the weathering hypothesis. Social Science & Medicine, 42, 589597.
Geronimus, A. T. (1996b). What teen mothers know. Human Nature, 7, 323352.
Geronimus, A. T. (1997). Teenage childbearing and personal responsibility. Political Science Quarterly,
112, 405430.
Geronimus, A. T. (2001). Inequalities in life expectancy, functional status, and active life expectancy
across selected black and white populations in the United States. Demography, 38, 227251.
Geronimus, A. T. (2004). Teenage childbearing as cultural prism. British Medical Bulletin, 69, 155166.
Gould, S. J. (1977). Ontogeny and phylogeny. Cambridge: Harvard University Press.
Handwerker, W. P. (1986). The modern demographic transition: an analysis of subsistence choices and
reproductive consequences. American Anthropologist, 88, 401417.
Hill, K., & Hurtado, A. M. (1996). Ache life history: The ecology and demography of a foraging people.
New York: Aldine de Gruyter.
Hill, K., & Kaplan, H. (1999). Life history traits in humans: theory and empirical studies. Annual Review
of Anthropology, 28, 397430.
Hill, E. M., Thomson, R. L., & Low, B. S. (1997). The role of future unpredictability in human risktaking. Human Nature, 8, 287325.
Kaplan, H. S., & Lancaster, J. B. (2003). An evolutionary and ecological analysis of human fertility,
mating patterns, and parental investment. In K. W. Wachter & R. A. Bulatao (Eds.), Offspring:
Human fertility behavior in biodemographic perspective (pp. 170223). Washington: National
Academies Press.
Kaplan, H., Hill, K., Lancaster, J., & Hurtado, A. M. (2000). A theory of human life history evolution:
diet, intelligence, and longevity. Evolutionary Anthropology, 9, 156185.
Kaplan, H., Hooper, P., & Gurven, M. (2009). The evolutionary and ecological roots of human social organization.
Philosophical Transactions of the Royal Society. Series B, Biological Sciences, 364, 32893299.
Knodel, J., Havanon, N., & Sittitrai, W. (1990). Family size and the education of children in the context of
rapid fertility decline. Population and Development Review, 16, 3162.
Koons, D. N., Metcalf, C. J. E., & Taljapurkar, S. (2008). Evolution of delayed reproduction in uncertain
environments: a life-history perspective. The American Naturalist, 172, 797805.
Lam, D., & Duryea, S. (1999). Effects of schooling on fertility, labor supply, and investments in children,
with evidence from Brazil. The Journal of Human Resources, 34(1), 160192.
Lamptey, P. R., Johnson, J. L., & Khan, M. (2006). The global challenge of HIV and AIDS. Population
Bulletin, 61, 324.
Lerner, R. (1984). On human plasticity. New York: Cambridge University Press.
Low, B. (2000a). Sex, wealth, and fertilityold rules, new environments. In L. Cronk, N. A. Chagnon, &
W. G. Irons (Eds.), Adaptation and human behavior: An anthropological perspective (pp. 323344).
New York: Aldine de Gruyter.

Hum Nat
Low, B. (2000b). Why sex matters: A Darwinian look at human behavior. Princeton: Princeton University Press.
Low, B. (2005). Womens lives there, here, then, now: a review of womens ecological and demographic
constraints cross-culturally. Evolution and Human Behavior, 26, 6487.
Low, B. S., Hazel, A., Parker, N., & Welch, K. B. (2008). Influences on womens reproductive lives:
unexpected ecological underpinnings. Cross-Cultural Research, 42, 201219.
Mackenbach, J. P., Kunst, A. E., Cavelaars, A. E., Groenhof, F., & Geurts, J. J. (1997). Socioeconomic
inequalities in morbidity and mortality in western Europe. The EU Working Group on Socioeconomic
Inequalities in Health. Lancet, 349, 16551659.
Mamdani, M. (1972). The myth of population control: family, caste, and class in an Indian village. New
York: Monthly Review Press.
Maternowska, M. C. (2006). Reproducing inequities: Poverty and the politics of population in Haiti.
Rutgers: Rutgers University Press.
Meara, E. R., Richards, S., & Cutler, D. M. (2008). The gap gets bigger: changes in mortality and life
expectancy, by education, 19812000. Health Affairs, 27, 350360.
Preston, S. (1975). The changing relation between mortality and level of economic development.
Population Studies, 29, 231248.
Promislow, D., & Harvey, P. (1990). Living fast and dying young: a comparative analysis of life-history
variation among mammals. Journal of the Zoological Society of London, 220, 417437.
Quinlan, R. J. (2007). Human parental effort and environmental risk. Proceedings of the Royal Society.
Series B, Biological Science, 274, 121125.
Rodgers, G. B. (1979). Income and inequality as determinants of mortality: an international cross-section
analysis. Population Studies, 33, 34351.
Roff, D. (2002). Life history evolution. Sunderland: Sinauer.
Schaffer, W. M. (1974). Optimal reproductive effort in fluctuating environments. The American Naturalist,
108, 783790.
Schultz, T. P. (1993). Returns to womens education. In E. M. King & M. A. Hill (Eds.), Womens education in
developing countries: Barriers, benefits, and policies. Baltimore: Johns Hopkins University Press.
Sen, A. (1993). The economics of life and death. Scientific American, 266, 4047.
Silventoinen, K., & Lahelma, E. (2002). Health inequalities by education and age in four Nordic countries,
1986 and 1994. Journal of Epidemiology and Community Health, 56, 253258.
Stearns, S. C. (1992). The evolution of life histories. Oxford: Oxford University Press.
United Nations (1999). Table 1: World population from year 0 to stabilization. The world at six billion. Available
at http://www.un.org/esa/population/publications/sixbillion/sixbilpart1.pdf. Accessed October 17, 2009.
United Nations Development Programme. (2007). Human development report 2007/2008. Fighting
climate change: Solidarity in a divided world. New York: Palgrave Macmillan. Retrieved from http://
hdr.undp.org/en/media/HDR_20072008_EN_Complete.pdf. Accessed November 3, 2009.
Valkonen, T., Sihvonen, A. P., & Lahelma, E. (1997). Health expectancy by level of education in Finland.
Social Science & Medicine, 44, 801808.
Wilson, E. O. (1975). Sociobiology: The new synthesis. Cambridge: Harvard University Press.
Wilson, M., & Daly, M. (1997). Life expectancy, economic inequality, homicide, and reproductive timing
in Chicago neighbourhoods. British Medical Journal, 314, 12711274.
World Health Organization. (2008). World health report 2008: Primary health care now more than ever.
Geneva: WHO.
World Health Organization. (2009). World health statistics: Indicator compendium, unedited version.
Geneva: WHO. Available at http://www.who.int/whosis/indicators/WHS09_IndicatorCompendium_20090701.pdf. Accessed November 16, 2009.
Nicola Bulled is a PhD candidate in anthropology at the University of Connecticut. She has conducted
epidemiologic research on HIV/AIDS and tuberculosis in developing and developed countries. She has an
MPH from Boston University and a BS in Biological Sciences from Colorado State University. Her
primary interests are examining the social and structural factors shaping health.
Richard Sosis is an associate professor of anthropology and director of the Evolution, Cognition, and Culture
Program at the University of Connecticut. His research interests include the evolution of cooperation, utopian
societies, and the behavioral ecology of religion. To explore these issues, he has conducted fieldwork with
remote cooperative fishers in the Federated States of Micronesia and with various communities throughout
Israel, including ultra-Orthodox Jews and members of secular and religious kibbutzim.