For Good Measure Read online

  While much is known about the gender wage gap,2 comparatively little is known about the gender asset or wealth gap, whether among couples (i.e., the intra-household distribution of wealth) or for the population as a whole. This is largely because data on asset ownership collected through household surveys—including in large-scale wealth surveys—have tended to be at the household rather than the individual level, constraining gender analysis. Analyses concerned with gender inequality have been limited to the study of household types, i.e., male or female sole-headed households in comparison to married couples.3 Gender analyses of households composed of couples are sometimes attempted by focusing on the sex of the respondent, who is typically the best informed on financial matters;4 but since wealth data is collected at the household rather than the individual level, such analyses do not shed light on the intra-household distribution of assets. The assumption that, in married couples, all assets are pooled and the benefits shared among all household members—i.e., the assumption of a unitary household—has prevailed for too long. However, in most legal systems, property rights are ceded to individuals, not households. As Doss, Grown, and Deere (2008) argue, analyses of “household wealth” ignore institutional frameworks governing individual property rights, as defined by marital regimes, inheritance laws, and social norms.

  Whether asset ownership is in fact pooled in marriage (and consensual unions) largely depends on a country’s default marital regime—the rules governing how property acquired prior to and during the marriage and how inheritances are treated should the union be dissolved (Deere and Doss, 2006). For example, under the prevailing separation of property regime in many African, Middle Eastern, and South Asian countries,5 all property acquired by individuals prior to or after marriage, including any inheritances received, are considered to be their own individual property, i.e., should a union dissolve, each person leaves with only their own personal property. In some countries that have traditionally had this default marital regime, such as the United Kingdom, divorce legislation reform has subsequently modified this outcome, so that property acquired during the marriage with the earnings of either spouse is pooled and divided equally. In this case, the outcome resembles partial community property, under which property acquired prior to marriage and any inheritances are considered individual property, while property acquired during the marriage is split equally among the spouses upon its dissolution.

  The main point is that institutional parameters shape the accumulation of wealth by individuals, and must be duly accounted for in data collection efforts and economic analysis. As an illustration, in Ecuador—where partial community property prevails and inheritance norms and practices are equitable—married women own 44% of couple wealth; conversely, in both Ghana and Karnataka, India—characterized by the separation of property marital regime as well as by male bias in inheritance—married women own only 19% and 9%, respectively, of couple wealth (Deere et al., 2013).

  Measuring the Gender Wealth Gap

  As mentioned above, when data on asset ownership is collected in household surveys, it has tended to be at the household rather than the individual level, constraining gender analysis. Among the large-scale wealth surveys included in the Luxembourg Wealth Study, for example, only the German Socio-Economic Panel collects data on individual ownership of a broad range of physical and financial assets, allowing analysis of the intra-household distribution of wealth (Grabka, Marcus, and Sierminska, 2015). Two other surveys collect partial data on what belongs to individuals: the United Kingdom Wealth and Assets Survey (on financial assets and liabilities, pension wealth, and real estate) and the Italian Survey of Household Income and Wealth (on real estate).6

  The multi-purpose surveys most frequently carried out in developing countries are the Living Standard Measurement Study (LSMS) and the Demographic and Health Surveys (DHS) Program. An analysis of a sample of 72 LSMS questionnaires across six world regions for the mid-2000s revealed that the great majority of countries collected data on household ownership of housing, land, livestock, and major consumer durables. Only 21% of these, however, collected data on who in the household owned the residence, 17% on who owned the land, and 14% on who owned nonagricultural businesses (Doss, Grown, and Deere, 2008). A subsequent analysis of 167 household survey questionnaires for 23 Latin American and Caribbean countries found that only 23 questionnaires, for 11 countries, collected gender-disaggregated ownership information on at least one asset, most frequently for the main residence (Deere, Alvarado, and Twyman, 2012). Since 2009, the DHS has included questions asking surveyed individuals whether they are owners or co-owners of the main residence and land.7 Thus, while it is increasingly possible to measure gender gaps with regard to specific assets, large lacunas remain in terms of being able to estimate total individual wealth and the gender wealth gap.

  One of the reasons why progress on measuring individual level wealth has been slow has been because of methodological concerns, such as whether reliable data on the valuation of assets can be elicited from respondents. Other issues include who should be interviewed in an asset survey, how ownership should be defined, how the value of assets should be measured, and whether all assets need be included in wealth estimates.8 The Gender Asset Gap Project was launched in 2009 to explore whether it was feasible to collect detailed, gender-disaggregated wealth data in developing countries, and to study the potential gender biases in the methods employed to do so. For this purpose, national-level household surveys were carried out in 2010 in Ecuador and Ghana and at the state level in India (Doss et al., 2011 and 2014). Two other projects are currently investigating some of these questions: the Methodological Experiment on Measuring Asset Ownership from a Gender Perspective (MEXA),9 and Evidence and Data for Gender Equality (EDGE).10

  The issue of who should be interviewed in a household wealth survey aiming to collect individual-level data has also been raised with respect to household income surveys. There is growing consensus that direct reporting is superior to reporting by proxy (where one household member reports on the income or assets of all other household members rather than just on their own resources).11 The MEXA report, for example, recommends that household surveys move beyond their reliance on asking a single respondent (whether the household head or “the most knowledgeable” person in the household) to include multiple respondents, beginning with the members of the main couple, if not all adults (Kilic and Moylan, 2016).

  The issue of how ownership should be defined has been raised primarily in the context of asset information,12 since there are various ways that it can be measured: reported ownership, documented ownership, or one or several of the components of the bundle of property rights. Documented ownership (having a deed or other form of documentation) tends to be the most secure form of ownership. However, housing and land titles are not always widely available in developing countries. To mitigate this problem, many recent wealth surveys first ask about reported ownership and then ask about documentation and, if available, whose names are on the document. In contexts where private property rights are not well defined, it may be useful to ask about the full range of rights separately (i.e., to use, to lease, to use as collateral, to sell or bequeath) to explore “effective rights.”13

  The valuation of assets is commonly measured by asking respondents what an asset could be sold for today in its present condition (potential sales price or realization value) and/or its replacement cost. Household income surveys often ask about the rental value of immovable property, whereby the present value of the asset can be estimated. All of these measures assume the existence of rental or sales markets for assets, although in developing countries some of these may be particularly thin. The Gender Asset Project, nonetheless, found that the incidence of nonreporting on these different value measures was relatively low (Doss et al., 2013). Another concern is whether knowledge about asset markets and hence values is gender biased, leading to over- or under-reporting depending on who is interviewed. Nevertheless, this is difficult
to determine in the absence of a benchmark such as administrative data on immovable property, which is rarely available in most developing countries (Doss et al., 2013; Deere and Catanzarite, 2016).

  What Can Be Done to Obtain Better Estimates of the Gender Wealth Gap?

  Collecting data on the ownership and value of all assets is a time-consuming process, leading to the question of whether there are any short cuts, particularly if an asset module is to be added to a multi-purpose household survey. The Gender Asset Gap Project, which collected data on ownership and value down to the last chicken in three developing countries, suggests that, as a minimum, data should be collected on all immovable property (i.e., the main residence, agricultural land, and other real estate), businesses, and financial assets. In the three countries covered by the project, immovable property and businesses ranged from 82% (Ghana) to 93% (India) of total household physical wealth.14 Nonetheless, the composition of wealth may vary across the wealth distribution, with consumer durables making up a large share of wealth among the poorest quintile. Thus, the range and number of assets that need to be included in a wealth survey depend on its specific objectives.

  Finally, for comparative purposes it is important for household wealth surveys to collect data on the marital regime—i.e., whether couples were married under civil, religious, or community law; and if the former, under what particular option if various are available. Moreover, to enrich gender analysis, it is important to collect data on how assets were acquired, who decides on their use, and—for potential use as an instrumental variable—on whether a respondent’s parents owned immovable property.

  Besides allowing analysis of the intra-household distribution of resources, the questions that gender-disaggregated wealth data could answer are many. Examples of the types of questions that could be analyzed include: How large is the gender wealth gap? Does it vary by countries’ level of economic development or across the distribution in any systematic way? To what extent is the gender wealth gap conditioned by the institutional framework of each country, specifically marital and inheritance regimes? Are there differences in magnitude between the gender wealth gap among couples and the population as a whole, and how does this relate to increases in the divorce rate and specific divorce legislation? Does the composition of assets owned by men and women differ? What are the sources of the gender wealth gap and how much of it is explained by the observable characteristics of men and women?

  Conclusions

  The different aspects of inequality discussed in this chapter have clear implications for measurement and statistics, and these have been highlighted in each of the sections. But they also raise important policy questions. For example, in arguing for the need to have measures of intra-household inequality with respect to income, consumption, and wealth, one might mention that in many countries social assistance is based on various kinds of household means tests, excluding from support those members of nonpoor households who are individually poor and get a small share of the household income and wealth.

  At the same time, an exclusive focus on vertical inequality, to the exclusion of inequalities across broadly defined groups based on, for example, ethnicity could mislead policy-makers in situations where vertical inequality is falling but horizontal inequality is rising, thus stoking social instability.

  As a final example (linking the gender wealth gap with intra-household allocation, since wealth affects bargaining power within the household), neglect of gender-specific wealth inequalities will mislead policy-makers on the final beneficiaries of transfer and other schemes targeted at the household level. Particularly in developing countries, but also in developed countries, a focus on horizontal inequality, intra-household inequality, and the gender wealth gap will pay policy dividends.

  Notes

  1. http://ec.europa.eu/eurostat/documents/3859598/5901513/KS-RA-07-006-EN.PDF.

  2. See Weichselbaumer and Winter-Ebmer (2005) for a meta-analysis of the gender wage gap internationally; World Bank (2012) for a good summary of findings for developing countries; and the discussion in Chapter 5.

  3. See Schmidt and Sevak (2006) for such an analysis with the US Panel Study of Income Dynamics; and Yamokoski and Keister (2006) for an analysis utilizing the US National Longitudinal Survey of Youth.

  4. See Neelakantan and Chang (2010) for such an analysis with the US Health and Retirement Survey and Ruel and Hauser (2013) for a similar study utilizing the Wisconsin Longitudinal Study.

  5. World Bank (2012, p. 162) provides a summary of the default marital regimes in many developed and developing countries.

  6. The content of these wealth surveys is described in www.lisdatacenter.org/frontend#/home.

  7. See https://dhsprogram.com/.

  8. OECD (2013) discusses some of the general issues. See Doss, Grown, and Deere (2008) and Doss et al. (2011) for some of the initial discussions of these issues from a gender perspective.

  9. MEXA was implemented in Uganda by the Development Data Group of the World Bank, with support from EDGE and the World Bank Living Standard Measurement Study–Integrated Surveys on Agriculture. See Kilic and Moylan (2016) for preliminary results on the experiment, which featured 5 survey treatments over 13 asset groups.

  10. EDGE is a project of the UN Statistical Division and UN Women in collaboration with the African Development Bank, the Asian Development Bank, FAO, the OECD, and the World Bank. It aims to develop guidelines for the collection of individual-level data on asset ownership and entrepreneurship, and is piloting data collection in seven countries.

  11. See Fisher, Reimer, and Carr (2010) on how men tend to understate the income of their wives compared with wives’ reports, hence potentially underestimating household income.

  12. In household income or employment surveys, it is usually assumed that the person who earns the income “owns” it in the sense of controlling its use. However, there is growing evidence from developing countries that women, in particular, may not always control the income they earn. See World Bank (2012), Fig. 2.9.

  13. On measuring land ownership in Africa, see Doss et al. (2015). An alternative, pursued in the MEXA experiment in Uganda, is to focus on economic ownership, defined as who keeps the proceeds of a sale should an asset be sold.

  14. The remaining share corresponds to livestock, agricultural equipment, and a broad range of consumer durables, including vehicles. Financial assets range from 2% (Ecuador) to 5% (Ghana) of gross household wealth (Doss et al., 2013).

  References

  Akerlof, G.A. and R.E. Kranton (2000), “Economics and identity,” Quarterly Journal of Economics, Vol. 115(3), pp. 715–753, https://doi.org/10.1162/003355300554881.

  Alderman, H. et al. (1995), “Unitary versus collective models of the household: Is it time to shift the burden of proof?,” World Bank Research Observer, Vol. 10(1), pp. 1–19.

  Badgett, M.V.L. and H.L. Hartmann (1995), “The effectiveness of equal employment opportunity policies,” in Simms, M.C. (ed.), Economic Perspectives on Affirmative Action, University Press of America, Washington, DC.

  Bourguignon, F. (2003), “The growth elasticity of poverty reduction: Explaining heterogeneity across countries and time periods,” in Eicher, T.S. and S.J. Turnovsky (eds.), Inequality and Growth: Theory and Policy Implications, The MIT Press, Cambridge, MA.

  Broman, C. (1997), “Race-related factors and life satisfaction among African Americans,” Journal of Black Psychology, Vol. 23(1), pp. 36–49.

  Brown, G., A. Langer, and F. Stewart (eds.) (2012), Affirmative Action in Plural Societies: International Experiences, Palgrave, London.

  Cederman, L.E., N.B. Weidmann, and K.S. Gleditsch (2011), “Horizontal inequalities and ethno-nationalist civil war: a global comparison,” American Political Science Review, Vol. 105(3), pp. 478–495.

  Collier, P. and A. Hoeffler (2004), “Greed and grievance in civil war,” Oxford Economic Papers, Vol. 56(4).

  Deere, C.D., G. Alvarado, and J. Twyman (2012), “Gender inequality in
asset ownership in Latin America: Female owners versus household heads,” Development and Change, Vol. 43(2), pp. 505–530.

  Deere, C.D. and Z. Catanzarite (2016), “Measuring the intra-household distribution of wealth in Ecuador: Qualitative insights and quantitative outcomes,” in Lee, F. and B. Conin (eds.), Handbook of Research Methods and Applications in Heterodox Economics, Edward Elgar Publishing, Cheltenham, pp. 512–534.

  Deere, C.D. and C. Doss (2006), “The gender asset gap: What do we know and why does it matter?,” Feminist Economics, Vol. 12(1&2), pp. 1–50.

  Deere, C.D. et al. (2013), “Property rights and the gender distribution of wealth in Ecuador, Ghana and India,” Journal of Economic Inequality, Vol. 11(2), pp. 249–265.

  Doss, C. et al. (2015), “Gender inequality in ownership and control of land in Africa: Myth and reality,” Agricultural Economics, Vol. 46(3), pp. 403–434.

  Doss, C. et al. (2014), “The gender asset and wealth gaps,” Development, Vol. 57(3–4), pp. 400–409.

  Doss, C. et al. (2013), “Measuring personal wealth in developing countries: Interviewing men and women about asset values,” Gender Asset Gap Project Working Paper, No. 15, November, https://sites.google.com/view/genderassetgap.

  Doss, C. et al. (2011), Lessons from the Field: Implementing Individual Asset Surveys in Ecuador, Ghana, India and Uganda, Indian Institute of Management, Bangalore, https://sites.google.com/view/genderassetgap.

  Doss, C., C. Grown, and C.D. Deere (2008), “Gender and asset ownership: A guide to collecting individual-level data,” World Bank Policy Research Working Paper, No. WPS4704, World Bank, Washington, DC, https://openknowledge.worldbank.org/bitstream/handle/10986/3468/WPS4704.pdf?sequence=6.