World Development 117 (2019) 127–137

Contents lists available at ScienceDirect

World Development
journal homepage: www.elsevier.com/locate/worlddev

Gender differences in risk behavior: An analysis of asset allocation

decisions in Ghana
Marya Hillesland ⇑
American University, Department of Economics, 4400 Massachusetts Ave, NW, Washington, DC 20016, United States

a r t i c l e i n f o a b s t r a c t

Article history: Within a developing country context, little is known about gender differences in risk preferences as
Accepted 4 January 2019 reflected in asset allocation decisions. Much of the empirical literature within development studies on
Available online 22 January 2019 risk and investment in assets centers on risk coping and risk management strategies of households.
With some exceptions, this literature primarily focuses on household and household-level outcomes.
JEL classification: Previous experimental studies have explored gender differences in risk preferences in developing coun-
D8 tries with inconclusive results. Using unique national-representative sex-disaggregated data with self-
G1
reported asset ownership and wealth information of individuals within households in Ghana from a
O1
D3
multi-country project, The Gender Asset Gap Project, this paper explores men and women’s risk prefer-
ences as reflected in asset allocation decisions through a decomposition method typically used to explore
Keywords: gender differences in wage employment. The study finds that although women hold significantly fewer
Gender difference risky assets than men in absolute terms and as a proportion of their wealth in the sample, men and
Risk preferences women do not have systematically different risk preferences. The results in this paper differ from the
Ghana results in many empirical studies in developed countries, where women are often found to be more risk
Wealth
averse than men. As the first study to look at gender differences in risk aversion in terms of asset alloca-
Asset allocation decisions
tion with in the developing country context, the results from this study provide evidence that gender dif-
Risk aversion
ferences in risk attitudes may vary by cultural context as the growing experimental literature in
developing countries suggests.
Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction of return. For risk averse individuals, investing in risky assets

requires that the expected return of risky assets is greater than
Physical and financial assets serve a number of important func- the return from risk-free assets. The difference must be large
tions. Physical assets—such as a household’s residence and vehi- enough to offset the value the individual places on the cost of tak-
cles—may provide current and future consumption value and ing the risk.
also be a means of production, generating future consumption Within a developing country context, little is known about gen-
flows. Financial assets have a monetary value that can be con- der differences in risk preferences as reflected in asset allocation
verted into future consumption. Both types of assets may generate decisions. Much of the literature within development studies on
profits or losses as well as rent or interest. They also may be held as risk and investment in assets focuses on risk coping and risk man-
a form of savings as a way to self-insure against possible future agement strategies of the household as opposed to individuals
economic hardships. within households. As a coping strategy, accumulating assets is
An individual’s propensity for risk influences how his or her an important form of self-insurance against consumption loss
assets are allocated between savings and riskier assets that may due to shocks. Poor households with relatively low asset holdings
generate profits or result in losses. Risk averse individuals prefer in developing countries often hold a significant share of their assets
to invest in secure assets or assets with a constant rate of return as precautionary savings in the form of cash or other assets to
over risky assets of the same expected value with a variable rate insure against possible negative shocks (Fafchamps, Udry, &
Czukas, 1998; Lee & Sawada, 2010; Park, 2006). Poor households
have an even greater propensity to save when faced with liquidity
⇑ Address: Food and Agricultural Organization of the United Nations, Viale delle
and credit constraints (Lee & Sawada, 2010).
Terme di Caracalla, 00153 Rome, Italy.
E-mail addresses: [email protected], [email protected]

https://doi.org/10.1016/j.worlddev.2019.01.001
0305-750X/Ó 2019 Elsevier Ltd. All rights reserved.
128 M. Hillesland / World Development 117 (2019) 127–137

In many low-income countries, where wage labor opportunities interventions that require some amount of risk taking will need to
are minimal, households invest their assets into different produc- be tailored to fit these differences to ensure gender inclusiveness.
tive activities. As a form of risk management, poor households with
lower resilience and higher sensitivity to shocks often seek to min-
imize risk by allocating their wealth into low-risk, low-return 2. Previous literature on gender and risk aversion
activities rather than higher-risk, higher-return activities to cush-
ion themselves from the effects from possible shocks (e.g. Many of the empirical studies that have explored gender differ-
Dercon, 2005; Zimmerman & Carter, 2003; Dercon & ences in risk preferences in terms of asset allocation decisions have
Christiaensen, 2011). done so using data from the United States, most of which focus on
The discussion on asset portfolios, risk management, and risk retirement portfolios with mixed findings (Riley & Chow, 1992;
coping strategies in developing countries largely focuses on the Bajtelsmit & VanDerhei, 1997; Sundén & Surette, 1998; Bernasek
household and household level outcomes. There are a few studies & Shwiff, 2001; Arano, Parker, & Terry, 2010). Jianakoplos and
that explore gender differences in asset allocation and accumula- Bernasek (1998) is one of the only empirical studies that explores
tion (Antonopoulos & Floro, 2005; Quisumbing, 2010; Oduro & gender differences in risk preferences as reflected in asset alloca-
Doss, 2018) and gender differences in the effects of negative house- tion beyond pension decisions. Using 1989 Survey of Consumer
hold shocks within developing countries (Dercon & Krishnan, Finances (SCF) data of 3143 households, Jianakoplos and
2000; Doss, 2001; Doss, McPeak, & Barrett, 2008). None, however, Bernasek (1998) find that single female-headed and couple house-
specifically explore gender differences in risk preferences as holds have greater relative risk aversion as reflected in financial
reflected in asset allocation decisions. asset allocation decisions than single male-headed households.
This gap is primarily due to the lack of sex disaggregated asset Comparing risk aversion of couple households to single households
and wealth data. Until recently, most household surveys that col- is problematic, however, as it assumes married couples act as an
lected wealth and asset data did so at the household level with individual (i.e. act in unison with no conflicting interests). Knowl-
the assumption that wealth held by individuals within the house- edge of decision-making within the household would help tease
hold is pooled. As such, sex-disaggregated wealth data is not read- out individual risk attitudes. Indeed, Jianakoplos and Bernasek
ily available, particularly within a developing country context. To (1998) admit their estimates are ‘‘clouded” by not knowing who
begin to address this gap, a nationally representative sex- makes the decisions about asset allocation within the household.
disaggregated asset and wealth survey was implemented in Ghana Within the experimental literature many studies focus on uni-
in 2010 as part of a multi-country project, The Gender Asset Gap versity students in the United States, largely finding women are
Project. more risk averse than men to varying degrees (see survey by
Using this unique sex-disaggregated data containing detailed Eckel & Grossman, 2008; Croson & Gneezy, 2009; Charness &
self-reported information on asset ownership and wealth from Gneezy, 2012) with some criticism (Nelson, 2015, 2016). The few
two individuals of the opposite sex in each household, this paper studies that explore gender differences in risk preferences within
explores gender differences in risk preferences in terms of asset al- developing countries also mostly find that men are less risk averse
locations in Ghana. Similar empirical studies use household-level than women (Fletschner et al., 2010; Cárdenas et al., 2012;
data from the United States (e.g. Jianakoplos & Bernasek, 1998; Khachatryan et al., 2015). There is some evidence, however, that
Sundén & Surette, 1998). This is the first study of its kind within gender differences in risk preferences may vary by country. For
a developing country context with individually self-reported example, Cárdenas et al. (2012) find that among a sample of chil-
assets. dren in Colombia and Sweden, girls are more risk averse than boys;
Previous experimental studies have explored gender differ- however, the gender differences in risk preferences is much greater
ences in risk preferences in a developing country with inconclu- in Columbia. Similarly, there is some evidence that gender differ-
sive results. Many of these studies find women are more risk ences in risk preferences change depending on how the back-
averse than men (Fletschner, Anderson, & Cullen, 2010; ground circumstances are framed (Schubert et al., 1999) and the
Cárdenas, Dreber, Von Essen, & Ranehill, 2012; Khachatryan, surrounding environment, such as the sex-composition of class-
Dreber, Von Essen, & Ranehill, 2015). However, some suggest rooms (Booth & Nolen, 2012; Booth et al., 2014).
the setting and surrounding environment may play a consider- Differences in findings across countries and environment sug-
able role in gender differences in preferences (Schubert, Brown, gest that it may be that differences found in men and women’s risk
Gysler, & Brachinger, 1999; Henrich et al., 2001; Gneezy, preferences are likely a consequence of men and women facing dif-
Leonard, & List, 2009; Booth & Nolen, 2012; Booth, Cardona- ferent constraints or due to learned societal traits, rather than
Sosa, & Nolen, 2014). reflecting actual risk preferences due to biological sex differences.
Ghana is a fitting place to examine whether women are more That is, men are not necessarily innately more risk taking than
risk averse than men as reflected in individual asset allocation women, but rather differences arrive through cultural pressures,
decisions because wealth is primarily held individually rather than which differ across countries.
pooled with other household members (Deere, Oduro,
Swaminathan, & Doss, 2013). Legal norms facilitate a strong sepa-
ration of property even within marriage, meaning individual prop- 3. Conceptualizing risk aversion
erty acquired before and during marriage—including inheritance
and gifts as well as all earnings from wages, salaries, and income Portfolio studies in high income countries often gauge risk aver-
activities—remains the property of the individual rather than joint sion as measured by the proportion of risky assets held to one’s
marital property (Deere et al., 2013; Deere & Doss, 2006). worth. Assuming that individuals are risk averse and that their
Gender differences in risk aversion is an important topic in utility functions can be characterized by the mean and variance
development. If risk aversion in terms of asset allocation decisions of final wealth, Mossin (1966) shows that in a competitive mar-
can proxy for risk aversion in other ways, such as in decisions on ket—where assets are perfectly divisible, assets can be sold at
technology adoption or in types of self-employment activities, gen- any point in time, there are no transaction costs, and the expected
der differences in risk aversion in Ghana means that programs and yield on an asset is a random variable whose distribution is known
 M. Hillesland / World Development 117 (2019) 127–137 129

~
to individuals—there is a general equilibrium for the market price In market equilibrium, Erðr2r Þ is constant across individuals. With con-
of risk.1
r
~
stant market price of risk, Erðr2r Þ, and a fixed non-variable rate of
Using this equilibrium assumption, Budd and Litzenberger r

(1972) extend Lintner’s (1970) examination of the relationship return, rf , Eq. (5) suggests that the proportion of risky assets to all
between the size of the market and the market for risky assets. If assets is inversely related to relative risk aversion for a given indi-
we assume that an investor has a utility function that is twice dif- vidual. This means that the lower an individual’s relative risk aver-
ferentiable and U 0 ðW Þ >0 and U 00 ðW Þ < 0, where W is the investor’s sion (and thus absolute risk aversion), the greater the optimal
asset wealth, then the market price of risk is equal to the inverse of demand for risky assets as a proportion of wealth, ak .
~
the sum of the measure of individual investors’ absolute risk aver- The market price of risk, Erðr2r Þ, assumes individuals have similar
r
sion, multiplied by the aggregate market value of all the risky
expectations about the risks of assets. Additionally, only aggregate
assets (Lintner, 1970; Budd & Litzenberger, 1972). Using this rela- risks should affect prices. Idiosyncratic risks are assumed to be
tionship, Friend and Blume (1975) use a Taylor series expansion to
diversified away, so that the marginal price of risk is constant
show that the ratio of the value of total liquid assets across k inves- across individuals. This means individual shocks do not affect the
tors is a function of individual investors’ risk tolerance. This means
market equilibrium and that no single individual is subject to a
that the proportion of wealth held in risky assets to total liquid random asset price shock that is not shared with everyone else
wealth is inversely related to Pratt’s (1964) measure of relative risk (i.e. individuals are only subject to covariate price shocks). If, for
aversion. Pratt’s (1964) measure of relative risk aversion is a single instance, an individual completely lost an asset (so that the return
dimensional outcome variable that captures an individual’s is less than the market price of the asset) due to an idiosyncratic
propensity to take risks with respect to wealth. In order to estimate
shock, it is assumed the loss is diversified away and does not affect
gender differences in risk preferences based on the asset allocation the marginal price of risk of holding this asset in the market.
decisions of individuals, a simple version of the Friend and Blume’s
Investments in formal financial assets in markets where indi-
(1975) measure is adapted slightly to better fit a context like viduals are price-takers—meaning that no single individual in the
Ghana.
market has the power to change prices—best meet the model’s
Suppose a risk averse individual k, who has a utility function assumptions. However, in a developing country context, only a
with respect to wealth, W, that is twice differentiable and small share of individuals access and use formal financial institu-
U 0 ðW Þ >0 and U 00 ðW Þ < 0, must decide what proportion, a, of her tions. Individuals are more likely to have access to informal or
worth to allocate her assets in risky, productive investments with semi-formal financial mechanisms locally and to save and invest
a random return, where the expected rate of return is ~r r , and in physical assets. In this context, the equilibrium is more likely
1  a in secure assets with a non-variable rate of return, r f . The to hold if the aggregate market is limited to a local market consist-
individual chooses a such that the expected value of wealth in ing of individuals with overlapping social networks engaged in



some future period, W tþ1 ¼ 1 þ rf W t þ ak W t ~r r  rf , maximizes risk-sharing and with similar expectations around the price of risk.
his or her expected utility: While households in all countries are vulnerable to shocks, house-



 holds in developing income countries are more likely to be subject
max E uk 1 þ r f W t þ ak W t ~r r  r f ð1Þ to a lack of or limited access to functioning formal insurance mech-
To simplify, let the non-variable rate of return equal zero, r f ¼ 0, so anisms, formal credit markets, or other formal institutions to help
that the difference in the return due to investing in risky assets is ~r r . them avoid consumption shortfalls when a shock occurs. However,
The first order condition is then studies suggest that in addition to self-insurance through savings,

 informal forms of insurance such as kinship networks help provide
E ~r r W t u0k W t þ ak W t ~r r ¼ 0 ð2Þ protection against loss of consumption due to idiosyncratic risks.

 Within the context of a developing country, the consumption
Using a first order Taylor series to expand uk W t þ ak W t r r around
0

risk sharing model states that when the market is complete,

W t , Eq. (2) is approximately
 
 idiosyncratic changes in household income that is not smoothed
E ~r r W t u0k ðW t Þ þ u00k ðW t Þ ak W t ~r r 0 ð3Þ through savings should be absorbed by other members of the risk
sharing network. It means that idiosyncratic income shocks should
This can be rearranged so that not affect consumption. While full risk sharing is often rejected by
 
Eð~r r Þ u0k ðW t Þ the data, many empirical studies find at least partial risk sharing.
ak ¼   ð4Þ
r2r W t u00k ðW t Þ Building on studies on risk sharing and networks, Ambrus,
Mobius, and Szeidl (2014) shows that a full insurance model
~
where ak is the optimal demand for risky assets, Erðr2r Þ is the price of among households in villages can be nearly reached through
r

risk, and r2r is the variance of the additional return to risky assets. loosely overlapping social networks. Among kinship networks,
Pratt’s (1964) measure of relative risk aversion for the kth individ- Chiappori, Samphantharak, Schulhofer-Wohl, and Townsend
ðW t Þ (2014) find nearly complete risk sharing among kin within villages
ual is C k ðW t Þ ¼ W t U}
U 0 ðW t Þ
so that Eq. (4) becomes
in Thailand. Gifts and transfers among these households reduce the

Eð~r r Þ 1 effect of liquidity constraints on household assets. In Vietnam,
ak ¼  ð5Þ
r2r C k ðW t Þ Sawada, Nakata, and Kotera (2017) take into account self-
insurance in the form of own use production of goods for consump-
tion and cannot reject full risk sharing in Vietnam. Other empirical
studies, although they generally reject full risk pooling, find evi-
1
The Expected Utility Model is often used to represent behavior under conditions dence of insurance among households in developing countries to
of risk. The Mean-Variance Model is a simplification of the Expected Utility Model, varying degrees (for example, Townsend, 1994, 1995; Grimard,
where utility can be expressed as the mean and variance of a probably distribution 1997; Dercon & Krishnan, 2000; Ogaki & Zhang, 2001; Ligon
that gives an investor different wealth outcomes at different probabilities. The two
models are equivalent when either investors’ utility functions can be represented by a
Thomas, & Worrall, 2002; Murgai, Winters, Sadoulet, & Janvry,
function that has only two moments or the portfolio return distribution is an elliptical 2002; Fafchamps & Lund, 2003; De Weerdt & Dercon, 2006;
distribution. While neither assumption is entirely realistic, the Mean-Variance Model Fafchamps & Gubert, 2007; Morduch, 2005; Mazzocco & Saini,
is considered a reasonable approximation to the solutions found in the Expected 2012; Ambrus et al., 2014).
Utility Model.
130 M. Hillesland / World Development 117 (2019) 127–137

Table 1
Risky and secure assets in Ghana.

Riskier Assets (higher return) More Secure Assets

 Stocks  Savings in formal and informal accounts and cash
 Registered businesses (as owner or part owner) that can be sold (i.e. there’s a market)  Treasury bills and bonds
 Commercial and residential real estate where there is a market

Most of the literature explores risk sharing across households, Friend and Blume (1975). Risk-free financial assets are informal
which assumes that within the household, members share risk. and formal savings accounts, cash holdings, and treasury bills
In Ghana, there is some evidence that this may not be the case. and bonds. Risky financial assets include stocks.2 Although com-
Goldstein (1999), for example, finds that men and women within mercial and residential real estate and formal business enterprises
the same households in Ghana do not necessarily share risks and are less divisible than financial assets, following Jianakoplos and
make different choices in their risk pooling groups and arrange- Bernasek (1998) and Friend and Blume (1975), commercial and res-
ments. Strategies used by individuals to reduce risk are often idential real estate that can be sold on the market and formally reg-
shared with individuals from other households, particularly those istered businesses are also included as risky assets (Table 1).3
of the same family lineage (Goldstein, de Janvry, & Sadoulet,
2004; Vanderpuye-Orgle & Barrett, 2009).
Within the Ghanian context, therefore, Eq. (5) holds if it can be 5. Data and wealth statistics
limited to a local economy, u; where individuals within house-
holds participate in risk sharing arrangements or other formal The data is from a sex-disaggregated asset and wealth house-
and informal insurance mechanisms and have similar expectations hold survey implemented in Ghana from May to July 2010, which
about riskiness of assets: was carried out as part of a multi-country gender and assets pro-
ject, The Gender Asset Gap Project. In most households, two individ-

Eð~rr Þ 1 uals of the opposite sex, who were best informed about the
ak ðWjuÞ ¼  ð6Þ
r2r C k ðW t Þ household’s assets, provided information on the household’s assets
and wealth as well as their own assets and wealth. The sampling
If there is limited risk sharing, and other formal and informal mech- frame is based on 144 enumeration areas from the national census.
anisms do not insure against idiosyncratic risks (i.e. so that most The number of enumeration areas selected within each of Ghana’s
individuals must reduce consumption in the face of an idiosyncratic ten regions was based on the region’s share of the total popula-
shock), members of the local economy will have different expecta- tion.4 Within each enumeration area, 15 households were randomly
~
tions about the riskiness of assets and the price of risk, Erðr2r Þ, will not selected to be surveyed. In all, 2170 households were surveyed.5 The
r

be constant across members. As a result, the price of risk of assets final sample for this analysis is 1341 males and 1690 females from
cannot be separated from individuals’ risk preferences and empiri- 2080 of these households.
cally it will be more difficult to capture any systematic difference in Table 2 presents gross wealth and asset allocations by sex and
risk preferences between men and women. Table 3 presents gross wealth and asset allocation across the
While resources are not necessarily pooled in households, we wealth distribution. The majority of assets are owned individu-
would expect decisions around assets to be influenced by an indi- ally.6 The value of any assets owned jointly is the value of the full
vidual’s role within the household. We would also expect that gen- asset divided by the number of owners, except for some of the reg-
der norms and institutions to partly determine the types of assets istered businesses in which the share of the business owned is
men and women hold. Women, for instance, may not necessarily reported. While one-fifth of the sample has debt, debt is minimal
have equal access to formal financial products as men, which influ- and the average net wealth is not statistically significantly different
ences the type of asset portfolios women may hold as compared to from average gross wealth. As such, gross wealth is used in this
men. Women may be more likely to hold safe assets not in formal study.
savings account, but in informal savings accounts or in the form of There is a substantial gender wealth gap. Nearly 42 percent of
cash. Because of this, this study uses an expanded definition of women and 56 percent of men in the sample hold positive wealth.
financial wealth to include all formal and informal financial assets. Of the 42 percent and 56 of women and men respectively, women
Since many men and women in Ghana hold the majority of their
2
wealth in non-financial assets, it would be ideal to analyze portfo- Few individuals in Ghana hold stock, although, Ghana has had a stock exchange
since 1990. Risky assets also include investments items such as art work and precious
lio decisions of individuals’ full range of financial and physical-
metals, but these assets are rare in Ghana and not found in the dataset.
asset wealth. However, it is difficult to classify many non- 3
In theory, portfolio investment decisions assume wealth fits closely with Arrow
financial assets as risky and risk-free. Consumer durables and (1965) and Pratt (1964) definition, where assets within a portfolio are infinitely
many productive assets are valued not just for their monetary divisible—liquid or non-lumpy—and can be reallocated without cost from one asset to
rewards, but also for their present and future consumption value. another. Financial assets come closest to this definition and are therefore most
frequently used in theoretical studies. Empirical studies, however, often include a
Additionally, data on the returns to many assets are not available
broader set of assets, even though they may not be infinitely divisible, under the
or possible to estimate with the data available. The next section implicit assumption that wealth as defined by Arrow and Pratt is highly correlated
describes how risky and non-risky assets are classified following with other measures of wealth (Meyer & Meyer, 2006).
4
previous literature. There are fewer enumeration areas in the Upper East Region due to conflict in
parts of the region.
5
Of the 2,170 households surveyed, both spouses were interviewed in 956
4. Defining wealth and the classification of risky and non-risky households. For the other 1214 households, the second respondent may be a different
family member (e.g. sibling, parent, parent-in-law) even if the first respondent is
assets in Ghana
married and lives with his or her spouse.
6
For assets that are owned jointly, as long as the individual can make the decision
This study uses the same division of risky and non-risky assets to sell the asset alone or in consultation, the value of the proportion of the asset
as previous studies such as Jianakoplos and Bernasek (1998) and owned is included in the individual’s wealth.
 M. Hillesland / World Development 117 (2019) 127–137 131

Table 2
Wealth and asset allocations by sex.

Women Men Total

(n = 1690) (n = 1341) (n = 3031)
Proportion who hold positive gross value of assets (%) 42 56 48
Of those who hold positive gross value of assets:
Mean value of risky assets of (GH¢) 1204.9 4417.6 2866.4
(5838.0) (20638.2) (15464.9)
Mean wealth (GH¢) 1426.8 4881.7 3213.6
(5970.0) (4881.7) (15666.7)
Mean alpha 0.21 0.37 0.29
(0.39) (0.46) (0.43)
Debt (formal business, real estate, or other) (GH¢) 39.7 108.4 72.4
(238.0) (948.6) (658.0)
Proportion who have debt (%) 20.1 21.3 20.6

Note: Statistics are unweighted. Standard deviations are in parenthesis. The debt value is based on a slightly smaller sample due to missing observations.

Table 3
Asset allocation across the wealth distribution.

Quintile 1 and 2 3 4 5
(n = 1575) (n = 267) (n = 602) (n = 587)
Percent female (%) 63 63 53 37
Mean value of risky assets (GH¢) 0 0.30 16.24 7093.10
– (2.76) (55.04) (23745.34)
Mean gross wealth (GH¢) 0 18.37 121.27 7838.20
– (8.60) (80.65) (23948.23)
Mean alpha 0 0.01 0.09 0.62
– (0.12) (0.27) (0.44)

Note: Statistics are unweighted. Standard deviations are in parenthesis.

hold GH¢ 1427 and men hold GH¢ 4882 average wealth.7 The mean 6. Empirical model
measure for relative risk aversion (alpha), the ratio of risky assets to
wealth, is 0.29 meaning that on average individuals hold about 29 Eq. (6) provides the theoretical basis for testing whether there
percent of their wealth in risky assets. Although not entirely compa- are differences in men and women’s relative risk aversion after
rable given that the countries are very different, this is not dissimilar controlling for other factors:
to the average values for the United States found in Friend and
Blume (1975).8
ak ¼ b0 þ b1 ln ðwealthk Þ þ b2 femalek þ b3 femalek  ln ðwealthk Þ
c
Men on average have a greater alpha than women in the sam-
þ local economykn um þhousehold v ariableskq cq
ple. Of those who hold wealth, women hold 21 percent of their
wealth in risky assets and men hold 37 percent. A t-test suggests þindiv idual v ariableskm wn þ 
that we can reject the hypothesis that the means are equal at a
0.1 percent significance level. If men and women in Ghana exhibit
where c ak is the relative risk aversion measure. It is the proportion of
constant relative risk aversion, this difference would suggest
risky assets to wealth held by the individual.  is the error term and
women are more risk averse than men on average with respect
b0 ; b1 ; b2 ; and b3 are parameters to be estimated as are the vector
to asset allocation decisions. However, many empirical studies find
of m coefficients of variables that capture risk sharing and similar
evidence of decreasing relative risk aversion as wealth increases
expectations about the market, um , the vector of q coefficients of
(see, for instance, Friend & Blume, 1975; Jianakoplos & Bernasek,
household characteristics, cq , and the vector of n coefficients of vari-
1998; Ogaki & Zhang, 2001). If we expect individuals in Ghana also
ables that capture individual characteristics, wn .
to have decreasing relative risk aversion, it would mean that the
Femalek is a dummy variable, and the ln(wealthk ) variable is the
proportion of risky assets to wealth would increase as wealth
natural log of gross wealth. Given that alpha increases as wealth
increases. Indeed, the mean alphas by wealth quintile in Table 3
increases (Table 3 and Fig. 1), and that theoretically at subsistence
and the asset allocation decisions by men and women across the
level (or zero wealth) relative risk aversion is infinite and decreases
wealth distribution in Fig. 1 provide evidence of decreasing relative
as wealth increases, it is expected that individuals in Ghana have
risk aversion, suggesting that the difference between the value of
decreasing relative risk aversion, such that b1 is positive.
men and women’s assets may explain the large difference in men
Table 4 presents the descriptive statistics for the variables that
and women’s proportion of risky asset holdings to wealth (i.e.
capture risk sharing and similar expectations about the market,um ,
the difference in men and women’s mean alphas).
the household characteristics, cq , and the individual characteris-
tics, wn . To control for an absence of risk sharing and other formal
7
This is a low wealth sample. The GLSS suggests income is also low in Ghana. The and informal insurance mechanisms, whether the individual is part
mean annual household income in Ghana in 2005 was GH¢1,217 (GLSS5 2008). The of a household that reduced consumption due to an idiosyncratic
mean annual household income in Ghana in 2005 for the bottom quintile was shock between 2005 and 2010 is included inum . Less than three
GH¢728 or GH¢116 per capita. For the top quintile, the mean annual household
percent of individuals are from households that reduced food or
income was GH¢1,544 or GH¢397 per capita (GLSS5 2008).
8
When the authors include additional assets, such as the estimated value for
non-food consumption in order to cope with an idiosyncratic
human capital and the market value of a family’s home in the measure of risky assets, shock. This suggests that most individuals that faced idiosyncratic
the ratios are much higher (see Table 3 in Friend & Blume, 1975). shocks in the sample are part of a risk sharing arrangement or have
132 M. Hillesland / World Development 117 (2019) 127–137

Notes: Average asset holdings. Estimates from the author. The darker assets are risky-assets; the lighter are non-
risky assets.

Fig. 1. Asset allocation across the wealth distribution by sex.

access to other formal and informal mechanisms to insure expects future inheritance is also included as a control in wn . Sim-
themselves. ilarly, whether the individual holds a pension is included as a con-
Similar expectations about the market requires that individuals trol. Debt is trivial, but it is also included as a control.
have access to information, which may be provided through face-
to-face communication as well as through technology such as cell
phones. Cell phones are included as a control. Differences in face- 7. Results
to-face communication differ by proximity to other households;
to control of this, a variable for whether the household is located There are a number of individuals who do not hold wealth in
in a rural zone is also included as a control. Districts are included the form of formal and informal financial assets, registered busi-
as a proxy for local networks. nesses that can be sold, and commercial real estate the individual
Household characteristics, cq , include the size of the household owns and holds the right to sell. Alpha is observed only when an
and household wealth, which is proxied by agricultural plots. individual holds positive wealth (ak is undefined for an individual
While financial assets are not necessarily pooled in households in without wealth). This means the relative risk aversion measure, the
Ghana, we would expect asset allocation decisions to be influenced proportion of risky assets to wealth, ak , is incidentally truncated
by household wealth. Agricultural plots are an imperfect measure and may result in a specification error if not addressed. Studies
of household wealth, as it does not capture the size and quantity such as Jianakoplos and Bernasek (1998) do not include house-
of land and it overlooks many urban households. However, the holds below a certain wealth threshold and, thus, do not need to
majority of households in Ghana in 2010 were agricultural house- address truncation.10 This study does not use a threshold; instead,
holds.9 In addition, land is a major asset that makes up a large share to address truncation, a two-step Heckman selection model is used.
of households’ wealth. The first stage of the Heckman selection model estimates the likeli-
Individual characteristics, wn , include age, marital status, and hood an individual has positive wealth. The second stage estimates
whether the individual has children under five, and children ages the relative risk aversion measure, ak , while incorporating informa-
six to 11. To capture the individual’s position in the household, tion from the first.
an individual’s decision-making power is proxied by whether the The estimated coefficients of wealth, b1 , are positive and statis-
individual reported he or she could make the decision alone to tically significant, suggesting as wealth increases, ak , the relative
engage in employment or income earning activities, and whether risk aversion measure, increases (Table 5). It provides evidence
he or she can decide how the earnings are used. that individuals in this sample exhibit decreasing relative risk
Education is a non-marketable (i.e. cannot be directly traded) aversion in terms of asset allocation decisions, ceteris paribus. Ide-
asset and is either controlled for or included in the value of risky ally, instrumental variables would be used to address potential
assets in other studies (see, for example, Friend & Blume, 1975). biases of simultaneity of wealth and the risk aversion measure;
It is included as a control in this study. Agriculture is an important appropriate instruments were unavailable however. As such, the
livelihood for many rural households in Ghana and influences how estimates of wealth should be interpreted with some caution. Even
productive assets are allocated and, as such, is included as a con- so, the results are consistent with the theoretical expectations and
trol. Since expectations around inheritance may influence how an previous studies (e.g. Friend & Blume, 1975; Jianakoplos &
individual allocates his or her assets, whether an individual Bernasek, 1998). Fig. 2 presents the mean predicted ratios across
wealth.
9
Nearly five million households of 6.6 million were estimated to engage in
10
cropping activities in the 12 month period prior to the survey (based on the GLSS6 Studies in the United States, such as Jianakoplos and Bernesek (1998), often only
which was implemented from October 2012 to October 2013). include households with wealth greater than US$1000 in their analyses.
 M. Hillesland / World Development 117 (2019) 127–137 133

Table 4
Descriptive Statistics.

Women Men Total

(n = 1690) (n = 1341) (n = 3031)
a
Mean age 43.9 45.3 44.5
(16.2) (16.7) (16.5)
Number of children 0 to 5 years old (%)
None 62.0 62.8 62.3
1 22.5 20.6 21.6
2 12.3 12.8 12.5
3 or more 3.32 3.9 3.6
Number of children 6 to 11 years old (%)
None 62.3 63.5 63.1
1 20.8 20.8 20.8
2 12.5 12.0 12.3
3 or more 3.9 3.65 3.8
Marital Status (%)
In a monogamous union 45.6 54.7 50.6
In a polygamous union 6.7 8.0 7.7
In a consensual union 10.4 13.4 12.7
Never married 6.6 13.4 9.4
Divorced, widowed, or deserted 30.7 10.5 19.5
Education (%)
No education or attended some primary school only 60.4 40.4 51.5
Attended some junior secondary school or equivalent 30.4 40.0 34.7
Attended at least some senior secondary school or vocational or technical training 7.0 13.8 10.0
Attended at least some university, professional training, or other post senior secondary education 2.3 5.7 3.8
Made decision whether to be employed or pursue an income-generating activity alone and can make the 37.0 40.1 38.7
decision how to spend the earnings (%)
Expects an inheritance (%) 8.0 8.1 8.0

Women Men Total

(n = 1690) (n = 1341) (n = 3031)
b
Has a pension or other type of retirement (%) 5.1 13.6 8.9
Has rights over agricultural land (%) 11.8 29.8 19.8
Owns place of residence and it can be sold (%) 8.9 27.2 17.0
Occupation is agriculture, animal husbandry, forestry work, fishing, or hunting (%) 33.5 44.1 38.2
Owns a mobile phone (%) 39.6 61.7 49.4
Household reduced consumption due to an idiosyncratic shock anytime between 2005 and 2010 (%) 3.1 2.2 2.7
Household is in a rural setting 65.1 66.5 65.8
Mean household size 4.2 4.1 4.2
(2.6) (2.8) (2.7)
Average number of agricultural plots the household holds (owned land or family land)
None 70.9 69.0 70.0
1 16.8 16.9 16.9
2 7.2 8.0 7.6
3 2.9 3.7 3.3
4 1.4 1.3 1.4
5 0.5 0.8 0.6
6 0.3 0.4 0.3

Notes: Statistics are unweighted. Standard deviations are in parenthesis.

Controlling for men and women’s differing characteristics, men ak , is due primarily to the difference in average wealth. A Oaxaca-
and women do not have statistically different levels of relative risk Blinder decomposition technique is used to test this. This technique
aversion. b2 and b3 are not individually statistically significant and is typically used in the labor market literature, but it can be used to
a Wald-test suggests there is not joint statistical significance for study differences in any outcome variable.
the coefficients (chi-squared = 0.07).11,12 Women are not more rel- Eq. (7) is estimated separately for men and women with a two-
atively risk averse than men; rather it is men and women’s different step Heckman selection model for each equation.13 Using the
characteristics that contribute to the considerable difference in aver- results from these two equations and the pooled equation from
age ratios of risky assets to wealth, ak : Since there is evidence of above, the difference in men and women’s average ratios of risky

 


decreasing relative risk aversion, it is possible that the difference assets to wealth, E am
xm  E a f
x f , is divided into explained
k k k k
in men and women’s average proportion of risky assets to wealth,
and unexplained components. The explained component is the
sum of the product of the estimated coefficients for the pooled equa-
11
The difference is also not significant across the wealth distribution. tion except for the coefficient for the dummy sex variable and differ-
12
While the dependent variable is shares between 0 and 1, theoretically, the shares ence of men and women’s expected values. The unexplained
could be above and below 1 and 0. As such, a Heckman two step model is used as it
fits the data best. However, similar results were found using a fractional probit model
13
which can be used when the dependent variable is a fraction between zero and one. I use a pooled regression with group indicator so that unexplained factors due to
These are available by request. sex are not transferred to the coefficients in the explained components.
134 M. Hillesland / World Development 117 (2019) 127–137

Table 5 coefficients from the coefficients from the pooled equation plus the
Risk preference model: select variables. product of the expected values of women and the difference
Dependent variable: Alpha Heckman Two-Step between the female coefficients from the pooled equation
Model coefficients:
Natural log of financial wealth 0.1494**
   0 
(0.0047)
 m f
Female 0.0123 E am
k jxm
k  E a f
jx
k k
f
¼ ⏟ x  x bpooled
 bpooled
sex
(0.0439)
Explained component
Interaction of Natural log of financial wealth and 0.0064   f 0  
Female (0.0072) m0
þ⏟ bpooled
sex þx bm  bpooled þ x bpooled  b f
Age 0.0060**
(0.0032) Unexplained component
Age squared 0.0001 ð8Þ
(0.0000)
Married or in a consensual union 0.0341
(0.0209)
where bpooled , bm , and bf are vectors containing the intercepts and
Constant 0.8659
(0.1471)
slope parameters (um , wn ; and cq ) for the pooled equation, the esti-
mated equation consisting of only men, and the estimated equation
Lambda 0. 1203
(0.0915) consisting of only women.
Table 6 presents the results. The explained part of the outcome
Observations 3031
Censored observations 1575 differential is significant and accounts for nearly the total differ-
Chi-squared 2252.64 ence between men and women’s mean proportion of risky assets
to wealth. The unexplained part, which is the effect of the unob-
Notes: Standard errors are in parentheses. *p < .10, **p < .05. Marital status includes
all married individual and those in consensual unions. Additional controls are served predictors, is also statistically significant but significantly
number of children of the individual ages 0–5, number of children of the individual smaller.
ages 6–11, whether the individual attended at least some senior secondary school The greatest contribution to the difference in men and
or vocational or technical training, whether the individual attended at least some women’s average proportion of risky assets to wealth in both
university, professional training, or other post senior secondary education, deci-
sion-making power of employment and income, whether the individual holds debt,
models is the difference in average wealth. The contribution of
whether the individual owns agricultural land he or she can sell, whether the other predicators are minimal. The results suggest that, in
individual owns place of residence he or she has the right to sell, whether the Ghana, women are not more risk averse than men in terms of
individual’s occupation is in agriculture, whether the individual expects an inher- asset allocation decisions, but because men and women exhibit
itance, whether the individual has a pension or other retirement source, household
decreasing relative risk aversion, women’s lower average wealth
size, household is in rural setting, number of plots of land the household owns as a
proxy for household wealth, whether the household reduced consumption due to
an idiosyncratic shock anytime between 2005 and 2010, and district fix effects. The
selection equation predicts the probability the individual holds wealth. Variables
that would explain whether or not an individual has positive wealth and does not Table 6
impact allocation of wealth are not available. It is assumed the nonlinearity of the Oaxaca-Blinder decomposition of men and women’s proportion of risky assets to
selection equation creates enough exclusion restrictions. The full model is in the gross wealth (alpha).
Appendix.
Oaxaca-Blinder
decomposition
Differential
Male 0. 3673
(0.0168)
.8

Female 0.2085
(0.0148)
Difference 0.1588**
(0. 0223)
Decomposition
Explained 0.1208**
(0.0163)
Unexplained 0.0380*
(0.0176)
Individual contributions to the explained
component of the decomposition of some of the
variables
Natural log of wealth 0. 1375**
(0.0164)
Married or in a consensual union 0.0043*
0

(0.0022)
0 2000 4000 6000 8000 10000 Age 0.0037*
Wealth (Ghanaian cedis) (0.0020)
Individual owns agricultural land (where there is a 0.0093*
market and individual and has the right to sell) (0.0042)
Notes: Predicted values are based on average characteriscs. Individual owns the place of residence (where there is 0.0158**
a market and individual and has the right to sell) (0.0048)
Fig. 2. Predicted Measure of Relative Risk Aversion (Risky Assets to Wealth). Occupation is agriculture, animal husbandry, forestry 0.0116**
work, fishing, or hunting (0. 0115)

Notes: Standard errors are in parentheses. *p < .10, **p < .05. The estimates are based
on using a pooled Oaxaca-Blinder approach and two-step Heckman models for the
component includes the coefficient on the sex dummy variable (indi- pooled, female, and male equations. The estimations for the individual contribu-
cator of group membership) in the pooled equation plus the product tions to the explained component of the decomposition of all predictors are avail-
of the expected values of men and the difference between the male able by request.
 M. Hillesland / World Development 117 (2019) 127–137 135

than men’s means women invest less in risky assets proportional Table 7
to their overall wealth. Risk preferences: Full model.

Dependent variable: Alpha Heckman Two-

Step Model
8. Discussion and conclusion
Female 0.0113
(0.0439)
Without taking into account wealth, there is a statistically sig- Natural log of mean wealth 0.1494**
nificant difference in how men and women allocate their wealth (0.0047)
between risky and non-risky assets in households in Ghana, sug- Female * Natural log of mean wealth 0.0064
(0.0072)
gesting gender differences in relative risk aversion. However, Age 0.0060*
because men and women exhibit decreasing relative risk aversion, (0.0032)
men and women do not have systematically different risk prefer- Age squared 0.0000
ences when wealth differences and differences in other character- (0.0000)
Married or in a consensual union 0.0341
istics are taken into account.
(0.0209)
The results of this paper differ from the results in many empir- Attended at least some senior secondary school or 0.0171
ical studies in the United States, where women are found to be vocational or technical training (0.0244)
more risk averse than men. As one of the first studies to look at Attended at least some university, professional training, 0.0315
gender differences in risk aversion in terms of asset allocation deci- or other post senior secondary education (0.0383)
Made decision whether to be employed or pursue an 0.0079
sions within a developing country, the results provide support for income-generating activity alone and can make the (0.0228)
the idea that gender differences in risk preferences may vary by the decision how to spend the earnings
surrounding environment—including by country—as some of the Has rights over agricultural land 0.0654**
experimental literature suggests. If risk preferences vary by county, (0.0261)
Owns place of residence and it can be sold 0.0740**
differences found in men and women’s risk preferences are less
(0.0240)
likely to reflect innate sex differences in risk preferences, but Expects an inheritance 0.0523*
rather differences in constraints, opportunities, or responsibilities (0.0268)
men and women face. That is, any systematic difference in risk atti- Pension or other type of retirement 0.0213
tudes between men and women found in studies is likely not based (0.0539)
Holds debt 0.0803**
on sex, but rather influenced by gender norms and roles, which (0.0194)
vary by country. These norms and roles play out differently in dif- Number of children under five 0.0230*
ferent contexts due to different cultural and environmental pres- (0.0118)
sures. So while women may be more risk averse in terms of Number of children 6 to 11 years old 0.0233**
(0.0116)
asset allocation decisions than men in the United States, this
Occupation is agriculture, animal husbandry, forestry 0.0687**
may not necessarily be the case for men and women in Ghana. work, fishing, or hunting (0.0272)
Average number of agricultural plots the household holds (owned land or
family land)
Funding
1 0.0570**
(0.0262)
This work was supported the American University’s Vice Pro- 2 0.0666*
vost for Graduate Studies Doctoral Student Research Award (0.0351)
2011-2012 and the Association for Social Economics William R. 3 0.0470
(0.0488)
Waters Research Grant 2011. I also received support from The Gen- 4 plots or more 0.0286
der Asset Gap Project. (0.0670)
Number of household members 0.0010
(0.0046)
Conflicts of interest Household reduced consumption due to an idiosyncratic 0.0328
shock anytime between 2005 – 2010 (0.0494)
Owns a mobile phone 0.0413
The author confirms there are no known conflicts of interest
(0.0322)
associated with this publication and there has been no significant Rural 0.0475
financial support for this work that could have influenced its (0.0394)
outcome. Constant 0.8659**
(0.1471)
Dependent variable: Has positive wealth Probit model
Acknowledgements
Age 0.0233**
(0.0091)
I want to thank Caren Grown, Cheryl Doss, Amos Golan, and Age squared 0.0003**
Maria Sagrario Floro for their many reviews and helpful sugges- (0.0001)
tions. I would like to extend my gratitude to the Gender Asset Married or in a consensual union 0.2045**
Gap team and the Department of Economics at University of (0.0611)
Attended at least some senior secondary school or 0.0440
Ghana, Legon; I especially want to thank Abena Oduro, Louis vocational or technical training (0.0939)
Boakye-Yiadom, and Hellen Enyonam at the University of Ghana, Attended at least some university, professional training, 0.6792**
Legon, for hosting me in the summer of 2011. I also thank the or other post senior secondary education (0.2015)
reviewers for their thoughtful and thorough reviews. Made decision whether to be employed or pursue an 0.2987**
income-generating activity alone and can make the (0.0602)
decision how to spend the earnings
Has rights over agricultural land 0.0884
Appendix
(0.0830)

Table 7 (continued on next page)

136 M. Hillesland / World Development 117 (2019) 127–137

Table 7 (continued) Cárdenas, J. C., Dreber, A., Von Essen, E., & Ranehill, E. (2012). Gender differences in
competitiveness and risk taking: Comparing children in Colombia and Sweden.
Dependent variable: Alpha Heckman Two- Journal of Economic Behavior & Organization, 83(1), 11–23.
Step Model Charness, G., & Gneezy, U. (2012). Strong evidence for gender differences in risk
taking. Journal of Economic Behavior & Organization, 83(1), 50–58.
Owns place of residence and it can be sold 0.1530** Chiappori, P. A., Samphantharak, K., Schulhofer-Wohl, S., & Townsend, R. M. (2014).
(0.0760) Heterogeneity and risk sharing in village economies. Quantitative Economics, 5
Expects an inheritance 0.0688 (1), 1–27.
(0.0960) Croson, R., & Gneezy, U. (2009). Gender differences in preferences. Journal of
Pension or other type of retirement 1.3956** Economic Literature, 47(2), 448–474.
(0.1387) Deere, C. D., & Doss, C. (2006). The gender asset gap: What do we know and why
Holds debt 0.1553** does it matter? Feminist Economics, 12(1–2), 1–50.
(0.0644) Deere, C. D., Oduro, A. D., Swaminathan, H., & Doss, C. (2013). Property rights and
Number of children under five 0.0390 the gender distribution of wealth in Ecuador, Ghana and India. The Journal of
(0.0386) Economic Inequality, 11(2), 249–265.
Number of children of 6 to11 years old 0.0155 Dercon, S. (2005). Risk, insurance, and poverty: A review. In Insurance against
poverty (pp. 9–37). Oxford; New York: Oxford University Press.
(0.0384)
Dercon, S., & Christiaensen, L. (2011). Consumption risk, technology adoption and
Occupation is agriculture, animal husbandry, forestry 0.2273**
poverty traps: Evidence from Ethiopia. Journal of Development Economics, 96(2),
work, fishing, or hunting (0.0766)
159–173.
Average number of agricultural plots the household holds (owned land or Dercon, S., & Krishnan, P. (2000). In sickness and in health: Risk sharing within
family land) households in rural Ethiopia. Journal of Political Economy, 108(4), 688–727.
1 0.1046 De Weerdt, J., & Dercon, S. (2006). Risk-sharing networks and insurance against
(0.0811) illness. Journal of Development Economics, 81(2), 337–356.
Doss, C. (2001). Is risk fully pooled within the household? Evidence from Ghana.
2 0.0541
Economic Development and Cultural Change, 50(1), 101–130.
(0.1105)
Doss, C., McPeak, J., & Barrett, C. (2008). Interpersonal, intertemporal and spatial
3 0.3304**
variation in risk perceptions: Evidence from East Africa. World Development, 36
(0.1553) (8), 1453–1468.
4 plots or more 0.0631 Eckel, C. C., & Grossman, P. J. (2008). Men, women and risk aversion: Experimental
(0.1869) evidence. In C. Plott & V. Smith (Eds.), Handbook of experimental economics
Number of household members 0.0205 results, 1 (pp. 1061–1073). Oxford: North-Holland.
(0.0149) Fafchamps, M., & Gubert, F. (2007). The formation of risk sharing networks. Journal
Household reduced consumption due to an idiosyncratic 0.0745 of Development Economics, 83(2), 326–350.
shock anytime between 2005 and 2010 (0.1590) Fafchamps, M., & Lund, S. (2003). Risk-sharing networks in rural Philippines. Journal
Owns a mobile phone 0.4902** of Development Economics, 71(2), 261–287.
(0.0581) Fafchamps, M., Udry, C., & Czukas, K. (1998). Drought and saving in West Africa: Are
Rural 0.1738 livestock a buffer stock? Journal of Development Economics, 55(2), 273–305.
Fletschner, D., Anderson, C. L., & Cullen, A. (2010). Are women as likely to take risks
(0.1250)
and compete? Behavioural findings from Central Vietnam. The Journal of
Female 0.1888**
Development Studies, 46(8), 1459–1479.
(0.0572)
Friend, I., & Blume, M. E. (1975). The demand for risky assets. American Economic
Constant 0.4725 Review, 65(5), 900–922.
(0.3584) Goldstein, M. (1999). Chop time, no friends: Intrahousehold and individual insurance
Lambda 0.1203 mechanisms in Southern Ghana. University of California, Berkeley, mimeo.
(0.0916) Goldstein, M., de Janvry, A., & Sadoulet, E. (2004). Is a friend in need a friend indeed?
Inclusion and exclusion in mutual insurance networks in Southern Ghana. In S.
Observations 3031 Dercon (Ed.), Insurance against poverty (pp. 9–37). Oxford; New York: Oxford
Censored observations 1,575 University Press. https://doi.org/10.1093/0199276838.003.0012.
Chi-squared 2252.64 Gneezy, U., Leonard, K. L., & List, J. A. (2009). Gender differences in competition:
Evidence from a matrilineal and a patriarchal society. Econometrica, 77(5),
Notes: Standard errors are in parentheses. *p < .10, **p < .05. The model also includes 1637–1664.
district fix effects. The selection equation predicts the probability the individual Grimard, Franque (1997). Household consumption smoothing through ethnic ties:
holds wealth. Variables that would explain whether or not an individual has posi- Evidence f from Cote d’Ivoire. Journal of Development Economics, 53(2), 391–422.
tive wealth and does not impact allocation of wealth are not available. While not Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R.
ideal, it is assumed the nonlinearity of the selection equation creates enough (2001). In search of homo economicus: Behavioral experiments in 15 small-
exclusion restrictions. scale societies. American Economic Review, 91(2), 73–78.
Jianakoplos, N. A., & Bernasek, A. (1998). Are women more risk averse? Economic
Inquiry, 36, 620–630.
Khachatryan, K., Dreber, A., Von Essen, E., & Ranehill, E. (2015). Gender and
preferences at a young age: Evidence from Armenia. Journal of Economic
Behavior & Organization, 118, 318–332.
References Lee, J., & Sawada, Y. (2010). Precautionary saving under liquidity constraints:
Evidence from rural Pakistan. Journal of Development Economics, 91(1), 77–86.
Ambrus, A., Mobius, M., & Szeidl, A. (2014). Consumption risk-sharing in social Ligon, E., Thomas, J. P., & Worrall, T. (2002). Informal insurance arrangements with
networks. The American Economic Review, 104(1), 149–182. limited commitment: Theory and evidence from village economies. The Review
Antonopoulos, R., & Floro, M. S. (2005). Asset Ownership along gender lines: Evidence of Economic Studies, 69(1), 209–244.
from Thailand Working Paper No. 418. NY: Levy Economics Institute. Lintner, J. (1970). The market price of risk, size of market and investor’s risk
Arano, K., Parker, C., & Terry, R. (2010). Gender-based risk aversion and retirement aversion. The Review of Economics and Statistics, 52(1), 87–99.
asset allocation. Economic Inquiry, 48(1), 147–155. Mazzocco, M., & Saini, S. (2012). Testing efficient risk sharing with heterogeneous
Arrow, K. J., (1965). Aspects of a theory of risk bearing, Yrjo Jahnsson Lectures, risk preferences. American Economic Review, 102(1), 428–468.
Helsinki. Reprinted in Essays in the theory of risk bearing (1971). Meyer, D. J., & Meyer, J. (2006). Measuring risk aversion. Foundation and Trends in
Bajtelsmit, V. L., & VanDerhei, J. L. (1997). Risk aversion and pension investment Microeconomics, 2(2), 107–203.
choices. In M. S. Gordon, O. S. Mitchell, & M. M. Twinney (Eds.), Positioning Morduch, J. (2005). Consumption smoothing across space: Testing theories of risk-
pensions for the twenty-first century. University of Pennsylvania Press. sharing in the ICRISAT study region of South India. In S. Dercon (Ed.), Insurance
Bernasek, A., & Shwiff, S. (2001). Gender, risk, and retirement. Journal of Economic against poverty. Oxford: Oxford University Press.
Issues, 35(2), 345–356. Mossin, J. (1966). Equilibrium in a capital asset market. Econometrica: Journal of the
Booth, A., Cardona-Sosa, L., & Nolen, P. (2014). Gender differences in risk aversion: Econometric Society, 768–783.
Do single-sex environments affect their development? Journal of Economic Murgai, R., Winters, P., Sadoulet, E., & De Janvry, A. (2002). Localized and incomplete
Behavior & Organization, 99, 126–154. mutual insurance. Journal of Development Economics, 67(2), 245–274.
Booth, A. L., & Nolen, P. (2012). Gender differences in risk behaviour: Does nurture Nelson, Julie A. (2015). Are women really more risk-averse than men? A re-analysis
matter? The Economic Journal, 122(558), F56–F78. of the literature using expanded methods. Journal of Economic Surveys, 29(3),
Budd, A., & Litzenberger, R. H. (1972). The market price of risk, size of market and 566–585.
investors’ risk aversion: A comment. The Review of Economics and Statistics, 54 Nelson, J. (2016). Not-so-strong evidence for gender differences in risk taking. Feminist
(2), 204–206. Economics, 22(2), 114–142. https://doi.org/10.1080/13545701.2015.1057609.
 M. Hillesland / World Development 117 (2019) 127–137 137

Oduro, A., & Doss, C. (2018). Changing patterns of wealth distribution: Evidence Schubert, R., Brown, M., Gysler, M., & Brachinger, H. W. (1999). Financial decision-
from Ghana. The Journal of Development Studies, 54(5), 933–948. https://doi.org/ making: Are women really more risk-averse? The American Economic Review, 89
10.1080/00220388.2018.1430769. (2), 381–385.
Ogaki, M., & Zhang, Q. (2001). Decreasing relative risk aversion and tests of risk Sundén, A., & Surette, B. (1998). Gender differences in the allocation of assets in
sharing. Econometrica, 69(2), 515–526. retirement savings plans. American Economic Review Papers and Proceedings, 88,
Park, Albert (2006). Risk and household grain management in developing countries. 207–211.
The Economic Journal, 116(514), 1088–1115. Townsend, R. M. (1994). Risk and insurance in village India. Econometrica: Journal of
Pratt, J. W. (1964). Risk aversion in the small and in the large. Econometrica: Journal the Econometric Society, 539–591.
of the Econometric Society, 122–136. Townsend, R. M. (1995). Consumption insurance: An evaluation of risk-bearing
Quisumbing, A. (2010). Do men and women accumulate assets in different ways. systems in low-income economies. The Journal of Economic Perspectives, 9(3),
Evidence from rural Bangladesh. Background paper prepared for the Food and 83–102.
Agricultural Organization of the United Nations (FAO) State of Food and Agriculture. Vanderpuye-Orgle, J., & Barrett, C. B. (2009). Risk management and social visibility
Riley, W. B., Jr, & Chow, K. V. (1992). Asset allocation and individual risk aversion. in Ghana. African Development Review, 21(1), 5–35.
Financial Analysts Journal, 32–37. Zimmerman, F. J., & Carter, M. R. (2003). Asset smoothing, consumption smoothing
Sawada, Y., Nakata, H., & Kotera, T. (2017). Self-production, friction, and risk sharing and the reproduction of inequality under risk and subsistence constraints.
against disasters: Evidence from a developing country. World Development, 94, Journal of Development Economics, 71(2), 233–260.
27–37.