Full text of DOI:10.1177/106939702237306

1CB0ar.or1sb1se7-rC7//u1Pl0tAu6R9ra3El9NR7T0e2Ase2La3r7IcN3h0V6/ENSoTvMemEbNeTr 2A0N02D WEALTH
Parental Investment, Human
Capital, and Cross-National
Differences in Wealth
Nigel Barber
Portland, Maine
Economists have recently turned to family variables in search of ex-
planations for why the economies of some countries grow more
rapidly than others. This research tested the hypothesis that cross-
national indicators of parental investment and human capital
would be predictive of GNP per capita of 147 countries. Regression
analysis found that close to two thirds of the variance in GNP was
explained by parental investment factors (including total fertility
rates, the adult sex ratio, polygyny, contraception, and school atten-
dance rates). The effects of parental investment/human capital on
wealth remained after controlling for climate, geography, endemic
diseases, religion, and government policies, even though each of
these was predictive of GNP. The data support the hypothesis that
wealth differences between countries are associated with human
capital differences accruing to differential parental investment.
Geography is one of the most important determinants of a coun-
try’s wealth. Thus, Hall and Jones (1996) found that countries hav-
ing temperate climates are more prosperous than those lying close
to the equator. Kenny (1999) also found that countries having
wealthy neighbors were much wealthier than those whose neigh-
bors were poor. Landes (1998) has suggested that greater economic
Cross-Cultural Research, Vol. 36 No. 4, November 2002 338-361
DOI: 10.1177/106939702237306
© 2002 Sage Publications
338
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
339
development for countries at higher latitudes,particularly those of
Europe, reflects climate, endemic disease, and the availability of
arable land, as well as technological innovation and other factors,
including a strong work ethic. This article tested the role of paren-
tal investment and human capital in predicting cross-national dif-
ferences in wealth while controlling for the important effects of ge-
ography, climate, disease, government policies, and religion.
Climate may be particularly important for economic growth
because it influences the health and well-being of individuals with
consequences for their desire to work. Consistent with this view,
many of the underdeveloped nations lie in the tropics, which have
an unpleasantly humid climate that discourages effort. Yet, sev-
eral researchers have reported that drought conditions reduce eco-
nomic growth (Kenny, 1999). The tropics also have many debilitat-
ing endemic diseases such as malaria, which has measurable
consequences for affected countries in terms of health care costs,
mosquito control expenses, reduced agricultural productivity and
lost labor (Banfield, 1998; Liese, 1998). Gallup, Sachs, and
Mellinger (1999) found that their malaria index (fraction of land
area subject to malaria and fraction of malaria cases that are of the
more severe P. falciparum type) was more strongly predictive of
national differences in GDP than any of 10 geographical, political,
and economic variables they tested. The malaria index was more
weakly predictive of economic growth.
In addition to climate and the related variable of endemic dis-
ease, there are at least five other categories of variables that might
be important in determining economic development. They include
but are not limited to the following:
• Geographic factors such as the availability of arable land and the
presence of a coastline and natural harbors suitable for shipping. Gal-
lup et al. (1999) found that countries were wealthier if more of the
population lived close to coastlines, if the capital city was closer to in-
ternational centers of trade, and if shipping costs were relatively low.
(Interestingly, natural mineral wealth seems not to be an important
factor given that abundance of natural resources was found to be neg-
atively correlated with economic growth; Sachs & Warner, 1997.)
High agricultural productivity was important in the case of early “hy-
draulic” civilizations that were based on extensive irrigation and
drainage systems that permitted food surpluses to be produced. An-
cient Greek civilization did not have a basis in arable land but was
founded on maritime trading, suggesting the importance of having a
coastline or access to ports (McClellan & Dorn, 1999).
Downloaded from ccr.sagepub.com by guest on August 12, 2015

340
Cross-Cultural Research / November 2002
• Most economists assume that government policies have important
consequences for country differences in wealth. Thus Communist
countries are generally supposed to have a lower rate of economic
growth and a lower standard of living. Gallup et al. (1999) found
that a country’s openness to trade and the quality of its governmental
institutions were each predictive of GDP and economic growth. In-
terestingly, economists are quite divided on whether government-
managed economies can grow as quickly as unregulated ones (Feng,
1996). Government policies, such as excessive foreign borrowing and
disproportionate military expenditures may also siphon off capital
that is necessary for economic development (Kick, Davis, Kiefer, and
Burns, 1998). Kenny (1999) found that the number of years of political
independence of a state was associated with its GNP per capita.
• Religion is another factor that sociologists have emphasized in ex-
plaining economic growth. Max Weber, for example, pointed to a con-
nection between the individualism associated with Protestantism
and the rise of entrepreneurial activity in the context of the Industrial
Revolution (Mook, 1987; Weber, 1930). Contrary to Weber’s (1930)
hypothesis, recent research has found that individualism in a country
is negatively correlated with GDP (Fedderke & Klitgaard, 1998) al-
though the number of the countries in the test was not reported and
could have been as few as 24, or as many as 118. Moreover, Swank
(1996) reported that communitarian politics is positively correlated
with economic growth for 25 countries. Early empirical research also
showed that Catholic countries in Europe had much slower economic
growth, using only one fifth as much electricity, for example, com-
pared to Protestant ones (Woytinski & Woytinski, 1953). Fedderke
and Klitgaard (1998) also reported that the proportion of a country’s
population that was Christian was positively correlated with GDP
but not with economic growth rates. Traditional religions of Africa
and elsewhere emphasize supernatural determination of good for-
tune (and ill fortune) and might thus be seen as militating against en-
trepreneurial effort, which is generally predicated on the belief that
businessenterprisecanberewardedbyeconomicsuccess. Islamictra-
dition rejects the lending of money for profit (or “usury”), which might
reduce the availability of capital for starting businesses (Graham,
1987).
• Technological innovations play an important role in industrializa-
tion. Thus, the Industrial Revolution in England was based on a novel
technology for mechanization of fabric production (Landes, 1998;
McClellan & Dorn, 1999). Innovations may diffuse rapidly in the mod-
ern world, however, so that their role in explaining cross-national dif-
ferences in economic growth is questionable. Innovation is also diffi-
cult to quantify. For these reasons, the possible role of scientific and
technological innovation were not investigated in this study.
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
341
• Recently, the theory of economic development has focused on human
capital as an important influence on economic growth (Fedderke &
Klitgaard, 1998; Healy, 1998; Kick, Davis, Lehtinen, & Wang, 2000).
Human capital can be loosely defined as the set of skills and abilities
acquired by individuals by virtue of rearing and education that helps
them to play a productive economic role as workers, managers, or en-
trepreneurs (Healy, 1998). Human capital can be increased by gov-
ernment investment in public education as well as by the efforts of
parents in socializing their children (Coleman, 1990).
Despite a considerable amount of empirical investigation into
the factors affecting economic growth, economists have achieved
surprisingly little consensus (Kindleberger, 1996; Porter, 1998;
Landes, 1998; Rosenberg & Birdzell, 1987). For example, research
on the role of democracy in economic growth has found strong posi-
tive effects,weak positive effects,and negative effects (Feng,1996).
It is possible that this inconsistency is due to the choice of depend-
ent variable, which has generally been restricted to the annual
growth rate. One problem with this practice is that annual growth
rates can fluctuate according to various economic cycles that have
little or no bearing on cross-national differences in wealth.
Another problem is that relative growth rates arguably exaggerate
the growth of poor countries while minimizing the growth of large
economies. Inappropriate measurement could explain some of the
inconsistencies in the literature. The use of GNP may be advanta-
geous because it cumulates the impact of a number of influences on
economic growth over time and therefore allows greater reliability
for investigating cross-national differences in economic develop-
ment, and that assumption guided the choice of dependent vari-
able in this study. Some economists prefer to study national pro-
ductivity using purchasing power parity estimates that correct
GNP for the purchasing power of local currencies. Others, such as
O’Hearn (1998),see GNP as a more accurate indication of standard
of living in a country, however, because purchasing power esti-
mates can be somewhat arbitrary and are vulnerable to gross dis-
tortion in the event of currency devaluations. From an empirical
perspective, the measures are so highly correlated that it cannot
make much difference which one is used (see the Discussion
section).
Another important point is that much of the literature has
focused on governmental influences, possibly paying too little
attention to family variables that can affect economic development
Downloaded from ccr.sagepub.com by guest on August 12, 2015

342
Cross-Cultural Research / November 2002
by altering human capital. Several recent studies have shown that
family variables relevant to human capital can produce larger
effects than those typically produced by governmental variables in
predicting cross-national differences in economic growth
(Fedderke, & Klitgaard, 1998; Kick et al. 2000; Swank, 1996). The
research reported here focused on parental investment and human
capital while treating the other explanatory categories as control
variables.
PARENTAL INVESTMENT AND HUMAN CAPITAL
Parental investment was defined by Trivers (1972) as parental
efforts that increase the chance of survival for one offspring at the
expense of the parent’s capacity to invest in other offspring. This
definition applies to all species with extensive parental care and
thus includes humans. According to Trivers’ definition, human
parental investment would be reduced in large families by close
spacing of births and by single parenthood. Parental effort
includes food, energy, time investment, vigilance, work, money,
social influence, emotional support, and any other finite parental
resource that contributes to the survival, or reproductive success,
of children (Barber, 2000a). Even though older children may con-
tribute resources to a family economy by gathering food,in the case
of hunter-gatherers, for example, there is always a net flow of
resources away from parents to children, even when children
approach maturity (Kaplan, 1994). In that sense, parental invest-
ment is an empirically verifiable phenomenon and the unidirec-
tional flow of resources is consistent with the predictions of evolu-
tionary biology because parental behavior was designed by natural
selection to enhance the survivorship, and reproductive success, of
offspring.
The main economic significance of variation in parental invest-
ment is that high parental investment increases human capital,as
discussed below, so that economic development is more likely in
those countries where parental investment is highest.
Assuming that high parental investment increases the eco-
nomic success of children, why do parents not always invest at a
very high level in each of their offspring? One probable reason is
high rates of child mortality in some countries.Under conditions of
high survival uncertainty, parents would be expected to pursue a
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
343
strategy of investing relatively little in a large number of offspring,
rather than investing a lot in a small number.
Developmental psychologists have proposed an evolutionary
theory of socialization according to which more difficult and uncer-
tain economic conditions are conducive to lower levels of parental
investment and a harsh, insensitive socialization process. Accord-
ing to the theory, children respond to a difficult early environment
by developing a manipulative interpersonal orientation with
delinquent tendencies, rather unrestricted sexual expression for
both sexes, and low occupational striving (Belsky, Steinberg, and
Draper, 1991). Reduced parental investment is adaptive because it
helps children to succeed in an environment with little potential
for economic progress. This theory can be easily modified to
account for socialization differences as a function of parental
income, country differences, and ecological differences such as the
potential for intensive agriculture (Barber, 2000b).
In economically developed societies, family size is constrained
by many factors, including limitation of parental time budgets by
occupational responsibilities and the lack of “free” child care ser-
vices provided by extended families and village networks. The
expense of raising children is also much greater in economically
developed countries for a variety of reasons including the costs of
modern medicine, housing, nutrition, education, and so forth, ser-
vices that are largely provided outside the monetary economy in
underdeveloped nations. Thus, in 1999, American families with
two children spent an average of $9,940 in expenses for the youn-
ger child (Lino, 1999). Such high costs might partly explain why
fertility is so much lower in developed countries (Population Refer-
ence Bureau, 1998).
It seems clear that developed countries provide an environment
that favors unusually high parental investment and that this
increases human capital, as reflected, for example, in large num-
bers of people receiving a lengthy education (Barber, 2000b). There
is good reason to believe that high parental investment also
increases achievement motivation with important implications for
career striving, entrepreneurial activity, and economic growth.
Some of the many types of evidence supporting this conclusion
include studies of language development as a function of social sta-
tus of parents (Hart & Risley, 1995), research on intelligence as a
function of parental stimulation that varies with income level
(Ramey, 1992), data on variation in achievement motivation of
Downloaded from ccr.sagepub.com by guest on August 12, 2015

344
Cross-Cultural Research / November 2002
children as a function of parental income (Argyle,1994),and obser-
vational evidence that parents of ambitious children support their
efforts to solve problems whereas less ambitious children have
parents who increase fear of failure (reviewed in Barber, 2000b).
Social psychologists have found that need for achievement
increases with parental affluence possibly because ambitiousness
helps children to exploit opportunities for social mobility. By con-
trast, children growing up in poverty or in underdeveloped coun-
tries have less possibility for advancement and tend to be less
ambitious and thus less likely to start businesses (McClelland,
Atkinson, Clark, & Lowell, 1976).
These societal differences are well illustrated by the contrast
between the United States, a highly entrepreneurial society, and
traditional India, in which a hereditary caste system diminished
opportunities for social progress. Social psychologist David
McClelland clarified this difference in his work on the develop-
ment of entrepreneurial spirit among Indian businessmen. Mc-
Clelland knew that entrepreneurs are highly ambitious people
because they scored high on his need for achievement scale. He
believed that if business people from an underdeveloped nation
could be taught to be highly ambitious, they would become more
enterprising. McClelland and his associates trained 50 Indian
businessmen from Hyderabad to think, talk, and act like highly
ambitious people. Lasting for only 40 days, the training program
was highly effective. The businesses of trainees grew more rapidly
than those of an untrained control group of Indian businessmen.
The trainees hired over twice as many new employees,thus raising
the living standards of thousands of people (McClelland & Winter,
1971).
McClelland and Winter’s (1971) work showed that achievement
motivation can be experimentally inculcated in adults. In the nor-
mal course of events, parents play an important role in training
their children to be ambitious. Research in the United States has
shown that parents inculcate a high need for achievement by train-
ing children to be independent from an early age. Early investiga-
tors focused on differences between Protestants and Catholics
(McClelland, Rindisbacher, & de Charms, 1955). Protestants were
more ambitious and this was attributed to earlier independence
training. Subsequent research showed that parents of ambitious
children actively train them to do things well.Psychologists visited
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
345
the homes of 9- to 11-year-old boys who were either high or low on
need for achievement (Rosen & D’Andrade, 1959). Children per-
formed difficult tasks, such as building towers with irregularly
shaped blocks using only one hand,in the presence of their parents.
Fathers and mothers of ambitious children were concerned about
the quality of their son’s performance.They set high standards,but
were also supportive, giving many useful hints, and were quick to
praise their son for his successes and not dwelling on failures. Par-
ents of boys who were low on need for achievement often became
visibly irritated at their son’s mistakes and rarely provided useful
hints about how to solve difficulties. Boys were criticized for fail-
ures but were not praised for success. Such experiences make chil-
dren reluctant to accept challenges. The less ambitious children in
the study generally had poorer parents and there is evidence that
poorer families generally provide far less emotional support for
children judging from analysis of conversations between parents
and young children (Hart & Risley, 1995). More recent research
supports the conclusions of Rosen and D’Andrade (1959); for
instance, Grolnick and Ryan (1989). Swank (1996) found a signifi-
cant positive association between achievement motivation and
economic growth for 25 countries in a reanalysis of the work of
Granato, Inglehart, and Leblang (1996).
Children can be trained to accept challenges, but they are
unlikely to receive such training unless they live in a society in
which achievement striving is rewarded for adults. Moreover,
teaching children to strive for success requires considerable paren-
tal time, effort, and empathy and is thus based on a high level of
parental commitment per child. Parental investment is higher in
societies with a monogamous marriage system and where fertility
rates are low because in each of these situations the parental
investment per child is increased because the efforts of two parents
are divided by a smaller number of children. High parental invest-
ment is also a feature of high sex ratio societies because women
control the marriage market and require a high level of paternal
investment and commitment as a precondition of marriage (Bar-
ber,2000a;Guttentag and Secord,1983).It was predicted that soci-
eties high on indicators of parental investment and development of
human capital (e.g., school attendance rates of children) would
havehighlevelsofeconomicgrowthasreflectedinGNPpercapita.
Downloaded from ccr.sagepub.com by guest on August 12, 2015

346
Cross-Cultural Research / November 2002
METHOD
SAMPLE OF SOCIETIES
The societies studied comprised the countries for which data on
1996 GNP per capita were provided in the World Population Data
Sheet (Population Reference Bureau, 1998). These data were used
rather than the most recently available statistics to allow
contemporaneity of the various independent variables given the
temporal lag in statistics for many developing countries. The 147
countries in the sample had an arithmetic average GNP of $5,180
and a log average of 3.20 (antilog $1,585), which indicates that the
arithmetic average is inflated by inclusion of a small number of
wealthy countries.
DEPENDENT VARIABLE
The dependent variable was the log of the GNP per capita for
1996 (Population Reference Bureau, 1998). GNP is defined as the
total value of goods and services produced by residents of a country
together with profits on property they own abroad. A log transfor-
mation was used to correct for nonlinearity in the distribution due
to the great range of variation (from $80 in Mozambique to $45,360
in Luxembourg).
INDEPENDENT VARIABLES
Independent variables intended to tap cross-national differ-
ences in parental investment included the sex ratio for 15- to 64-
year-olds (CIA, 1998), total fertility rate (or expected number of
children produced by the average woman in her reproductive
career) (Population Reference Bureau,1998),and the proportion of
women using any form of contraception (Population Reference
Bureau, 1998). The rationale is that women use contraception to
reduce their number of offspring and to plan the timing of repro-
duction,both of which increase parental investment.Polygyny was
coded based on the predominant ethnic group within a country (see
Barber, 2000a). Countries were coded as 1 = monogamous, 2 =
polygyny permitted but rare, or 3 = polygyny permitted and com-
monly practiced. Primary school attendance rates (1990-1999)
(UNICEF, 2000) were used as a measure of investment in human
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
347
capital, which is clearly related to parental investment because
parents support children while they attend school. Another possi-
ble measure of human capital, the proportion of children aged
10-14 years at work was excluded because it was highly negatively
correlated with school attendance for the obvious reason that chil-
dren who are at work cannot be at school at the same time. Female
labor participation, another possible measure of human capital
investment was omitted because the definition of female employ-
ment varies from country to country with some countries including
agricultural labor, for example, and others not. Female literacy
rates also had to be excluded because of their strong negative cor-
relation with total fertility rates.
CONTROL VARIABLES
As further clarified in the section on statistical analysis below,
control variables were defined as alternative explanations to
parental investment/human capital, and it is reasonable to think
of them in terms of clusters of independent variables that might
affect economic activity in a country. Measures of endemic disease,
which would reduce economic activity because of reduced vigor of
work, economic costs of combating illness, and loss of human capi-
tal due to absence from school and subtracted years of productive
work life, included rates of AIDS/HIV infection (percentage of
adults infected) (UNAIDS/World Health Organization [WHO],
1998) and schistosomiasis (or bilharzia), a parasitic flatworm (the
liver fluke) found in water that afflicts some 200 million people in
74 countries around the world (WHO, 2001) and has devastating
chronic effects on health that prevents many of the afflicted from
working (Gallup & Sachs, 2000). Malaria affects more people than
any other endemic disease, some 270 million worldwide. Rates of
malaria were calculated from data provided online by Rollback
Malaria (2001). However, preliminary analysis found that the cor-
relation between GNP and malaria rates was slightly smaller than
a one-zero coding of countries with malaria, so the latter was
preferred.
Many economists have pointed out that malaria saps the eco-
nomic vitality of underdeveloped nations, but it is possible that
what would show up as effects of malaria are really attributable to
the effects of living in a tropical climate.Some of these possible con-
founding influences were addressed by controlling for climate.
Downloaded from ccr.sagepub.com by guest on August 12, 2015

348
Cross-Cultural Research / November 2002
Tropical climate, as such, was found to be much more weakly
related to GNP than temperate climate. Countries were coded as
temperate if they were described as such in the geographical
descriptions in the CIA World Factbook (1998). Countries having
several different climate types were not coded as temperate even if
they included some temperate regions. Average country latitude
was defined as the line bisecting a country into two equal parts.
These values were obtained from the World Factbook (CIA, 1998)
as was the proportion of arable land in a country, which should be
positively related to GNP if the generation of a food surplus that
allows people to leave agricultural occupations is necessary for
stimulating economic growth (Landes, 1998). Whether countries
have a coastline was also coded because such countries have the
potential for increasing their prosperity through maritime trade.
The average rainfall in inches was calculated as the mean of the
different locations within a country for which these data were pre-
sented in The Statesman’s Year-book (Hunter, 1998).
Apart from disease and geographical/climatic factors, it is likely
that GNP could be affected by government policies. Many econo-
mists have pointed out that Communist governments stifle enter-
prise. Few countries are still under Communist regimes but the
cumulative effect of suppressing free enterprise ought to be still
present in the case of countries having a history of Communism.
Countries were coded as having, or not having, a history of Com-
munist government where Communism was defined as a single-
party system with government ownership of most enterprises, a
definition that included Mexico, for example. Borrowing policies
are also potentially important. External public debt as a percent-
age of GNP was obtained from Kurian (1997).Military expenditure
as a percentage of GDP was obtained from the same source. Politi-
cal corruption is also potentially important as a negative index of
the importance of good government for economic prosperity
(Ehrlich & Lui, 1999; Mauro, 1995). It was not included in this
study because of the unavailability of relevant data for many of the
countries in the sample.
Religion was coded based on the most widely held creed within a
country based on encyclopedia descriptions (Grolier’s, 1997). Five
dummy variables were coded as such: Islam, Roman Catholicism,
other Christian religions (i.e., Protestants and Eastern Orthodox),
Buddhism, and traditional religions (which are most common in
Africa). Country population for 1996 was also used as a control
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
349
variable in the combined regression analysis (Population Refer-
ence Bureau, 1998).
STATISTICAL ANALYSIS
Variables were screened for outliers using a criterion of three
standard deviations from the mean. By this criterion, China and
India were outliers on the population variable. Removal of these
cases did not affect the outcome for the population variable, how-
ever, so these cases were included in the final analysis. The foreign
debt variable also had potential outliers whose removal did not
alter the result.
The regression analysis was carried out in two steps. First sepa-
rate models were developed for geographic/climatic effects,govern-
ment policies,endemic diseases,religion,and parental investment.
The purpose of these various models was twofold: to determine
what proportion of the variance that was explained by each of
these clusters separately and to identify factors for inclusion in a
combined model that allowed parental investment indicators to be
assessed with the other factors controlled. Some readers may feel
that which variables are considered independent variables and
which are control variables is somewhat arbitrary and, in view of
the comparative novelty of several of the “control” variables in this
exploratory research, this is hard to dispute. Variables that pro-
duced a marginally significant t value in the preliminary models (p
< .10) were added to the final equation.
RESULTS
Meansandstandarddeviationsofthevariablesarepresentedin
Table 1. Results for diseases, government, religion, and geography
are summarized in text to conserve space rather than being pre-
sented in the usual tables. The theoretically central parental
investment model is presented in Tables 2 and 3 and the combined
model in Tables 4 and 5.
Each of the three diseases, bilharzia, HIV/AIDS, and malaria,
was a significant negative predictor of log GNP rs(145) = –.35,–.36,
and –.60, ps < .05, respectively. According to the regression analy-
sis, diseases explained 40% of the variance in log GNP with a large
effect of malaria and modest significant effects of bilharzia and
Downloaded from ccr.sagepub.com by guest on August 12, 2015

350
Cross-Cultural Research / November 2002
TABLE 1
Means and Standard Deviations of Variables
Variable
M
SD
Log GNP
HIV/AIDS (%)
3.20
4.41
0.68
4.34
Malaria
0.49
0.50
Bilharzia
0.44
1.00
Military budget
Foreign debt
Communist history
Roman Catholic
Other Christians (Protestants and Eastern Orthodox)
Buddhism
2.21
59.33
0.27
0.37
0.20
1.47
70.33
0.44
0.48
0.40
0.06
0.34
Traditional religions
Islam
0.14
0.20
0.34
0.40
Rainfall (inches)
Temperate climate
Latitude
46.36
0.22
26.16
0.76
29.68
0.42
17.95
0.43
Coast
Arable land (%)
Polygyny
15.76
1.71
14.59
0.87
Population
Total fertility
36.93
3.49
133.41
1.84
Contraception (%)
Elementary school attendance (%)
Sex ratio (15-64 years of age)
42.92
50.37
98.29
20.87
22.71
4.96
HIV/AIDS (standardized regression coefficients of .50,.15,and .13,
respectively).(Note that “variance explained”is used here and sub-
sequently in a statistical sense and is not intended to imply causa-
tion.) HIV/AIDS was eliminated from the combined model as add-
ing little to the prediction of log GNP with the other diseases
controlled.
National debt and Communist history were significant negative
predictors of log GNP, rs(145) = –.24, –.19, respectively, ps < .05,
but military budget was not correlated with log GNP. Government
policies explained just 9% of the variance in log GNP with signifi-
cant negative effects of foreign debt and Communist history but no
effect of military budget (standardized regression coefficients of
.23, .17, and .04, respectively).
Roman Catholic and other Christian countries were wealthier,
rs = .30, .28, respectively, ps < .05. Countries practicing traditional
religions and Islam were poorer, rs of –.44, –.24, respectively, ps <
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
351
TABLE 2
Correlations Between Log GNP Per Capita and
Parental Investment with Means and Standard Deviations
Total
Sex
Variable
GNP
Polygyny Fertility Contraception School
Ratio
Sex ratio
School
Contraception
Total fertility
Polygyny
.36*
.76*
.60*
–.73*
–.60*
–.14
–.50*
–.51*
.64*
—
–.15
–.70*
–.62*
—
.29*
.58*
—
.21*
—
—
*p < .05.
TABLE 3
Regression Analysis of Parental Investment and Log GNP
(Standardized Regression Coefficients in Parentheses)
Predictor
B
SE B
t
Sex ratio (15-64 years of age)
School attendance
Contraception
Total fertility
Polygyny
Constant
.029 (.21)
.0049 (.16)
.0030 (.09)
.0069
.0021
.0016
.018
.039
.69
4.20**
2.35*
1.88
8.42**
3.59**
—
–.16
(.43)
(.18)
–.14
.21
NOTE:Theequationaccountedfor64%ofthevariance,F(4,141)=49.26,p<.001.
*p < .05. **p < .01.
.05. Buddhism was not correlated with national wealth. In the
regression analysis, religions explained 33% of the variance in log
GNP. The regression analysis confirmed the correlations in terms
of significance and direction of effects. Standardized regression
coefficients were as follows: traditional religions .33, other Chris-
tian .26, Roman Catholic .25, Islam .14, and Buddhism .00.
Correlational analysis of geography/climate found that having a
coastline, mean latitude, and temperate climate were significant
positive predictors of log GNP, rs(145) = .26, .56, and .56, respec-
tively, ps < .05. Neither arable land nor rainfall was significantly
correlated with log GNP, rs(145) = .10, –.10, respectively, ps > .05.
Geography and climate accounted for 52% of the variance in GNP
with highly significant positive effects of latitude, temperate cli-
mate, and coastline (standardized regression coefficients of .53,
.33, and .26, respectively). Rainfall and arable land emerged as
Downloaded from ccr.sagepub.com by guest on August 12, 2015

TABLE 4
Correlations Between Independent Variables in Combined Model
Variable
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
1. Roman Catholic
2. Islam
.06 .03 .00 –.02 .16* .02 –.04 –.36* –.11 –.13 .22* .24* .14 –.13 –.11 –.30* –.39* –.25*
.02 –.08 –.07 –.15 –.18* –.04 –.03 .40* –.02 .19* –.21* –.15 .02 .10 –.14 –.20* –.25*
—
—
3. Other Christians
(Protestants and
Eastern Orthodox) .43 .08 .14 –.05 .25* –.14 .12 –.31* –.05 –.35* .10 .23* .03 –.29* –.06 –.20*
—
4. Traditional
5. Bilharzia
6. Malaria
7. Sex ratio
8. School
9. Contraception
10. Fertility
11. Population
12. Polygyny
13. Communist
14. Debt
.31* –.06 –.10 .12 –.21* .23* –.10 .50* –.05 .46* –.39* –.33* –.24* .41* .40*
–.30* –.15 –.14 .04 –.19* .05 –.11 .38* .00 .45* –.32* –.36* –.24* .30*
–.61* –.03 –.23* .23* –.46* .10 –.19* .48* .02 .51* –.41* –.50* –.19
.08 .15 .04 .04 .10 –.10 –.31* –.14 .22* –.15 .29* .10
.50* .26* .09 –.23* .41 –.07 .03 –.49* –.10 –.64* .44*
.27* .26* .16* –.02 .19* –.11 .03 –.51* .17* –.62*
–.65* –.13 –.32* .12 –.40* .26* –.21* .64* –.08
.02 .11 .16* .08 –.05 –.07 .06 .04
–.42* –.13 –.08 .00 –.34* .11 –.13
.31* –.21* .16* –.18 –.10 .07
.26* .17* –.06 .08 –.11
—
—
—
—
—
—
—
—
—
—
—
15. Temperate
climate
16. Rain
17. Arable land
18. Coast
.53* .07 .21* –.29
–.46* .26* –.13
.36* –.01
–.11
—
—
—
—
19. Latitude
—
*p < .05.

Barber / PARENTAL INVESTMENT AND WEALTH
353
TABLE 5
Regression Analysis of Combined Predictors and Log GNP
(Standardized Regression Coefficients in Parentheses)
Predictor
B
SE B
t
Roman Catholic
Islam
.075
–.17
(.05)
(.10)
.043
.052
1.74
3.27**
Other Christians (Protestants
and Eastern Orthodox)
Traditional religions
Bilharzia
.047
–.21
.046
–.12
.017
.010
.0050 (.15)
–.063 (.17)
–.00044 (.09)
.0065 (.01)
(.03)
(.11)
(.07)
(.09)
(.12)
(.33)
.051
.060
.021
.041
.0041
.00090
.00098
.011
.00015
.023
0.92
3.50**
2.19*
2.93**
4.15**
11.00**
5.10**
5.73**
2.93**
0.28
Malaria
Sex ratio (15-64 years of age)
School attendance
Contraception
Total fertility
Population
Polygyny
Communist history
External debt
Temperate climate
Rain
Arable land
Coast
–.35
–.0050 (.05)
.19 (.12)
.0025 (.11)
–.0045 (.10)
(.23)
.046
.00029
.049
.00069
.0014
.048
.0011
.42
7.61**
1.72
3.88**
3.62**
3.21**
1.73
.083
.012
.77
(.05)
(.29)
Latitude
Constant
9.09**
—
NOTE:Theequationaccountedfor88%ofthevariance,F(19,127)=50.37,p<.001.
*p < .05. **p < .01.
significant predictors of log GNP in a direction opposite to the sign
of their (nonsignificant) correlation coefficients. National wealth
increased with rainfall and declined with the proportion of arable
land (standardized regression coefficients of .15 and .13, respec-
tively). The direction of the effect for arable land was contrary to
the prediction that increased capacity for food production would
facilitate economic development.
Correlations for the parental investment factors are presented
in Table 2 along with means and standard deviations.
All of the independent variables were significantly correlated
with log GNP. All of the variables reached statistical significance
in the regression analysis (Table 3) with exception of contracep-
tion, which was marginal (p < .10). The strongest effect was pro-
duced by total fertility. The effect of school attendance was weak,
despite a large correlation (Table 2). Country wealth (log GNP)
increased with the sex ratio, with school attendance, and with
Downloaded from ccr.sagepub.com by guest on August 12, 2015

354
Cross-Cultural Research / November 2002
contraception and was inversely related to total fertility and
polygyny. All of these effects were in the predicted direction.
Together, the independent variables explained close to two thirds
of the variance (64%) in log GNP.
The combined regression model is presented in Table 5 and the
correlations between variables in Table 4. The combined model
explained 88% of the variance. All of the parental investment vari-
ables remained significant in the combined model with the excep-
tion of polygyny.Loss of significance for polygyny was probably due
to controlling for a number of variables that were significantly cor-
related with polygyny, including Islam, traditional religions, lati-
tude, total fertility, school attendance, contraception, malaria, and
bilharzia (Table 4). It can be seen from Table 4 that total fertility
was strongly correlated with latitude,with contraception,and with
school attendance. Latitude was also strongly correlated with
malaria and with temperate climate.
The effect size of school attendance was increased by the multi-
ple statistical controls of other variables in the combined models
(the standardized coefficient increased from .16 to .33). A number
of variables in the combined model just failed to reach statistical
significance. They included Roman Catholic, external debt, and
coast.The effect sizes for all of the religious variables were reduced
and the other Christian variable did not approach significance.
The largest effects, in descending order, were school attendance,
latitude, Communist history, total fertility, contraception, the sex
ratio, temperate climate, traditional religion, rainfall, arable land,
Islam, malaria, and population size. National wealth decreased
significantly with population size in the regression analysis, even
though the simple correlation was not significant,r(145) = –.06,p >
.10. The negative effect of arable land was confirmed in the com-
bined model.
Results thus indicated (a) that parental investment and human
capital explain a substantial fraction of the variance in GNP and
(b) that these effects remain after controlling for a variety of other
factors affecting economic development. Yet it is still possible that
parental investment does not contribute any additional variance
over the other variables in the combined model. This question was
addressed by entering all of the 13 control variables in another
regression analysis.Between them,they explained 75% of the vari-
ance, which was significantly less than the 88% accounted for in
the combined model, Fisher’s z = 4.50, p < .0001. Using the same
method of subtracting factors from the combined model, the 13% of
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
355
the variance uniquely explained by parental investment/human
capital compared to 9% for government, 8% for geography/climate,
6% for religion, and 5% for endemic disease.
DISCUSSION
Results showed that GNP is predictable from all of the clusters
of factors examined. In terms of the variance explained, the most
effective prediction was provided by parental investment/human
capital followed by geography/climate, endemic diseases, religion,
and government, in that order. As far as the variance uniquely
explained by each of the clusters,the order was somewhat different
with parental investment/human capital first, government and
geography/climate tied in second,followed by religion and endemic
disease. Clearly all of these clusters of factors are important for
explaining country differences in wealth, even though most of the
variance is shared between the different clusters (i.e., their unique
contribution is always much less than the variance explained in
the separate models with the solitary exception of government).
Even though there are several substantial correlations among the
factors of the combined model (Table 4), multicollinearity was
apparently not a major problem in the sense that removal of sus-
pectvariablesdidnotaltertheerrorestimatesofthoseremaining.
Needless to say, the ordering of importance for different clusters
may have been biased by the selection of variables and by the num-
ber of factors representing each cluster. With these qualifications,
it is clear that variables tapping parental investment are
extremely useful in explaining cross-national differences in
wealth and are possibly more important in this respect than the
political and economic variables conventionally applied to this
problem. The results thus add generality to recent research show-
ing that family variables relevant to parental investment (family
size, single parenthood, marriage rate, childhood workforce partic-
ipation) are modestly predictive of annual economic growth in
developedcountries(Kicketal.,2000).Mostofthepreviousquanti-
tative research in this field has restricted itself to analysis of
growth rates that are important for predicting future prosperity
but has paid little attention to the more fundamental issue of why
some countries are richer than others. GNP has some advantages
over growth as a dependent variable, the most important of which
is that country differences are probably more stable whereas
Downloaded from ccr.sagepub.com by guest on August 12, 2015

356
Cross-Cultural Research / November 2002
growth rates fluctuate from year to year due to endogenous busi-
ness cycles that can vary greatly from country to country introduc-
ing noise in statistical analyses and possibly contributing to incon-
sistent findings (Feng, 1996). Many researchers have addressed
this problem by looking at average growth rates over several
decades.
Another important qualification is that the correlational results
reported here are amenable to many different interpretations and
do not mean that the independent variables have a causal relation-
ship with the dependent variable. It is possible that there is no
causal relationship present. It is also possible that the direction of
causality is reversed. Thus, the use of contraception as a measure
of fertility planning and hence as an index of high parental invest-
ment may be in error because contraception might be a response to
national wealth as much as a cause. For example, wealthy coun-
tries typically provided better career opportunities for women and
they may limit their fertility by using contraception to take advan-
tage of these opportunities. Similarly, even an elementary school
education might be impossible in poor countries because of the eco-
nomic contribution of children to their families. Limited educa-
tional opportunities might thus be an effect as well as a cause of
poverty. In the same vein, it seems likely that educational opportu-
nities are increased in developed countries so that wealth could
cause education as much as education causes wealth. From a his-
torical perspective, it is clear that the Industrial Revolution in
England preceded universal education, which supports the
hypothesis that wealth increases education.
The results are clearly preliminary and are intended only to
suggest that family variables are of potential importance for
causal explanation of national differences in wealth. Of the many
possible statistical techniques that can be applied to distinguish
correlation and causation, reviewers have generously suggested
the following. A longitudinal approach could be used to analyze
changes in national wealth as a function of changes in family vari-
ables with constant country effects statistically removed. Causal
modeling would also be appropriate. Path models that include
reciprocal causation seem particularly desirable because of the
apparent bidirectionality of causation previously suggested.
This study replicated Gallup et al. (1999) in finding a strong
negative relationship between presence of malaria and GNP.
Malaria may be a proxy for a number of geographic variables that
are correlated with the disease, the most obvious of which are
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
357
latitude and climate with which it was significantly correlated
(Table 4). It is thus hardly surprising that the effect of malaria was
greatly reduced in the combined model (the standardized coeffi-
cient declined from .50 to .09). The finding that bilharzia and HIV/
AIDS are also negatively associated with the wealth of a country
adds generality to the disease hypothesis used to explain country
cross-national differences in wealth.
It is interesting that despite controlling for all of the other vari-
ables in the combined analysis, geographic latitude maintained a
substantial effect size (.29). This is unlikely to be due to climatic
factors, such as heat and humidity (that were controlled indirectly
in terms of temperate climate and rainfall) but it could reflect the
cumulative effects of numerous uncontrolled endemic diseases and
parasites such as leprosy and river blindness.
Another interesting result was the counterintuitive finding that
the proportion of arable land in a country is negatively related to
log GNP, at least with other geographic/climatic factors controlled.
This would appear to falsify the hypothesis concerning the essen-
tial role of food production in permitting industrial development
(Landes, 1998). Yet there is at least one important possible con-
founding variable, namely the productivity of agricultural land.
This is higher in the case of advanced societies that use more scien-
tific farming methods so that advanced societies would need less
acreage to satisfy their needs (Gallup & Sachs, 2000). Another
issue is that in developing countries, such as Bangladesh, more
people live on farms, which may have the effect of boosting the
acreage of arable land, particularly in marginal regions. The pro-
portion of workers in agriculture was not included in the analysis
because this measure (which is negatively correlated with urban-
ization) is arguably a result of industrial development and eco-
nomic growth drawing population to cities rather than a cause.
Taken at face value, the negative relationship between arable land
and wealth adds generality to the equally counterintuitive finding
of earlier research that mineral wealth of a country is negatively
correlated with economic development (Sachs & Warner, 1997).
Gallup et al.(1999) found that there was no relationship between a
country’s oil and natural gas reserves and its GDP.
Even though the effect sizes of most of the variables were
reduced in the combined model (with two, polygyny and other
Christian, disappearing and three, Roman Catholic, external debt,
and coast, just missing statistical significance), at least one indica-
tor of all of the clusters of variables remained significant. This
Downloaded from ccr.sagepub.com by guest on August 12, 2015

358
Cross-Cultural Research / November 2002
shows that predicting GNP is complex and suggests that no one
cluster, however well chosen, can account for all of the explainable
variance.
The measurement of GNP is itself rather complex and contro-
versial. Some economists prefer to measure it in constant dollars
(the approach used here) but others believe that a country’s wealth
should be measured in terms of the purchasing power of its citi-
zens. The cost-of-living adjustment to national productivity is
problematic, however, and it is vulnerable to major distortions as a
result of currency devaluations, for example, so that O’Hearn
(1998) argues that GNP is preferable as a measure of the standard
of living in a country. When the purchasing power parity method is
used, the United States, having a relatively low cost of living,
moves up the ranks,for example,and Switzerland,with a high cost
of living, moves down. A substantial fraction of the unexplained
variance could be due to such imprecision in the dependent vari-
able. The lack of validity of the dependent variable as a measure of
country differences in wealth can be quantified. The correlation
between log GNP and log purchasing power using the most recent
statistics (World Bank, 2000) was r(152) = .96, implying 92%
shared variance and providing an error estimate of 8%. This sug-
gests that the maximum explainable variance of log GNP would be
92%.Given the possibility of such error,as well as potential inaccu-
racies in the calculation of GNP for various countries,the variables
examined in this article could be tapping close to all of the explain-
able variance. Yet there are many variables that other scholars
have identified as potentially important that merit further investi-
gation. They include, but are not limited to, enrollment in second-
and third-level education, the role of scientific/technological inno-
vation, degree of bureaucratic regulation, relations among corpo-
rations, enforcement of contract laws, efficiency of government,
political corruption and organized crime, governments encourag-
ing personal freedom,openness to international trade,institutions
such as common law and science, a history of colonial domination,
tax rates, savings rates, inflation, domestic investment, income
inequality, gender inequality, and female labor participation (see
the Method section) (Ehrlich & Lui, 1999; Fedderke & Klitgaard,
1998; Forsythe, Korzeniewicz, & Durrant, 2000; Niggle, 1998).
Entry of any of these variables into the combined model could
reduce any of the effect sizes in the existing equation. In conclu-
sion, it has been shown that consistent with previous research on
economic growth, family variables tapping parental investment
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
359
are important for explaining cross-national differences in wealth.
These results are purely correlational, of course, and many inter-
pretations are possible, one of which is that there is no causal rela-
tionship between parental investment and national wealth but
that their association is caused by a third variable, such as urban-
ization of the population. Assuming a causal relationship, high
parental investment might increase human capital and increased
human capital could promote economic development. This hypoth-
esis is empirically testable.
References
Argyle, M. (1994). The psychology of social class. London: Routledge.
Banfield, J. (1998, September). Malaria: Africa’s public enemy no. 1. Afri-
can Business, pp. 35-36.
Barber, N. (2000a). The sex ratio as a predictor of cross-national variation
in violent crime. Cross-Cultural Research, 34, 264-282.
Barber,N.(2000b).Why parents matter:Parental investment and child out-
comes. Westport, CT: Bergin & Garvey.
Belsky, J., Steinberg, L., & Draper, P. (1991). Childhood experience, inter-
personal development,and reproductive strategy:An evolutionary the-
ory of socialization: Child Development, 62, 647-670.
CIA. (1998). World factbook. Washington, DC: Author.
Coleman, J. S. (1990). Foundations of social theory. Cambridge, MA: Har-
vard University Press.
Ehrlich, I., & Lui, F. T. (1999). Bureaucratic corruption and endogenous
economic growth. Journal of Political Economy, 107, S270-S293.
Fedderke, J., & Klitgaard, R. (1998). Economic growth and social indica-
tors: An exploratory analysis. Economic Development & Cultural
Change, 46, 455-490.
Feng, Y. (1996). Democracy and growth: The sub-Saharan African case
1960-1992. Review of Black Political Economy, 25, 95-125.
Forsythe, N., Korzeniewicz, R. P., & Durrant, V. (2000). Gender inequali-
ties and economic growth: A longitudinal evaluation. Economic Devel-
opment and Cultural Change, 48, 573-618.
Gallup, J. L., & Sachs, J. D. (2000). Agriculture, climate, and technology:
Why are the tropics falling behind. American Journal of Agricultural
Economics, 82, 731-737.
Gallup, J. L., Sachs, J. D., & Mellinger, A. D. (1999). Geography and eco-
nomic development (CID Working Paper No.1).Cambridge,MA:Center
for International Development.
Graham,D.(1987,May).Sharialoosensitsgrip.Euromoney,pp.137-138.
Granato, J., Inglehart, R., & Leblang, D. (1996). The effects of cultural val-
ues on economic development: Theory, hypotheses, and some empirical
tests. American Journal of Political Science, 40, 607-631.
Grolier’s.(1997).Grolier’smultimedia encyclopedia.Danbury,CT:Author.
Downloaded from ccr.sagepub.com by guest on August 12, 2015

360
Cross-Cultural Research / November 2002
Grolnick, W. S., & Ryan, R. M. (1989). Parents’ styles associated with chil-
dren’s self-regulation and competence in school. Journal of Educa-
tional Psychology, 81, 143-154.
Guttenberg,M.,& Secord,P.F.(1983).Too many women:The sex ratio ques-
tion. Beverly Hills, CA: Sage.
Hall, R., & Jones, C. (1996). The productivity of nations (NBER Working
Paper No. 5812). Cambridge, MA: National Bureau of Economic Re-
search.
Hart, B., & Risley, T. (1995). Meaningful differences in the everyday experi-
ences of young American children. Baltimore, MD: Paul H. Brookes.
Healy, T. (1998, June/July). Counting human capital. OECD Observer,
31-33.
Hunter,B.(Ed.).(1998).Thestatesman’syear-book.NewYork:St.Martin’s.
Kaplan,H.(1994).Evolutionary and wealth flows theories of fertility:Em-
pirical tests and new models. Population and Development Review, 20,
753-791.
Kenny, C. (1999). Why aren’t countries rich: Weak states and bad neigh-
borhoods. Journal of Development Studies, 35, 26-47.
Kick,E.L.,Davis,B.L.,Kiefer,D.M.,& Burns,T.J.(1998).A cross-national
analysis of militarization and well-being relationships in developing
countries. Social Science Research, 27, 351-352.
Kick, E., Davis, B., Lehtinen, M., & Wang, L. (2000). Family and economic
growth: A world-system approach and a cross-national analysis. Inter-
national Journal of Comparative Sociology, 41, 225-244.
Kindleberger, C. P. (1996). World economic primacy: 1500-1900. New York:
Oxford University Press.
Kurian, G. T. (1997). The illustrated book of world rankings. Armonk, NY:
Sharpe Reference.
Landes, D. S. (1998). The wealth and poverty of nations: Why some are so
rich and some so poor. New York: Norton.
Liese, B. H. (1998). A brake on economic development. World Health, 51,
16.
Lino, M. (1999, Winter). Expenditures on children by families, 1999. Fam-
ily Economics and Nutrition Review, pp. 56-74.
Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics,
110, 681-712.
McClellan, J. E., & Dorn, H. (1999). Science and technology in world his-
tory. Baltimore: Johns Hopkins University.
McClelland, D. C., Atkinson, J. W., Clark, R. W., & Lowell, E. L. (1976). The
achievement motive. New York: Irvington.
McClelland, D. C., Rindlisbacher, A., & de Charms, R. C. (1955). Religions
and other sources of parental attitudes toward independence training.
In D. C. McClelland (Ed.), Studies in motivation. New York: Appleton-
Century-Crofts.
McClelland, D. C., & Winter, D. G. (1971). Motivating achievement motiva-
tion. New York: Free Press.
Mook, D. G. (1987). Motivation: The organization of action. New York:
Norton.
Niggle, C. J. (1998). Equality, democracy, institutions, and growth. Journal
of Economic Issues, 32, 523-530.
Downloaded from ccr.sagepub.com by guest on August 12, 2015

Barber / PARENTAL INVESTMENT AND WEALTH
361
O’Hearn,D.(1998).Inside the Celtic tiger:The Irish economy and the Asian
model. London: Pluto.
Population Reference Bureau. (1998). World population data sheet. Wash-
ington, DC: Author.
Porter, M. E. (1998). The competitive advantage of nations. New York: Free
Press.
Ramey, C. T. (1992). High-risk children and IQ:Altering intergenerational
patterns. Intelligence, 16, 239-256.
Rollback Malaria. (2001). Country profiles. Retrieved from http://www.
rbm.who.int
Rosen, B. C., & D’Andrade, R. (1959). The psychological origins of achieve-
ment motivation. Sociometry, 22, 185-218.
Rosenberg,N.,& Birdzell,L.E.,Jr.(1987).How the West grew rich:The eco-
nomictransformationoftheindustrialworld.NewYork:BasicBooks.
Sachs, J., & Warner, A. (1997). Natural resource abundance and economic
growth (NBER Working Paper No.5398).Cambridge,MA:National Bu-
reau of Economic Research.
Swank, D. (1996). Culture, institutions and economic growth. American
Journal of Political Science, 40, 660-679.
Trivers, R. L. (1972). Parental investment and sexual selection. In B. Campbell
(Ed.),Sexual selection and the descent of man (pp.136-179).Chicago,IL:
Aldine Atherton.
UNAIDS/World Health Organization. (1998). Report on the global HIV/
AIDS epidemic. Geneva, Switzerland: Author.
UNICEF. (2000). The progress of nations 2000. Retrieved from
http://www.unicef.org/pon00/statistics.html
Weber,M.(1930).The Protestant work ethic (T.Parsons,Trans.).New York:
Scribners.
World Bank. (2000). World development indicators database. Retrieved
from http://www.ciesin.org/lw-kmn/guides/sid.html
World Health Organization.(2001).Global distribution of schistosomiasis.
Retrieved from http:///www.who.int/ctd/schisto/geofull.jpg
Woytinski, W. S., & Woytinski, E. S. (1953). World population and produc-
tion. New York: Twentieth Century Fund.
Nigel Barber received his Ph.D. in biopsychology from Hunter College of
the City University of New York and has had teaching appointments at
Memidji State University and Birmingham Southern College.His research
interests in evolutionary psychology include parental investment theory,
physical attractiveness fashions,and gender stereotypes.He is the author of
Why Parents Matter: Parental Investment and Child Outcomes (2000,
Bergin and Garvey) and The Science of Romance (2002, Prometheus).
Downloaded from ccr.sagepub.com by guest on August 12, 2015