Policy Research Working Paper 10733 Gender and Fiscal Policy A Methodological Proposal and Its Application to Jordan and Armenia Jon Jellema Caren Grown Alan Fuchs Matthew Wai-Poi Sailesh Tiwari Mariano Sosa Poverty and Equity Global Practice March 2024 Policy Research Working Paper 10733 Abstract Fiscal policies affect households and individuals in a variety incidence analysis framework, enabling the implicit valua- of ways. Even though these effects are likely to be different tion of unpaid work that is typically undertaken by women for men and women, conventional tools of fiscal incidence on activities such as cooking, cleaning, and caring for chil- analysis are typically unable to capture these gender differ- dren and the elderly. Applying this methodology to the cases ences. Using a particular type of incidence analysis known of Jordan and Armenia—two countries with very different in the literature as the Commitment to Equity framework, approaches to fiscal policy and cultural norms around the this paper proposes a methodology to overcome this chal- economic and social roles of men and women—the paper lenge. A particular novelty the paper introduces is the also highlights some of the insights that this engendered explicit incorporation of social reproduction into the fiscal perspective could add to standard fiscal incidence analysis. This paper is a product of the Poverty and Equity Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at j.jellema@worldbank.org; afuchs@worldbank.org; mwaipoi@worldbank.org and stiwari@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Gender and Fiscal Policy – A Methodological Proposal and Its Application to Jordan and Armenia* Jon Jellema Caren Grown Alan Fuchs Matthew Wai-Poi Sailesh Tiwari Mariano Sosa JEL Classifications: H22, H23, J16 Keywords: taxes, transfers, incidence analysis, gender * This paper was produced by the Global Solutions Group for Fiscal and Social Policies and Equity of the Poverty and Equity Global Practice of the World Bank. The work was done under the managerial oversight of Benu Bidani (Practice Manager, Poverty and Equity Global Practice Unit) and the overall guidance of Carolina Sanchez Paramo (Global Director, Poverty and Equity) and Hana Brixi (Global Director, Gender). Several rounds of helpful comments and feedback were received throughout the process from a technical advisory group that included David Coady, Susan Himmelweit, Abena Oduro, Nora Lustig, Daniela Casale, Janet Stotsky, and Stephen Younger. The authors also gratefully acknowledge the financial support received from the Hewlett Foundation and the Global Tax Program (GTP). 1 Introduction Public revenue and expenditure policies, or taxes and transfers, shape individuals’ behaviors, consumption, savings and investment patterns, participation in labor markets, and access to social safety nets and other publicly provided goods and services. Poor and vulnerable women, men, and children make decisions about how much to work outside the home for pay or inside the home without pay, about how much to save and spend, about what to buy, and about which services are essential for which household members. These decisions are often made under severe financial, social, and cultural constraints, which fiscal policies can either ameliorate or exacerbate. Fiscal incidence analysis (FIA), which accounts for the receipt of public expenditures (transfers) and contributions to public revenues (taxes), can demonstrate how policies affect welfare, purchasing power, and wealth, and can identify policies that are most effective in redistributing public resources to those who need it most. While FIA has become an integral component for assessing the equitable nature of fiscal systems, most FIA work does not attempt to describe how fiscal policy differentially impacts men, women, and children. Incidence accounting at the individual level is challenging for two reasons. First, fiscal policies either support or hinder activities a household member pursues, rather than individual household members or their characteristics. Second, when tax burdens and transfer benefits are linked directly to individuals or their characteristics, they may or may not be shared broadly among all household members. For instance, a woman receiving a cash transfer may not necessarily control how that additional income is used. Analogously, a man’s statutory liability for income taxes may not mean he alone will bear the entire economic burden of any income lost to tax payments. Gender norms influence the activities in which men and women participate. Fiscal policy either supports economic and social activities through public expenditures or makes such activities more expensive by imposing taxes. For example, Barnett and Grown (2004) identify four differences in economic activity which condition the impact of fiscal policies: gender differences in paid formal and informal employment and wages, gender differences in unpaid care work or participation in social reproduction, gender differences in asset ownership, and gender differences in consumption behavior. The imposition of taxes on the wages of those who are formally employed or the support of subsidy expenditures encouraging consumption of basic commodities, for example, link fiscal policy directly to economic and gender-based inequality. International frameworks have also implicitly connected fiscal policy to gender inequality. For example, the Sustainable Development Goals 2030 (SDGs) have listed gender equality (SDG 5) and reduced inequality (SDG 10) as separate goals; while SDG Target 5.4, “Recognize and value unpaid care and domestic work through the provision of public services, infrastructure and social protection policies...” and SDG Target 10.4, “Adopt policies, especially fiscal, wage and social protection policies, and progressively achieve greater equality” are listed separately. Both dimensions of equality will likely be addressed through a common set of fiscal and social policies.1 1. Social and fiscal policies supporting goals and activities in targets 5.4 and 10.4 will be funded at least partially by countries’ own revenues, which also links these SDG targets explicitly to tax and revenue collection policy. While grants, donor aid, and concessional loans will partially fund or otherwise support public policy, the contribution from domestic resources is expected to be equally large if not larger (Lagarde and Gaspar 2018). For an empirical look at how increased mobilization of domestic resources may weaken the equity impacts of expenditures, see Lustig (2018). 2 This paper makes two main contributions to the literature on the equity impacts of fiscal policy. First, it uses the Commitment to Equity (CEQ) framework to bring taxes and benefits together to examine the differential impacts of fiscal systems on activities that men and women undertake at the household level.2 Developed in 2008, the CEQ framework examines how taxes and spending (including indirect subsidies, education, and health expenditures) affect income inequality and poverty at the country level. Since the development of this methodology, it has been applied to more than 105 countries across the world.3 Second, this paper explicitly brings social reproduction into the FIA framework. Social reproduction is the production and maintenance of human capabilities, often undertaken by women, such as cooking, cleaning, and caring for children, the elderly, and other dependents; subsistence activities like fetching water; and the provision of cultural, social, physical, and educational guidance necessary for the household members to thrive and increase their capability. Social reproduction is crucial for the development of human capabilities; it is therefore a necessary condition for the short- and long-term growth of economies and the social and cultural institutions supporting the economies. We treat social reproduction as part of a household’s total social and economic activity. Social reproduction – how it is accomplished and who accomplishes it – is therefore subject to the same type of constraints that inform other household decisions. We suggest a household-level typology to understand which households are involved in which activities critical for social reproduction and, based on a large literature about the gender division of paid and unpaid work,4 make assumptions about who in those households is responsible for those activities. This analytical choice allows estimation of the likely impact of fiscal policies on individuals – based on their household roles – to incorporate a deeper understanding of gender inequalities in economic life. The rest of the paper is organized as follows. Section 2 provides a summary of studies that have investigated the differential impacts of fiscal policies on men and women and situates this paper’s contributions within that literature. Section 3 includes a discussion of gender, social reproduction, and the market economy and outlines the various ways in which a gender lens might feature within a fiscal incidence analysis. Section 4 introduces the CEQ and proposes a methodology for practitioners in examining the different ways in which fiscal policies support or undermine equality between men and women. Section 5 presents the application of the proposed methodology to the cases of Jordan and Armenia, highlighting some of the main results. The final section concludes with some reflections on the application of this methodology to other countries and the kind of data that could be useful to extend it further. 2. The CEQ approach was developed by the Commitment to Equity Institute (CEQ Institute) at Tulane University. The methodology, imple- mentation guidelines, applications, and software of the CEQ approach can be found in Lustig, Nora, ed. 2018. Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty. Brookings Institution Press, pp. 3–55. As such the methodology is a registered trademark. 3. The World Bank and the CEQ Institute are responsible for approximately two-thirds of the existing or ongoing CEQ assessment exercises while the CEQ Institute and other partners are responsible for the remaining one-third. 4. See Charmes (2019, 2022); Rubiano and Vollaz (2019); Dinkelman and Ngai (2022); and Pailhe, Solaz, and Stanfors (2020). 3 2 Gender Dimensions in Fiscal Policy Early studies investigating the differential impacts of fiscal policy across sex and gender5 looked at the expenditure or tax sides of fiscal policy in isolation. For example, Selden and Wasylenko (1995), Demery (1996), Sahn and Younger (2000), and Pritchett and Filmer (1999) examined the incidence of education and health expenditures separately for men and women using a sample of 57 Demographic and Health Surveys (DHS) from 41 countries. Glick, Saha, and Younger (2004) developed a more systematic attempt to “engender” benefit incidence analysis, examining gender gaps in benefits from health, education, and infrastructure (water supply, electricity, and roads).6 Using household survey data from Bulgaria, Ghana, Jamaica, Madagascar, Mauritania, Pakistan, Peru, Uganda, and Viet Nam, they calculated the share of benefits as the proportion of males/females in any given quin- tile that benefit from a particular public service in these sectors and documented large gender gaps in access to these services. Grown and Valodia (2010) investigated the differential burden of taxes across genders via household types.7 For example, households might have male, female, or joint income earners; or male or female adult majorities; or younger- or older-age dependents (or both). Tax incidence8 is estimated and compared across the different household types and across the post-tax expenditure distribution. More recently, Komatsu and others (2021) used sociological, cultural, and institutional information regarding the role of women in agricultural production in Ethiopia to better illuminate why a statutorily neutral agricultural income tax and rural land use fee nonetheless created larger economic burdens for households distinguished by having female agricultural landowners. Greenspun (2019) used the Grown and Valodia (2010) household categories but brought both the expenditure and tax sides together within the CEQ assessment FIA framework.9 The CEQ assessment architecture is fully generalizable to any social, economic, or fiscal context. More recently, Ambel, Wondigmagegn, and Yonis (2022) estimated the differential impacts of fiscal policy within a CEQ assessment framework for Ethiopia. The approaches described above capture certain aspects of gender relations at the household level: it is the household that is classified according to type, and it is across households of different types that the incidence of fiscal policy is compared. Estimating a complete intrahousehold distribution of benefits or burdens from fiscal policy requires fine-grained individual data which is often not available. The reason is neoclassical economic 5. Sex refers to the biological differences that characterize males and females, while gender is the social construction of male and female identity based on biological differences (see page 5). 6. More recent work looking at the differential impact of benefits or transfers created from public expenditure includes Aziz, Gemmell, and Laws (2013) for New Zealand; Doorley and Keane (2020) for Ireland and the United Kingdom; and Garcia-Peña (2019) for Barbados. 7. Grown and Valodia (2010) piloted the proposed methodology in eight countries: South Africa, Ghana, Uganda, Morocco, India, Mexico, Argentina, and the UK. 8. Incidence is measured as cumulative taxes paid as a share of post-tax household expenditure. Other studies estimating the differential inci- dence of direct tax regimes include Doorley and Keane (2020) for Ireland and the UK; Aziz, Gemmell, and Laws (2013) for New Zealand; Lahey (2017) for Australia; Garcia-Peña (2019) for Barbados; Rossignolo (2018) and Rodriguez-Enriquez and Aguila (2017) for Argentina; and Bakker and Philipps (2018) for Canada. Other studies estimating the differential incidence of indirect taxes include Garcia-Peña (2019) for Barbados; Vijil (2017) for Nicaragua; Casale (2012) for South Africa; Fragoso and Gonzales (2010) for Mexico; and Bucheli and Olivieri (2013) for Uruguay. 9. Greenspun (2019) covered five Latin American countries. 4 theory assumes that households maximize the welfare of all household members, and that the household functions collectively and has a single, cohesive objective. Neoclassical economic theory also assumes that individuals maximize their own personal utility, including, for example, deciding whether and how to share any personal income gains or agreeing to share tax burdens. Household-based surveys do not yet systematically capture how the income-based benefits received and the income losses from tax payments are shared among individual members of the household.10 Understanding how the income gains and losses generated by transfers and taxes are allocated among individual household members is necessary to understand how fiscal policies create individual impacts. A class of empirical models has attempted to capture intrahousehold resource allocations using information on exclusive or assignable expenditures; an example of the former is spending on toys exclusively for children and of the latter is expenditure on clothing for boys, men, girls, and women (Lewbel and Pendakur 2008; Dunbar and others 2013; Chiappori and others 2002; Bose-Duker and others 2021; Bargain 2022). In describing how the intrahousehold allocation of resources creates inequality within the household, or varying levels of resource poverty for different members of the same household, these models allow a description of how intrahousehold inequality changes from pre-fiscal to post-fiscal income levels. However, three practical disadvantages of empirical models of intrahousehold allocations make it hard to apply them widely and transparently to gendered fiscal incidence analysis. First, in low-income settings, and for low- income households in low- and middle-income settings, assignable expenditures are infrequent and account for small shares of total household expenditure. Infrequent, marginal expenditures might provide unreliable estimates of how (total) income gains and losses are shared.11 Second, available evidence suggests consumption magnitudes (for individual items) are themselves often imprecisely estimated. In Mexico in 2016, for example, the coverage rate – defined as the cumulative consumption expenditure magnitude for an item covered in a household survey divided by cumulative consumption expenditure (for the same item) from national accounts – of alcohol and tobacco expenditures was less than 10 percent while the coverage rate of education expenditures was more than 160 percent (Zwijnenburg and others 2021). The third disadvantage helps distinguish our approach from empirical approaches to intrahousehold decision-making that make use of exclusive or assignable expenditures. In order to determine how individuals would allocate income gains to maximize their own individual utility, it is usually assumed that preferences for single individuals without dependents or spouses also describe the preference of similar individuals in households with children or elderly dependents (or both); with one, two, three, or more caretakers who may also be income earners; and with members that have both a fiduciary responsibility and intrinsic motivation to care for and 10. Individual-level data helpful for more precise estimates of the incidence of fiscal policies on individuals is time-use data, or records summarizing the differential time-use patterns of males and females, children, and the elderly, and (potentially) household members who are either temporarily or permanently absent. Beyond that, detailed individual-level information on assets, employment, income, and consumption expenditure could be used to provide individual-level impact estimates of fiscal policies. 11. Past expenditures on single items may also be unreliable for predicting future intrahousehold allocation or decision-making processes. 5 provide for others (for example). As Bargain (2022) notes, applying this assumption in developing countries is problematic because “it is not common to find people living alone.”12 The approach taken in this paper instead exploits household demographic composition and economic information to guide estimates of which household activities benefit the most (on net) from fiscal policies and which individual household members fill those roles. It also falls squarely within the “accounting” approach to fiscal incidence analysis which attempts to map fiscal facts (e.g., the taxes collected and the expenditures made) to the behaviors observed in a cross-section of individuals. This analytical choice also limits us to a description of how total economic activity at the household level is supported or hindered by fiscal policies and for whom within the household that such support or hindrance will matter most. These limitations are sufficiently different from those relevant for the intrahousehold bargaining literature, and we do not propose our method as a substitute for those approaches. The methodology we propose is generalizable to most fiscal, socioeconomic, and cultural contexts for the following reasons. First, we include a broader set of activities and points of interaction with the fiscal system, such that individuals who spend most of their productive hours working without remuneration can nonetheless be made more productive by, or face burdens from, fiscal policy. Second, we propose procedures for a gendered fiscal incidence analysis where only the household-level data required for standard CEQ assessments is available and suggest modifications to those procedures when richer individual-level data is available. 3 Gender Inequality, Social Reproduction, and Fiscal Policy Gender is a normative social construct defining and differentiating the roles, rights, entitlements, responsibilities, and social obligations of women and men. Gender is also a set of expectations providing positive and negative incentives for broad-based compliance with those roles, responsibilities, and obligations. Gender relations are embedded in markets and institutions and are integral to how an economy functions. Economic institutions, policies, and the implicit rules and traditions that govern aspects of economic life can exacerbate, ameliorate, or simply modify existing gender inequalities.13 The nature, degree, and social, cultural, and institutional location of these gender differences vary from one society to the next; historically they have favored men (Beneria, Berik, and Floro 2016). 12. Potential solutions to this problem provide estimation strategies to generate individual budget shares from household expenditure data alone without data on single individuals (Dunbar and others 2013). However, these solutions must also restrict preferences across household compositions or across people in order to identify individual allocation shares. Either solution flattens estimated fiscal policy impacts precisely where we believe these impacts would exhibit variability that enhances or discourages gender equity. 13. See chapters in Berik and Kongar 2021, including Berik and Kongar, Beneria and Sen, Seguino, and Braunstein. See also Duflo (2012), and World Bank (2012). 6 Fiscal policy can reduce some gender inequalities directly. For example, conditional cash transfers providing household-level transfers for every child enrolled in school can create positive incentives for girls to remain in school longer and for households to reduce the time spent by school-age girls in care for younger dependents.14 Here too a direct reduction in gender inequality in one area (time spent caring for younger dependents) may lead indirectly to a reduction in gender inequalities in another area (a reduction in the wage gap for externally supplied labor). Fiscal policy can also directly reduce or exacerbate gender differences in paid employment. Around the world, women’s labor force participation rates are lower than men’s (ILO 2023). Women enter and exit the labor force more frequently than do men, which means their participation is more discontinuous than men, and they are more likely to be in part-time and seasonal jobs (ILO 2023). Women on average earn less than men, even after controlling for standard human capital variables (age, education, job experience), though the gap has narrowed in many countries over the last decade (Chamberlain, Zhao, and Stansell 2019). Women tend to dominate in informal employment, especially in low-income settings. Revenue policy can promote female employment by eliminating higher marginal rates on secondary earners in married couples; public expenditures can support unemployed women through unemployment insurance or informally employed women through social protection benefits. Here too, direct impacts today can lead to indirect impacts tomorrow. For example, women’s participation in the paid labor force can help rebalance expectations and social norms regarding women’s and men’s participation in social reproduction activities. An increase in labor force participation is good for the macroeconomy, and any policy encouraging labor force participation can potentially reduce gender employment gaps.15 Fiscal policy can also support or undermine social reproduction. As noted earlier, social reproduction is the production and maintenance of human capabilities. It involves both daily household maintenance using women’s and men’s unpaid labor time and the commodities that men and women purchase. Social reproduction also includes unpaid activities such as fuel and water collection; cooking and cleaning and caring for children, the elderly, and other dependents; and providing cultural, social, athletic, recreational, and educational instruction that encourages the thriving and increased capability of dependents in many walks of life. In the short run, social reproduction activities may positively impact labor productivity when workers are nourished, cared for, and replenished at home (Braunstein and others 2020). Social reproduction may also lead to direct increases in purchasing power. For example, consider the exchange of clothing between a mother with an older child and a mother with a younger child: even if no money changes hands, the time, effort, and social 14. In some countries, conditional cash transfers have also negatively affected maternal employment. For instance, Molyneux and Thomson (2011) noted that in a survey of women who received transfers from the Opportunidades program (Mexico), between one-third and half reported that the demands of the program conflicted with their income-earning activities, which echoed findings from other cash transfer programs in Latin America. 15. However, the gender wage gaps may be unchanged. 7 transactions that are wrapped up in the exchange of clothing have an opportunity cost for individuals who have arranged the exchange.16 Social reproduction includes investments (and associated returns) in human capabilities through education, health care, and the building of social capital. These investments raise future productive capacity while creating a virtuous cycle within social reproduction itself – the more educated, healthy, and capable the younger generations are, and the more positive social connections they benefit from, the generation to follow them will benefit as well. Social reproduction is therefore not only a necessary input into economic production but is also a public good, encouraging positive externalities and limiting negative externalities of current and future economic systems. In short, the work of social reproduction is vital for the paid economy to function and an important criterion for fiscal policy to consider. To the extent that women participate more and men less in unpaid social reproduction activities – which is the case in most countries – fiscal policy supporting social reproduction activities may also be enhancing the productivity of women across all of their activities. For example, public expenditure on roads, transport, and energy services directly supports social reproduction by reducing the amount of time that women spend in unpaid work. Cash transfers and other fiscal expenditures that enhance the bargaining power of women caretakers can affect the amount and type of expenditures, savings, and other investments made by those women. However, in general, empirical studies are inconclusive regarding gender differentials in expenditures and investments. Some studies find that women tend to spend a higher proportion of income under their control on goods that enhance the well-being and capabilities of children, such as food, education, and health care (Thomas 1993; Haddad and others 1997; Quisumbing and Maluccio 2000; Armand and others 2016). However, Benhassine and others (2015) found that, in Morocco, the gender of the unconditional cash transfer recipient has no differential impact; while Akresh, De Walque, and Kazianga (2016) found better outcomes among children (relative to regional macroeco- nomic conditions) when cash transfers were given to fathers rather than to mothers in Burkina Faso. Public benefits directly supporting women may also indirectly support the unpaid social reproduction activities that women are often responsible for. However, indirect taxes (such as a value-added tax or an excise tax) on expenditures that women are typically responsible for may both directly reduce the bargaining power of women and indirectly discourage the social reproduction activities in which women are typically involved. Finally, in many developing countries, women are denied the right to own and inherit a property or other productive assets. Men hold the majority of land and housing titles, and social norms may dictate that business- es are owned by male family members, and women may supply labor to them. Nonetheless, in recent decades, women have made strides in some areas of property ownership, most notably in entrepreneurship and business development. Taxes and transfers can promote or discourage female entrepreneurship and land ownership. 16. Neither the inputs to this transaction, the transaction itself, nor the results of the transaction are visible in a typical household survey. 8 As noted in these examples, fiscal policy directly and indirectly impacts gender inequality and social reproduction. It directly supports social reproduction through provision of goods and services, creation of incentives for behaviors, and setting of prices for goods and services that support social reproduction. However, standard microdata-based fiscal incidence analysis (FIA) distributes cumulative burden and benefit totals uniformly to all household members regardless of the roles that members play in an economic activity (including social reproduction) of the household. In what follows, we propose a series of steps to bring a gendered perspective within the CEQ assessment fiscal incidence analysis framework in a transparent manner and without additional data requirements. 4 Methodology We begin this section with a brief summary of the CEQ assessment method and how the CEQ “income concepts” framework lends itself to a gender-differentiated fiscal incidence analysis. Following this introduction, we discuss in which dimensions of gender inequality we might expect to observe or deduce evidence of fiscal policy impact. We then propose a methodology for examining the impact of fiscal policy on gender by expanding the CEQ assessment method using standard datasets and techniques. CEQ Assessments A CEQ assessment accounts for and quantifies the nature, character, and magnitude of benefits received as well as the burdens experienced by individuals and households from the execution of a fiscal policy. A CEQ assessment generates counterfactual income distributions and computes poverty and inequality indices based on the distribution of each of these income concepts. These pre-fiscal and post-fiscal incomes contain more or fewer of the benefits created from public expenditure or tax contributions to public revenues. The analysis starts with pre-fiscal income (referred to as “market income” or “market income + pensions”) which is earned or unearned from wages and salaries or profits from enterprises, self-employment activities, or as returns on capital and other private income sources (Figure 1). Net market income is what would be left behind after all relevant income taxes are paid and households and the necessary deductions and withholdings have been taken out. If the household is a recipient of any direct cash transfer from the government, this would be added to net market income to arrive at disposable income. At this stage, depending on the consumption basket of the household, various indirect taxes (VAT and excise) as well as indirect subsidies (energy subsidies) would be paid or accrued. Adding or subtracting these as appropriate would yield consumable income or post-fiscal income. Finally, accounting for any in-kind benefits received from the government on things like education and health and discounting associated copayments and user fees would yield final income. 9 Figure 1. Overview of the Commitment to Equity Methodology TAXES TRANSFERS MARKET INCOME (SUBSTRACTS (ADDS TO INCOME) Wages and salaries, income from FROM INCOME) capital, private transfers; before WHO government taxes, social security RECEIVES? contributions and transfers. INCOME TAXES & PENSION CONTRIBUTIONS NET MARKET INCOME DIRECT CASH WHO PAYS? TRANSFERS HOW MUCH? HOW MUCH? DISPOSAL INCOME INDIRECT INDIRECT SUBSIDIES TAXES CONSUMABLE INCOME IN-KIND CO-PAYMENTS TRANSFERS AND USER FEES (HEALTH, EDUCATION) FINAL INCOME Source: Lustig (2018). How do gender norms and roles play out in this framework? Broadly three types of activities create exposure to fiscal policies: as individuals earn income, as individuals become consumers, and as individuals provide unpaid labor to the household or become investors in household care (Figure 2). 10 Figure 2. Economic and Social Activities and Fiscal Policy HOW/WHERE DO INDIVIDUALS EXPERIENCE FISCAL POLICIES? WHICH POLICIES? As suppliers of external labor As consumers Income taxes including tax credits and Direct transfers - can target recipient/ taxable income rules eligibility; can target control of benefit MSME finance Indirect taxes and subsidies Maternal/parental leave policies that address wage gaps As participants in social reproduction In-kind transfers: education; health; As investors in human capital water; sanitation, and hygiene (WASH) services; childcare elderly care Labor-market activation programs Infrastructure Provision of housing and other assets Communications/internet and media First, as income earners, individuals supply labor (externally, or outside the household) both formally and informally in market-based arrangements. Most often, fiscal policies surrounding the provision of labor or wages will have an impact on pre-fiscal income (also known as market income) or net market income, which is pre-fiscal income minus any direct taxes owed on wages or other earned or unearned income. For example, the personal income tax statutes and rate schedules may privilege the “primary” income earner and penalize the “secondary” income earner; secondary income earners are more often women than men. Policies subsidizing paid parental leave may ameliorate or exacerbate inequalities with respect to at-home care for young dependents. Subsidized micro, small, and medium-sized enterprise (MSME) finance – which in principle should lead to high- er market or net market incomes – may not benefit women and men equally when women are unable to be deed holders of productive assets. Individuals must also demand goods and services and make decisions regarding how much to consume or save; these activities also create exposure to fiscal policies. For example, direct transfer programs like conditional or unconditional cash transfer programs effectively supplement budgets for essential goods and services like food, clothing, shelter, and energy; direct transfers have an impact on disposable income. Indirect taxes on consumption activities – VAT, excise taxes, customs duties, and sales taxes – effectively reduce consumption of goods and services (for a given budget) while indirect subsidies on consumption activities will effectively increase consumption (for a given budget). 11 Finally, the third type of exposure to fiscal policy occurs when individuals within a household provide unpaid labor and invest in the care of all household members and the communities in which they exist. These activities belong to social reproduction as defined above.17 The public provision of education, health care, WASH, and housing services directly provide social reproduction components at low cost. Providing transport and communication infrastructure enables faster access to essential social services for those involved in social reproduction (and therefore lower opportunity cost). Which Gender-Based Impacts Are Observed or Deducible? The impact of an individual fiscal policy, or fiscal policy in general, on gender equality will be multifaceted. To quantify fiscal policy impact, we would like to know (among other things) whether the policy (1) supports social reproduction, (2) creates incentives for more or higher-quality female labor force participation, (3) creates incentives for more or higher-quality male participation in social reproduction, (4) increases the individual purchasing power of women, (5) increases individual agency18 for women, and (6) increases the per-capita cumulative purchasing power of the household. The approach developed in this paper does not permit estimation of the impact of fiscal policy on gender inequality or women’s agency along all these dimensions. This is in part due to data limitations. To understand whether and how agency is affected by fiscal policy, for example, would require a wide perspective on the tradeoffs that households face, how options are weighted and judged by all household members, how compromises are made, and which characteristics inform bargaining positions.19 Understanding the potential for agency also calls for an investigation into prevailing social norms, legal rules, and cultural expectations that may encourage (or limit) agency outside the household.20 Irrespective of data limitations, our inability to categorize a fiscal policy as “good” or “bad” for gender equality is also partly due to conflicting impacts along these different dimensions. For example, giving benefits to mothers who stay at home to look after their children reduces gender inequality in total incomes while exacerbating gender inequality in labor force participation. In the same vein, generous maternity leave has been shown to discourage women returning to the labor market while voucher systems for childcare have been shown to discourage the use of market-provided childcare services in some contexts (Spiess and Wrohlich 2008). 17. These activities should be distinguished from other types of unpaid labor, for example, in subsistence agriculture or labor supplied by household members to a family business. 18. Agency is the ability to define goals and act on them; see Donald and others (2020). 19. See Bakhtiar and others (2022) for an experimental examination of the processes determining intrahousehold allocations and the impact of those processes on women’s agency. 20. When such data is available, or when FIA can be complemented with a discussion of social norms and institutions that limit policy-based progress on gender inequality, a gendered FIA will be much richer and more nuanced. Such information is available from, for example, the Social Institutions and Gender Index; Afrobarometer; the World Values Survey; and Women, Business and the Law. These sources, among others, can be used to complement FIA and triangulate policy-relevant advice. 12 Consider also introducing work requirements into welfare programs. If the labor force participation requirement leads to less or lower quality time of single female parent households in social reproduction activities, that could lead to worse outcomes among dependents (Kalil and others 2022). Higher labor force participation may lead to total lifetime income reductions for the same women (assuming that their incomes later in life are at least partially supported by intrafamily transfers from working-age members to elderly dependents). On the other side of the same coin, conditional cash transfer programs may require additional time spent in nonremunerated social reproduction activities by caregivers – who are most often women – even while generating better human capital outcomes among dependents. Lastly, the “same” policy from a fiscal standpoint may have different gender inequality impacts depending on the mode of delivery. Bjorvatn and others (2022) demonstrate in an experimental setting in Uganda that a childcare subsidy delivered as an in-kind service was at least as effective as an equivalent cash grant in increasing a mother’s total income and labor productivity; that the childcare subsidy also led to increased labor income for fathers while the equivalent cash grant had no impact; and that the childcare subsidy led to higher enrollments while the equivalent cash grant did not. The two policies are equivalent from a fiscal expenditure point of view, and yet the positive impact of the childcare subsidy on women, children, and men is greater than that of the equivalent cash grant. We can begin to understand how fiscal policy affects gender inequality, social reproduction, and care by analyzing household demographics and social organization. The sex and age composition of households plays a significant role in determining care needs as well as which taxes and transfers supporting or undermining social reproduction are relevant: children enrolled in public school receive in-kind education services; women of child- bearing age have greater health care needs; households with older, non-working older male members may qualify for old-age support. The employment status of household members plays a role too. For instance, the employment status of a married couple can influence effective marginal tax rates on “secondary” income. Who takes part – and how frequently – in social reproduction can influence the impact that education service provision has: preschool children require more care while school age children require less, but school-age girls may be called on to provide some or all the care of younger dependents. And if women (rather than men) are more responsible for social reproduction, they may not look for paid employment. Household Typologies: Bringing Care and Social Reproduction into the Picture The size and organization of the household are observable in household budget surveys, income and expenditure surveys, living standards measurement surveys, and other survey instruments used in standard CEQ assessments. We propose, therefore, that the CEQ assessment toolkit be adapted to estimate fiscal policy impacts on households according to their demographic (sex and age) and employment-disaggregated characteristics. 13 We present below a list of generic household types that can be tailored to specific country contexts. At a minimum, these types should specify the sex and age ranges of household members, the relationships among the household members (for example, married or in consensual union, with and without dependents), and the employment or earning status of the principal adults in the household. Together, that information is important in establishing what the care and social reproduction needs of various households may be. Constructing the household typology and undertaking the fiscal incidence analysis should follow a few steps. The first step is to compile information from censuses, household surveys, labor force surveys, and other sources, such as time use if available,21 to get a big-picture understanding of the nature of gender inequalities in employment and social reproduction. The second step is to use this data to construct economic and demographic categories. A generic set of groupings that can be adopted in most countries is as follows: Economic categories 1. Single male/female member households, who are employed or not employed. In many countries, these are older retired adults. 2. Households with married couples or consensual unions.22 This category should be divided into at least four subcategories: where both principal adults are employed, where only the principal male is employed, where only the principal female is employed, and where no adult is employed. 3. Households with working-age adults and dependents, disaggregated by which members are employed. Examples of this type of household may be an adult child living with an elderly parent or younger-generation children living with older-generation children (from the same parents). Demographic categories Using the economic categories above, households should then be further classified by their care needs:23 1. without dependents and with dependents 2. total dependents 3. dependents disaggregated by age: for example, children 0 to 3 years old, children ages 4 to 12, adults over 75, and/or other age compositions relevant in a particular country context. 21. Additional contextual information is available in sociological, legal, and anthropological studies and in databases summarizing the institutional and legal structures that inform gender-based equity. See for example the Social Institutions and Gender Index, Afrobarometer, the World Values Survey, and the Women, Business and the Law databases. 22. In those countries where same sex marriages or consensual unions are recognized and such data exist, the types could also be male/male and female/female. Consensual unions also include polygamous structures. 23. In the first stage of work, country pilots classified households into separate economic and care categories. However, for fiscal incidence analysis, it is important to use a single combined classification. 14 Analyzing these categories is important for understanding country-level patterns. For the fiscal incidence analysis, a single household classification system should be created by combining the economic and care categories.24 That is because tax and transfer policies condition simultaneously all of the economic and care activities that households and household members pursue. For example, personal income tax policy will determine how many household members will spend time in paid employment (as well as how long each member will spend in paid employment), how many dependents from each age bracket can be supported from wage earnings, and how much time each household member will spend on social reproduction activities. Table 1 illustrates the advantages and disadvantages of each of these categories from a fiscal incidence perspective in the context of Armenia.25 Beginning with Type 2 (T2) households, the needs and opportunities of individual household members shape household organization and the incentives informing choices and behaviors at home, in the labor market, and in markets for goods and services. T2 categorizes households according to the contributions (to household income) of individual male and female members, while T3 assumes that not all household earners have equal contributions or bargaining power within households. T4 implicitly identifies the presence of children and the elderly to capture the household’s care needs and domestic responsibilities, and their interaction with incomes and labor decisions. While initially the demographic and employment categories can be studied separately, the fiscal incidence analysis would ideally be performed on the integrated employment/care category (see T4 in the last column of Table 1). Table 1. Household Typology, Armenia (T1) Demographic (T2) Income (T3) Gender-shares composition contributions of HH earnings (T4) Mixed typology Types Female majority Earners are a Female earnings >60% Earners are female or male majority female majority, of total earnings, majority + dependent (or equal numbers) Earners are a male male earnings >60% children + dependent of household majority, equal of total earnings, elderly, Earners are male members number of male No earners majority + dependent and female earners children + dependent elderly Usefulness/ Exploits variation in Introduces Exploits variation in Captures care needs and relevance consumption patterns women’s economic bargaining power domestic responsibilities, for gender and gender-specific opportunities and and interaction with analysis demands or needs household roles incomes and labor decisions Source: Based on Fuchs and Gonzalez (2022). 24. In some countries, like Jordan, the cell sizes for a combined economic/demographic classification may be too small for a fiscal incidence analysis. In this case, the analyst will need to make choices about which classifications make most sense in any particular country context (see Rodriguez, Wai-Poi, and Woodham 2022). 25. A household type defined by household headship is included in Fuchs and Gonzalez (2022). However, Grown and Komatsu (2010) discuss the lack of relevant information in standard household survey techniques that ask household members to identify the head of household. 15 After the household types have been constructed, the next step is to nest them within pre-fiscal and post- fiscal income distributions so that fiscal impacts along gender, social reproduction, and income-disaggregated lines can be estimated and analyzed (see Table 2).26 Household types can also be disaggregated by other key characteristics including poverty status according to national or international poverty lines, rank in the income distribution, rank in the urban or rural or provincial-level income distributions, and rank in the conditional (upon household type) income distribution. (See Armenia examples in the next section and in the Appendix.) Table 2. Examples of Disaggregated Household Types, by Income Rank Quintiles of prefiscal or postfiscal Single member Male Male not Female Female not income disribution households employed employed employed employed Q1 Q2 Q3 Q4 Q5 Female Not employed Dual Male employed with children Married or employed employed with children and Consensual with children with children (by age range other adults Unions (by age range) (by age range) as feasible) (by age range) Q1 Q2 Q3 Q4 Q5 After the household types are located in the pre-fiscal and post-fiscal income distributions, the next step is to generate a fiscal incidence analysis (for example, using the CEQ assessment framework) to summarize which of the household types most often benefit from policies that support social reproduction, which households experience the biggest burden of indirect taxes on different commodities, or which households most often confront 26. In countries where income data are not available or less reliable, consumption expenditure can be used to construct a welfare distribution. 16 disadvantageous tax rates on secondary incomes. Finally, the completed FIA is then used to analyze whether, as executed, the policies support an equitable distribution between different household members via a mapping of household roles and functions across individual households.27 When additional microdata beyond the standard household budget or income or expenditure survey is available, it can be incorporated to more precisely identify individuals who fulfill the roles or most often carry out the activities supported by public expenditures or made costlier via the application of taxes. Time-use data, for example, which records how individuals spend their waking hours, are used to capture who most often carries out care activities or other aspects of social reproduction. Then, each household type’s share of taxes and transfers can be further allocated to roles or activities and the individuals that most often perform those roles or carry out those activities. 5 Applying the Method to Jordan and Armenia This section discusses the application of this methodology in Armenia and Jordan. Both countries have recently completed CEQ assessments and are using the methodology described here to dig deeper into potential impacts of fiscal policies on gender inequality. The fiscal and social structures of the two countries provide useful contrasts and illuminate how country specificity may be important in the identification and operationalization of household categories relevant for gender-specific incidence analysis.28 5.1 Armenia Armenian women have low labor force participation rates. Less than 60 percent of working-age women participate in the labor force, compared with just over 70 percent of men (World Bank 2023a). Women are also more likely to work part-time positions than men. Women are underrepresented in science, technology, engineering, and math fields, accounting for only 38 percent of graduates in these fields (World Bank 2023b). In 2019, one-quarter of women workers were employed in agriculture, while women were virtually excluded from industrial activities, such as construction and mining. Men consistently earn higher wages than women, regardless of age and educational attainment; even tertiary-educated women employees earn only 72 percent of the average wage earnings of tertiary-educated men. That evidence suggests that women are segregated both across and within sectors of employment, facing exclusion from higher-paying sectors and higher-paying positions. 27. This likely requires a consultation with in-country experts or primary or secondary data sources to determine who most often fills the roles that are either supported by or made costlier by fiscal policy. 28. Armenia is a transition economy with an aging population wherein much of the redistributive impact of fiscal policy is due to pensions. Jordan, on the other hand, is a country with strong social norms against female participation in the labor market. 17 Preliminary demographic analysis reveals that over a quarter of Armenian households are composed of a male and female couple with children, with other adults also living in the household (Fuchs and Gonzalez 2022). Most single-headed households – where the self-identified household head is not living with a spouse – are sustained by females; single female-headed households account for almost one-fifth of all households. a. Definition of Household Types Three typologies were tested for the fiscal incidence analysis (see Table 3). Ultimately the T3 category was used most often while T2 was used in some analyses due to sample size limitations. Table 3. Household Types Based on Armenia ILCS 2020 Demographics + Income (proxy for intrahousehold roles and bargaining power) (T1) A. (Labor) earners are a female majority • Calculated from the gender of Gender of B. (Labor) earners are a male majority household members who report household earners C. Equal number of male and female positive labor earnings in labor markets labor earners • Labor earnings include salaries and D. No (labor) earners in the household self-employed earnings; exclude incomes from agricultural sales and own consumption (T2) A. Larger share of female earnings • Calculated over total labor earnings Gender sustaining (>60% of household labor earnings) reported by gender of household household earnings B. Larger share of male earnings members (>60% of household earnings) • Labor earnings include salaries and C. No earners in the household self-employed earnings; exclude incomes from agricultural sales and own consumption Demographics + Income (proxy for intrahousehold roles and bargaining power) + Presence of children and elderly (proxy for care needs and domestic responsibilities) (T3) • Typology (3) of gender sustaining • Calculated from gender and age of all Gender sustaining household labor earnings household members (self-reported) household earnings • Combined with the presence of in household interview roster; labor + presence of retirees who may contribute to earnings reported by gender of children and domestic tasks and care household members elderly • Combined with the presence of dependent household members: - Children <6 years old - Elderly 80+ years old Note: ILCS = Integrated Living Conditions Survey. 18 Within the T3 hybrid set of economic and care subcategories, the following disaggregations were created and located within the post-fiscal income and geographic distributions (Table 4). Two of the most common subcategories in the general population are households where male earnings account for more than 60 percent of all earnings, and households in these two subcategories are more often urban and in the top three quintiles of the income distribution (rather than in the bottom two quintiles). The most common subcategory of households where females account for greater than 60 percent of all earnings are those households with no young- or old-age dependents; this subcategory is also far more likely to be urban-based and far less likely to belong to the bottom two quintiles of the income distribution. Table 4. Household Types Based on Armenia ILCS 2020 Demographics + Income (proxy for intrahousehold roles and bargaining power) + Presence of children and elderly (proxy for care needs and domestic responsibilities) Population Bottom 40 Top 60 Rural Subcategory T4 Share (%) (%) (%) (%) Dependents Female (>60% earnings) 1.2 37 63 46 + Elderly Male (> 60% earnings) 3.4 51 49 44 No labor earnings 1.5 61 39 54 Dependents, Female (>60% earnings) 3.9 46 54 43 No elderly Male (> 60% earnings) 19.7 45 55 46 No labor earnings 5.6 59 55 46 No dependents, Female (>60% earnings) 4.4 35 65 29 Elderly Male (> 60% earnings) 7.4 43 57 36 No labor earnings 7.3 39 61 34 No dependents, Female (>60% earnings) 10.5 31 69 30 No elderly Male (> 60% earnings) 28.1 31 69 38 No labor earnings 6.8 47 53 61 Note: ILCS = Integrated Living Conditions Survey. b. Findings: Impact of Taxes and Transfers across Household Types In Armenia, direct taxes and the public contributory pension system have a large impact on improving the relative income position of either female-headed households or households where female earnings account for at least 60 percent of cumulative household income. At the same time, dual earner households contribute the highest pre-fiscal income share to income taxes while benefiting the least from direct transfers, which is an implicit disincentive for households to allocate all their labor supply to formal labor markets. Taken together, and conditional on cultural practices or community- or household-level arrangements that prioritize men as “primary” income 19 earners, these facts may indicate that fiscal policy in Armenia seems to be associated with women (1) staying out of the labor force, (2) taking lower-paying employment, and/or (3) working in economic activities that tend to be in the informal sector.29 The fiscal system significantly reduces gender gaps in incomes for all household types, but the gaps for households with dependents remain negative and the largest, indicating that dependents create demands on care- givers more so than they contribute to the supply of caregiving. On the other hand, the presence of stay-at-home elders reduces the gender gap relative to a household type with dependents but no assistance from elders. The reduction is most significant from pre-fiscal to disposable incomes, highlighting the association with household labor market decisions and the associated PIT liabilities.30 For reference, the relatively small gender gap at pre-fiscal income in households with neither older-age nor younger-age dependents is actually reversed and becomes a gender “bonus” at the disposable and consumable income concepts. These patterns are consistent with older-age dependents contributing resources to care, social reproduction, and household production simultaneous with creating expenses associated with their care needs; they are also consistent with those expenses being assigned more frequently to women. c. Incorporating Time-Use Data Worldwide, women spend over three times as much of their time on unpaid work, and approximately one-third as much of their time on paid work, as men (Charmes 2019). The volume of unpaid work may mean women have less time than men to engage in paid work; or women may feel restricted to jobs allowing for a balance between paid and unpaid work. Either way, a one-hour decrease in time spent on unpaid care work is correlated with a five- percent increase in women’s labor force participation (Ferrant and others 2014). Women in Armenia are much more likely than men to miss economic opportunities due to childcare, older dependent care and support, and other social reproduction activities. Women experience “time poverty” linked to responsibilities for unpaid domestic tasks and non-utilization of childcare services. A contributing fac- tor to women’s burden is the low enrolment of children in preschool institutions, at only 30 percent nationally and 17 percent in rural areas in Armenia (SCRA 2021). Although there are government-provided childcare ser- vices, childcare payments are not tax deductible and private childcare centers do not receive nontax benefits (USAID 2019). 29. For example, informal employment, household production for self-consumption, and other forms of work outside labor markets such as domestic work at one’s own household, and childcare. Domestic tasks, including childcare and care for other household members are generally classified as out-of-the labor-force. This paper situates all social reproduction activities squarely within the set of economic activities that a household pursues. 30. Contributory pension incomes (and other benefits) also have a significant gender-equalizing effect, particularly among households with elders 65 to 79 years old. 20 Time-use data confirms women are overburdened by household and family responsibilities, including care for children and the elderly.31 Time spent on childcare and elderly care (as a share of total hours of activity in a regular weekday) is three times as much for women than men (Table 5). The pattern is independent of geographic location and much more significant in households with children, especially young children.32 Table 5. Time Use Survey Responses, Armenia QUESTION: HOW MANY HOURS DID YOU SPEND IN A TYPICAL WEEKDAY/WEEKEND ON THE FOLLOWING ACTIVITIES? Work and Elderly commuting Domestic Sleeping care Childcare to work Studying Leisure tasks Transport Others Location Female 31.3 1.9 9.8 14.4 4.8 17.7 12.6 4.9 0.4 respondent Armenia Male 32.7 1.3 3.5 26.8 3.1 19.8 3.2 4.7 0.4 respondent Armenia Women with tertiary education spend more time at work than women with no tertiary education while men spend close to 27 percent of their time in employment. Adding up time spent on elderly care, childcare, and domestic tasks – all of which contain elements of social reproduction – reveals that women (nationally) spend three times as much on those three tasks as men. Even the group of women spending the most time working – those with tertiary education or living in the capital Yerevan – spend three to four times as much on these activities as their men counterparts. The groups of women who spend the least time working – those in households with children, with or without elderly dependents – also spend three to four times as much on these activities as their men counterparts. 31. The survey for “Assessing Gender Disparities in Time Allocation and Household Responsibilities in Armenia” was conducted in 2022, by the World Bank, in partnership with CRRC Armenia. It aimed to complement results from the pilot study for the Engendered CEQ analysis and address current knowledge gaps on the decisions and perceptions of adults in Armenia related to (1) time use, (2) gender roles and responsibilities within households, (3) childcare access, (4) access to elderly care, and (5) opinions and preferences on gender disparities. 32. Men also consistently spend more time on leisure activities than women; the gap in leisure hours is also larger in households with young children. 21 Interviewed women and men ages 65 and above allocate 5 percent of their time to childcare, indicating that elderly dependents in Armenia can contribute to childcare within the household (or family). Women living in households with young children reduce their time allocated to childcare slightly with the presence of elderly members (from 21 to 18 percent). Empirical results indicate that gender roles and disproportionate time allocations to social reproduction activities constrain women’s time and labor market opportunities, which conditions women’s interactions with the fiscal system. The gender gap (in labor market participation) is consistently higher for households with dependents, who most likely need care instead of helping out by providing it. The fiscal system significantly reduces gender gaps for all household types, but the gaps for households with dependents remain negative and the largest. Aid from stay-at-home (healthy) elders can reduce gender gaps in labor earnings for households with primary school children (or older) while children and elders can aid with household chores and care for family members in need. Access to affordable and convenient care for children and elderly family members can therefore influence their labor decisions and interactions with the fiscal system, for women in particular, and potentially for every household member. Results from the Engendered CEQ assessment with time-use data in the Armenia example demonstrates the intricacies in which a model of intrahousehold bargaining would have to consider for it to be useful within a fiscal incidence model. It appears that older dependents can contribute in a positive way to household incomes or social reproduction and care responsibilities of females within the household (or both). The type of contribution made by an older dependent, however, likely depends on the dependent’s gender and health status. Furthermore, in Armenia, the positive nature of the contribution has been neutralized at post-fiscal income stages, which indicates (among other things) that the incorporation of older-age dependents into the household is likely negotiated by the household’s adults. Both the starting positions of the negotiating members and the outcomes are likely affected too by the nature of the contribution as well as the gender of the older-age dependent. A single policy, therefore, may create impacts at different decision or negotiation points, and intrahousehold bargaining models that completely describe interactions and the incentives that constrain those interactions at each of those points will lack generality. 5.2 Jordan To create household types in the Jordan Engendered CEQ assessment, Rodriguez, Wai-Poi, and Woodham (2022) explored whether households could be grouped into commonly occurring categories that exhibit demographic or organizational differences salient in the allocation of care and social reproduction responsibilities on the one hand and the incidence of taxes and spending on the other.33 A relatively small number of categories are needed for tractable analysis. Each category should neither represent too many people so as not to obscure differences between them nor too few people so as to lack representativeness of the entire population. The statistical tradeoff between large groups with precisely estimated average characteristics and smaller groups that would likely show greater deviation from a national average but exhibit less precise estimates should also be taken into account. 22 In Jordan, Rodriguez, Wai-Poi, and Woodham (2022) proposed five “care” subcategories to represent the important differences in the amount of home-based care and social reproduction necessary as well as in the share of these activities accomplished by different household members (Figure 3). Three of the five categories represent a quarter to a third of each of the following Jordanian population groups: married with dependents, married with dependents and other adults, and no dependents.34 The single adult with dependents category is smaller (at 5 per- cent); it was included in the analysis as the implications for care responsibilities are significant. A single adult must bear all of the burden, perhaps with some help from older children or elderly members. Figure 3. Care Categories in Jordan Married couple 2+ unmarried and 1+ other adults: adults with 1+ children (at dependents Married, Other least one is 0-5), no seniors (seniors or other adult seniors and 1+ children (all 6-14), no seniors 31 children) 1+ seniors, no 11 children 1+ seniors, 1+ children 23 man Married couple 31 woman with: 5 dependents 1+ children (at Married, Married adults, other least one is 0-5), no seniors 2+ unmarried adults, 1+ children (all no dependents 6-14), no seniors 1+ seniors, no Single man or woman with: children 1+ children (at least one is 0-5), no seniors 1+ seniors, 1+ 1+ children (all 6-14), no seniors children 1+ seniors, no children 1+ seniors, 1+ children Source: Rodriguez, Wai-Poi, and Woodham (2022). 33. The others category makes up 11 percent of the population. 23 The no dependents group tends to be richer than average and when married or cohabitating (more frequent than single adults with no dependents), the primary woman is less likely (on average) to be working. Single adults with dependents are also richer than average, and single women with a dependent are far more likely to work. Married couples with dependents have younger children while married couples with other adults have older children; both groups are poorer than average. The “fiscal” subcategories (Figure 4) consider the same household demographic dimensions, but each dimension is considered from a different perspective: • Children matter: the amount of education benefits received will depend upon how many children are in the household and whether they are enrolled in a public school (without concern for the child’s age). • Elderly matters: the amount of pension benefits will depend upon how many elderly members there are in the household and whether they are receiving a public pension. • The working status of married couples matters: the relative value of dependent-based income tax deduction will depend upon whether both, one, or neither of a married couple are working. Figure 4. Fiscal Categories in Jordan No workers,1 children in school 14 35 No workers,1 23 Children in school4 11 Pensioners3 no dependents2 28 1 no workers: 3 pensioners include: 4 children in school Single head of No children No worker includes: household is not in school One worker working workers • +1 children public pension in school, couple is working 1+ pensioner Source: Rodriguez, Wai-Poi, and Woodham (2022). 24 Fiscal and care types are based on the same set of demographic characteristics (number, sex, and age of members, as well as employment and schooling status), thus in principle, overlapping membership is possible, and ideally the categories could have been consolidated for fiscal incidence analysis. However, in Jordan, there is not enough overlap to warrant an integrated categorization for analytical purposes. For example, half of the households in the married couple with dependents care category are also classified as 1–2 workers with children in school (a fiscal category), but at the same time, a third, while also having one or two workers, enters in the no dependents category (because their children are preschool age or not in school, or because they have no children).35 Impact of taxes and transfers for households in fiscal and care groups In Jordan, the only fiscal types that are net beneficiaries of the fiscal system – that is, the value of the benefits received is greater than the payments in taxes – are those that have children in school (Figure 5). This reflects the fact that a large part of the benefits, and the progressivity, of the fiscal system in Jordan comes in the form of an in-kind education (Rodriguez, Wai-Poi, and Woodham 2022). Within the care subcategories, net benefits received from the fiscal system are significant only for the married couple with dependents category (and close to zero for all other care categories). This group receives large education benefits, and large amounts in indirect subsidies (Figure 5). When excluding in-kind health and education benefits, none of the care types are substantial net beneficiaries of the fiscal system (Figure 5). Even the married couple with dependents household category is made neither better off nor worse off (on average and in cash terms) when in-kind benefits are left out. Though this category receives the largest amount of direct social protection transfers36 and indirect subsidies for electricity and water, of all fiscal and care groups, they are at the same time the largest payers of direct (personal income tax) and indirect taxes (goods and services tax and streamlined sales tax) in absolute terms. Taken together, these findings indicate that the fiscal system in Jordan neither supports nor hinders social repro- duction activities. Households with dependents access essential basic services in health and education more often, but receipt of an in-kind service does not improve the material well-being or the net cash position of a household in the current period.37 As the vast majority of Jordanian households have social reproduction responsibilities, this is a missed opportunity for supporting an average Jordanian family. 34. Jordan’s demographic structure may be representative of countries with very low female labor force participation. 35. However, much of it comes from the now discontinued bread transfer. 36. The opportunity cost of the acquisition of those services (days in paid employment lost) may in fact contribute negatively to the household’s current net cash position while simultaneously raising their expected future net cash position. 25 Figure 5. Payments of Taxes and Benefits of Public Spending by Household, by Fiscal and Care Categories (Percentage of Market Income) 100 (percentage of market income) Fiscal revenue/expenditure 80 60 40 20 0 -20 Direct taxes Indirect taxes - direct Indirect taxes - indirect Direct transfers Indirect transfers - direct Indirect transfers - indirect In-kind spending (education) In-kind spending (health) Total impact Source: Rodriguez, Wai-Poi, and Woodham (2022). For households in the single adult, dependents care category, fiscal policy through consumable income reduces the poverty headcount by over 50 percent. For most other care and fiscal groups, poverty falls only slightly from market income to disposable income and then plateaus when moving to consumable income, accounting for indirect taxes and subsidies (see Figure 6).38 37. Official poverty is measured using the household per capita consumption aggregate and results in a poverty rate for Jordanians of 15.7 percent. As a standard in the CEQ approach, we do not report the poverty impact of in-kind transfers, since these benefits are neither cash (as all the taxes and other spending benefits are), nor are considered when constructing the poverty line. 26 Figure 6. Poverty Impact Change between Income Concepts for Fiscal and Care Groups 30 20 Percentage 10 0 Market Disposable Consumable Final income income income income Poverty (national poverty line) No workers, no dependents No workers, children in school Pensioners One/two worker, no dependents One/two worker, children in school No dependets Single adult, dependents Married couple, dependents Other care categories All Source: Rodriguez, Wai-Poi, and Woodham (2022). In Jordan, income inequalities across the care and fiscal categories are driven largely by inequality within any one category; this is true whether pre-fiscal or post-fiscal income concepts are used to estimate inequality (Figure 7 and Figure 8). The fiscal group with the highest inequality within its members is the workers, no dependents group while the no workers and children in school group demonstrates the least (Figure 9). For care categories, the married couple with dependents group shows the largest within-group inequality while the single adult with dependents category demonstrates the least inequality between its members (Figure 10). 27 Figure 7. Inequality Decomposition by Fiscal Categories 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Market Disposable Consumable Final income income income income Within Between Total Source: Rodriguez, Wai-Poi, and Woodham (2022). Figure 8. Inequality Decomposition by Care Categories 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Market Disposable Consumable Final income income income income Within Between Total Source: Rodriguez, Wai-Poi, and Woodham (2022). 28 Across fiscal categories, fiscal policy produces the largest reduction in inequality for the workers, school- age children group, from market to final income (Figure 9) while each group keeps its market income ranking vis- à-vis within-group inequality. That is, the workers, no dependents group shows the most within-group inequality, and the no workers and children in school shows the least within-group inequality at either pre-fiscal or post-fiscal income. Across care categories, the married couple with dependents group also shows the largest fall in its with- in-group inequality when moving from market to final income (Figure 10). For the fiscal categories, however, within-group inequality rankings do not change from pre-fiscal to post-fiscal income. Figure 9. Contribution to Within-Group Inequality by Fiscal Categories 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 Market Disposable Consumable Final income income income income No workers, no dependents No workers, children in school Pensioners One/two worker, no dependents One/two worker, children in school Source: Rodriguez, Wai-Poi, and Woodham (2022). 29 Figure 10. Contribution to Within-Group Inequality by Care Categories 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 Market Disposable Consumable Final income income income income No dependents Single adults, dependents Other care categories Married couple, dependents Source: Rodriguez, Wai-Poi, and Woodham (2022). All told, in Jordan, the receipt of in-kind benefits, primarily education, determines which fiscal or care groups receive positive net benefits from the fiscal system. When focusing only on cash benefits, the fiscal system is very close to neutral for groups. The Jordanian fiscal system does reduce within-group inequalities but has only a very small impact on inequalities between groups which are very low at the outset. Most of the within-group inequality reduction happens when direct taxes and transfers are added to the fiscal system.39 c. Incorporating Time-Use Data Jordan has one of the lowest rates of female labor force participation in the world at around 15 percent, and during the COVID-19 pandemic, unemployment rates among women rose dramatically. In Jordan, as elsewhere, women bear more responsibility for childcare during school closures as well as care for other family members who have fallen sick. Women ages 15 to 44 spend over thirteen times as much time on chores and nearly five times as much time on childcare as men of the same age group (World Bank 2020). 30 38. These findings reinforce the general (non-gendered) findings presented in Rodriguez, Wai-Poi, and Woodham (2022) where it was shown that the direct fiscal instruments (direct transfers and direct taxes) were the most cost-effective in reducing overall vertical inequality. 31 Time-use data from Jordan reveals that time spent in social reproduction activities has not decreased for working women, creating a powerful disincentive to labor force participation (World Bank 2020; Alhawarin and others 2020). Women spend on average four hours a day on household chores and another two hours on childcare; men spend around one hour combined on both (Figure 11). While women in paid employment work fewer unpaid hours around the house, they still do more unpaid work than working men and work more total hours than working men (Figure 12). This means that in addition to spending more time on social reproduction generally, women also spend less time (relative to men) on leisure and self-care activities. Figure 11. Hours per Day Spent on Household Chores and Childcare, by Gender 14 12 10 8 Hours 6 4 2 0 Men Women Men Women Pre-COVID COVID Indirect care Direct care Source: Rodriguez, Wai-Poi, and Woodham (2022). Figure 12. Daily Paid and Unpaid Hours of Work, by Gender 14 12 10 8 Hours 6 4 2 0 Women Women Men Women Inactive Paid employment Unpaid care Paid Source: Rodriguez, Wai-Poi, and Woodham (2022). 32 Married women in Jordan, even with no children, are far less likely to work for pay, indicating that social norms, including those related to fertility expectations, play a role in the household resource allocation process. Women with less housework and fewer care responsibilities are not more likely to be in paid employment than those who do more hours of work at home; however, the presence of other women in the household with paid employment increases the chance of women working themselves. Conversely, having other women in the household who do not work for pay (and therefore could potentially help out with care responsibilities) does not increase the likelihood of women working. These patterns indicate it is likely that social norms and expectations regarding roles that constrain women from accessing labor market opportunities rather than time allocations to unpaid social reproduction activities. For Jordanian women, being married generates the most extra unpaid work.40 Young children increase unpaid work modestly; school-aged children do not add to the burden. The presence of another adult woman in the house reduces women’s unpaid work, but another adult man increases women’s unpaid work. Unsurprisingly, there is high demand in Jordan among working women and women seeking employment for childcare; the hourly rate these women indicate they are willing to pay is similar to that being currently charged in the market. Only 18 percent of Jordanian women who are employed use formal childcare services but a further 69 percent would use those services if they were available, as would nearly half of unemployed Jordanian women. The increased avail- ability of childcare could also help shift childcare expectations of married Jordanian women, potentially reducing barriers to labor force entry.41 d. Additional Analysis Figure A1 of the appendix reveals that having dependents – especially younger dependents – creates financial bur- dens that have not substantially decreased (in cash terms) via fiscal policy. Time-use data (Figure 12) shows that dependents create significant care responsibilities that are met by females whether or not they are working outside the home. Additional analysis in Jordan that places each household type in the pre-fiscal and post-fiscal income distributions would reveal whether and how fiscal policy creates impacts within household types. This is essential in understanding how, for example, households with multiple income earners and multiple dependents could be provided relief from high total tax burdens via tax credits for young dependents. 39. Single women with dependents spend nearly as much time in unpaid work as married women; however single women can share neither income generating responsibilities nor social reproduction responsibilities with an adult partner and will by necessity be the main contributor in both areas. 40. Current utilization of, and willingness to use childcare services in Jordan is low: only 9 percent of households with children under five years of age use some form of childcare; over half rely upon family, friends, or neighbors (World Bank 2022). Only 36 percent of non-users would be prepared to use formal childcare: 16 percent if it were free care, a further 16 percent if it were either free or paid, and 4 percent only if it were paid. 33 6 Conclusions National advocates and international frameworks like the Sustainable Development Goals have called upon governments to use the policy tools at their disposal to reduce income inequality and gender inequality. However, development stakeholders do not have a compact, low-cost, and transparent tool for estimating how fiscal policies are ameliorating or exacerbating inequalities in opportunities or outcomes between genders. This paper has sought to develop such a tool. While our approach cannot illuminate all dimensions of gender inequality, it represents a transparent and low-data-cost way to expand on the well-established and globally prevalent CEQ assessment methodology to help stakeholders determine whether fiscal policy targets gender inequalities in economic production and social reproduction. Our proposal involves no extra data but can incorporate additional data when such data are available. We developed a method for applying FIA to household typologies based on underlying survey information that helps put in focus the different roles men and women play, and the different activities they pursue in paid labor outside the house and in unpaid social reproduction activities inside the house. By bringing social reproduction activities – which are critical for the continued strengthening of social, cultural, legal, and economic institutions –together with economic activities, we can begin to estimate whether and how fiscal policy supports household members in their individual and varied roles. The application of a gendered perspective in fiscal incidence analysis is revealing in a number of ways. It adds greater richness to the standard analysis of fiscal policy’s impact on poverty and inequality by highlighting which male-type or female-type households within a welfare distribution gain or lose from fiscal policy. Because it adds information on the sex, age, and employment status of individual household members, it also reveals the extent to which fiscal policy supports household members in their individual roles and activities. In Armenia, for instance, the household type defined by the sex of the main wage earner was critical for understanding where and how fiscal policy affects household members differently. The direct tax burdens in households where women accounted for 60 percent (or more) of wage earnings were lower than in households where men accounted for 60 percent (or more) of wage earnings. At first glance, it appears that direct tax policy reduces gender gaps in wage earnings and implicitly encourages women’s labor force participation. However, reduced tax burdens in households with women as the primary wage earner might equally reflect more limit- ed opportunities for women, including higher frequency of informal employment, household production for self-consumption, or other economic activities that avoid taxes. In Jordan, women are overwhelmingly responsible for social reproduction activities including the care of dependents, but fiscal policy supporting social reproduction activities does not lead to greater participation by women in paid labor. That is because, in Jordan, social norms and other cultural features create the expectation that married women with or without children, and in households with or without other dependents, will remain out of the workforce. For this reason, among others, fiscal policy is less effective at reducing income inequality between households with and without dependents and with and without women as income earners. However, fiscal policy could play a role by supporting services for children and others who need care. 34 The addition of time-use data can make the analysis even richer. Time-use data demonstrated that in both Armenia and Jordan, some children and elders can aid with household tasks and care for family members. This participation in social reproduction from some dependents can reduce gender gaps in labor earnings. In these countries, therefore, a formal extension of affordable and convenient public childcare or elderly care services should positively influence women’s labor supply decisions as well as provide access to fiscal policies that support social reproduction. In sum, the results discussed in this paper from the cases of Jordan and Armenia bode well for the application of the proposed Engendered CEQ methodology in other countries, in different fiscal and socioeconomic settings, and in settings with different cultural expectations regarding the economic and social roles of men and women. In both Armenia and Jordan, richer results were discovered in a contextualized household, in the application of the CEQ assessment methodology across different household types, and in the incorporation of time-use data as a link between household organization and the fiscal policies under examination. The flexibility of the approach should allow analysts and researchers to carry out an engendered analysis, which in turn can help policy makers understand how fiscal policies treat individuals in their different roles and improve reform efforts that benefit men, women, and children. Going forward, a number of methodological extensions can be explored. We already noted the incorporation of time-use data. Further work can be done to incorporate savings into the notion of income. More granularity can be added to taxes. For instance, it would be interesting to examine in any particular country context the structure of rates and deductions in personal income taxes; for instance, higher marginal rates on secondary income affects women more than men as they are the ones considered to be secondary earners. How can the CEQ be applied to countries with schedular, e.g., different rates for personal income, capital income, and capital gains, as opposed to global rate structures? Other extensions include adding presumptive taxes for small entrepreneurs, land and property taxes, and corporate income taxes, including whether and how dividends are allocated to income. Early work from some countries suggests these other taxes all have different income effects on male and female taxpayers and entrepreneurs (Komatsu and others 2021). Finally, future research can explore a range of policy simulations, for instance, removing subsidies or tax expenditures and replacing them with direct transfers. These extensions can further enrich an already useful tool for understanding fiscal policy. 35 Appendix A Table A1. Example of an Application of Household Types by Quintile in Armenia Household typology All Q1 Q2 Q3 Q4 Q5 Bottom-40 Top-60 Urban Rural Gender of household adult members (T1) Members are female majority 30.1% 31.0% 28.2% 30.8% 28.4% 31.8% 29.6% 30.3% 32.1% 27.1% Members are male majority 20.3% 17.5% 21.6% 18.9% 21.6% 21.7% 19.5% 20.7% 17.5% 24.1% Equal number of females and males 49.7% 51.5% 50.1% 50.2% 50.0% 46.5% 50.8% 48.9% 50.3% 48.8% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Gender of household earners (T2) Earners are female only 16.5% 17.2% 12.3% 19.0% 15.0% 19.0% 14.8% 17.7% 18.5% 13.6% Earners are male only 28.3% 27.2% 31.8% 29.5% 26.6% 26.3% 29.5% 27.5% 25.8% 31.8% Earners are both female and male 34.0% 27.4% 33.0% 34.3% 40.2% 35.3% 30.2% 36.6% 37.7% 28.8% 34 No labor earners 21.2% 28.3% 22.9% 17.2% 18.2% 19.4% 25.6% 18.3% 18.0% 25.8% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Gender sustaining household earnings (T3) Female (>60% earnings) 20.1% 20.0% 15.4% 23.2% 19.0% 22.7% 17.7% 21.6% 22.8% 16.2% Male (> 60% earnings) 58.7% 51.7% 61.7% 59.6% 62.8% 57.9% 56.7% 60.1% 59.3% 58.0% No labor earnings 21.2% 28.3% 22.9% 17.2% 18.2% 19.4% 25.6% 18.3% 18.0% 25.8% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Gender sustaining household earnings + presence of children & elderly (T4) Dependents + Female (>60% earnings) 1.2% 1.3% 0.9% 1.2% 1.6% 1.0% 1.1% 1.2% 1.1% 1.3% elderly Male (> 60% earnings) 3.4% 4.7% 4.1% 2.7% 2.8% 2.9% 4.4% 2.8% 3.3% 3.7% No labor earnings 1.5% 3.0% 1.6% 1.1% 1.0% 0.9% 2.3% 1.0% 1.2% 2.0% Dependents, Female (>60% earnings) 3.9% 4.8% 4.2% 3.4% 3.9% 3.3% 4.5% 3.5% 3.8% 4.1% no elderly Male (> 60% earnings) 19.7% 20.0% 24.5% 20.3% 18.6% 15.2% 22.2% 18.1% 18.0% 22.1% No labor earnings 5.6% 9.8% 6.6% 4.4% 3.5% 3.7% 8.2% 3.9% 4.1% 7.7% Table A1 continued Household typology All Q1 Q2 Q3 Q4 Q5 Bottom-40 Top-60 Urban Rural Gender sustaining household earnings + presence of children & elderly (T4) No dependents, Female (>60% earnings) 4.4% 5.0% 2.7% 6.8% 4.4% 3.2% 3.8% 4.8% 5.3% 3.2% elderly Male (> 60% earnings) 7.4% 7.5% 8.6% 7.0% 6.0% 8.0% 8.0% 7.0% 8.1% 6.4% No labor earnings 7.3% 6.3% 7.8% 6.7% 8.0% 7.8% 7.0% 7.5% 8.2% 6.1% No dependents, Female (>60% earnings) 10.5% 8.9% 7.7% 11.8% 9.1% 15.2% 8.3% 12.0% 12.6% 7.6% no elderly Male (> 60% earnings) 28.1% 19.6% 24.4% 29.6% 35.4% 31.8% 22.0% 32.3% 29.8% 25.8% No labor earnings 6.8% 9.1% 6.9% 5.0% 5.7% 7.1% 8.0% 5.9% 4.5% 10.0% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 35 Table A2. Example of an Application of Household Types by Quintile and Other Characteristics, in Armenia Household typology All Q1 Q2 Q3 Q4 Q5 Bottom-40 Top-60 Urban Rural Gender of household adult members (T1) Members are female majority 100.0% 20.7% 18.8% 20.5% 18.9% 21.1% 39% 61% 63% 37% Members are male majority 100.0% 17.3% 21.3% 18.7% 21.3% 21.4% 39% 61% 51% 49% Equal number of females and males 100.0% 20.8% 20.2% 20.2% 20.1% 18.7% 41% 59% 60% 40% Gender of household earners (T2) Earners are female only 100.0% 20.9% 14.9% 23.1% 18.2% 23.0% 36% 64% 66% 34% Earners are male only 100.0% 19.3% 22.5% 20.9% 18.8% 18.6% 42% 58% 54% 46% Earners are both female and male 100.0% 16.1% 19.3% 20.2% 23.6% 20.7% 35% 65% 65% 35% No labor earners 100.0% 26.7% 21.6% 16.2% 17.2% 18.3% 48% 52% 50% 50% Gender sustaining household earnings (T3) Female (>60% earnings) 100.0% 20.0% 15.3% 23.1% 18.9% 22.6% 35% 65% 67% 33% 36 Male (> 60% earnings) 100.0% 17.6% 21.0% 20.3% 21.4% 19.7% 39% 61% 59% 41% No labor earnings 100.0% 26.7% 21.6% 16.2% 17.2% 18.3% 48% 52% 50% 50% Gender sustaining household earnings + presence of children & elderly (T4) Dependents + Female (>60% earnings) 100.0% 22.3% 14.8% 20.6% 26.2% 16.1% 37% 63% 54% 46% elderly Male (> 60% earnings) 100.0% 27.3% 24.2% 15.5% 16.3% 16.8% 51% 49% 56% 44% No labor earnings 100.0% 39.7% 21.3% 14.6% 12.9% 11.5% 61% 39% 46% 54% Dependents, Female (>60% earnings) 100.0% 24.7% 21.2% 17.3% 19.8% 16.9% 46% 54% 57% 43% no elderly Male (> 60% earnings) 100.0% 20.3% 24.8% 20.6% 18.9% 15.4% 45% 55% 54% 46% No labor earnings 100.0% 35.1% 23.5% 15.6% 12.6% 13.2% 59% 41% 43% 57% No dependents, Female (>60% earnings) 100.0% 22.4% 12.2% 30.7% 20.0% 14.6% 35% 65% 71% 29% elderly Male (> 60% earnings) 100.0% 20.2% 23.2% 18.9% 16.2% 21.6% 43% 57% 64% 36% No labor earnings 100.0% 17.4% 21.2% 18.2% 22.0% 21.2% 39% 61% 66% 34% No dependents, Female (>60% earnings) 100.0% 17.0% 14.5% 22.4% 17.3% 28.9% 31% 69% 70% 30% no elderly Male (> 60% earnings) 100.0% 13.9% 17.3% 21.1% 25.1% 22.6% 31% 69% 62% 38% No labor earnings 100.0% 26.9% 20.5% 14.9% 16.8% 20.9% 47% 53% 39% 61% Example of an application of household types by quintile and other characteristics, in Pakistan Another example of how the household typology can be constructed is in Figure A1, based on the Pakistan House- hold Integrated Economic Survey (HIES). Household types are represented by different bubble colors, and the frequency wherein a household type is represented by the size of the bubble. In this schematic, household types are placed into quintiles of the post-fiscal income distribution to demonstrate which households have access to natural gas subsidies. The figure demonstrates that larger households with multiple dependents and multiple wage earners are much more prevalent among the bottom 40 percent of the population (ranked by post-fiscal income) and very few households overall in the bottom 20 percent of the population have access to natural gas subsidies. Figure A1. Household Types in Pakistan 160 140 120 100 80 60 40 0 1 2 3 4 5 1-2 members, no dependents 1-2 dependents, any age Multiple dependents, multiple earners Young + old-age dependents; multiple working-age adults >10 members total 37 References Akresh, Richard, Damien de Walque, and Harounan Kazianga. 2016. “Evidence from a randomized evaluation of the household welfare impacts of conditional and unconditional cash transfers given to mothers or fathers.” Development Research Group Human Development and Public Services Team Working Paper 7730, World Bank, Washington, DC. Alhawarin, Ibrahim, Ragui Assaad, and Ahmed Elsayed. 2020. “Migration Shocks and Housing: Short-Run Im- pact of the Syrian Refugee Crisis in Jordan.” Institute of Labor Economics (IZA) Discussion Paper 13969. Ambel, Alemayehu, Wondimagegn Tesfaye, and Manex Yonis. 2022. “A Gendered Fiscal Incidence Analysis for Ethiopia: Evidence from Individual-Level Data.” Policy Research Working Paper WPS 10130. World Bank, Washington, DC. Armand, Alex, Orazio Attanasio, Pedro Carneiro, and Valérie Lechene. 2016. “The Effect of Gender-Targeted Conditional Cash Transfers on Household Expenditures: Evidence from a Randomized Experiment.” CEPR Discussion Paper, DP11465. Aziz, Omar, Norman Gemmell, and Athene Laws. 2013. “The Distribution of Income and Fiscal Incidence by Age and Gender: Some Evidence from New Zealand.” Victoria University of Wellington Working Paper in Public Finance No. 10/2013. Bakhtiar, Mehrab, Marcel Fafchamps, Markus Goldstein, Kenneth L. Leonard, and Sreelakshmi Papineni. 2022. “Women’s Empowerment and the Intrinsic Demand for Agency: Experimental Evidence from Nigeria.” NBER Working Paper w30789. Bakker, Isabella, and Lisa Philipps. 2018. “Gendering the Analysis of Tax Expenditures: Bridging Two Solicitudes in Canadian Fiscal Policy.” 2018 New York University Tax Policy (colloquium presentation). Bargain, Olivier. 2022. “Income Sources, Intra-Household Allocation, and Individual Poverty.” Commitment to Equity (CEQ) Working Paper Series 121, Tulane University, Department of Economics. Barnett, Kathleen and Caren Grown. 2004. Gender Impacts of Government Revenue Collection: The Case of Taxation. Commonwealth Secretariat Economic Paper 62. Commonwealth Secretariat. Benhassine, Najy, Florencia Devoto, Esther Duflo, Pascaline Dupas, and Victor Pouliquen. 2015. “Turning a Shove Into a Nudge? A ‘Labeled Cash Transfer’ for Education.” American Economic Journal: Economic Policy 7 (3): 86–125. 38 Berik, Günseli, and Ebru Kongar, eds. 2021. Routledge Handbook of Feminist Economics. Abingdon: Routledge. Berik, Günseli, Lourdes Beneria, and Maria Floro. 2016. Economics as if All People Mattered. New York: Taylor and Francis. Bjorvatn, Kjetil, Denise Ferris, Selim Gulesci, Arne Nasgowitz, Vincent Somville, and Lore Vandewalle. 2022. “Childcare, Labor Supply, and Business Development: Experimental Evidence from Uganda.” Gender, Growth, and Labor Markets in Low Income Countries Program Working Paper 67. IZA. Elissa Braunstein, Rachid Bouhia, Stephanie Seguino. 2020. “Social Reproduction, Gender Equality and Econom- ic Growth.” Cambridge Journal of Economics 44 (1): 129–156. Browne, James. 2011. The Impact of Tax and Benefit Reforms by Sex: Some Simple Analysis. London: IFS. Browning, Martin, Pierre-Andre Chiappori, and Arthur Lewbel. 2013. “Estimating consumption economies of scale, adult equivalence scales, and household bargaining power.” Review of Economic Studies 80 (4): 1267–1303. Boerma, Ties, Ahmed Reza Hosseinpoor, Emese Verdes, and Somnath Chatterji. 2016. “A global assessment of the gender gap in self-reported health with survey data from 59 countries.” BMC Public Health 16: 675. Bose-Duker, Theophiline, Michael Henry, and Eric Strobl. 2021. “Children’s Resource Shares: Male Versus Fe- male-Headed Households,” Journal of Family and Economic Issues 42 (4): 573-585. Bucheli, Marisa, and Cecilia Olivieri. 2013. Gender Equity in Taxation in Latin America and the Caribbean. Wash- ington, DC: Inter-American Development Bank. Candia, Bernardo, and Eduardo Engel. 2018. “Taxes, Transfers, and Income Distribution in Chile: Incorporating Undistributed Profits.” CEQ Working Paper No. 82. Tulane University, Department of Economics. Casale, Daniela Maria. 2012. “Indirect taxation and gender equity: Evidence from South Africa.” Feminist Eco- nomics 18 (3): 25–54. Chakraborty, Lekha, Marian Ingrams, and Yadawendra Singh. 2018. “Fiscal Policy Effectiveness and Inequality: Efficacy of Gender Budgeting in Asia Pacific.” NIPFP Working Paper No. 18/224. New Delhi. Chamberlain, Andrew, Daniel Zhao, and Amanda Stansell. 2019. “Progress on the Gender Pay Gap: 2019.” Glass- door Economic Research Report. 39 Charmes, Jacques. 2022. “Variety and Change of Patterns in the Gender Balance Between Unpaid Carework, Paid Work and Free Time Across the World and Over Time: A Measure of Well-Being?” Wellbeing, Space, and Society. Volume 3. Charmes, Jacques. 2019. The Unpaid Care Work and the Labour Market: An Analysis of Time Use Data Based on the Latest World Compilation of Time Use Surveys. Geneva: International Labour Organization. Chiappori, Pierre-André, Bernard Fortin, and Guy Lacroix. 2002. “Marriage Market, Divorce Legislation, and Household Labor Supply.” Journal of Political Economy 110 (1): 37–72. De Henau, Jérôme, Susan Himmelweit, Zofia Lapniewska, and Diane Perrons. 2016. Investing in the Care Economy: A gender analysis of employment stimulus in seven OECD countries. International Trade Union Confederation. De Henau, Jérôme, Debbie Budlender, Fernando Filgueira, Ipek Ilkkaraçan, Kijong Kim, and Rafael Mantero. 2019. Investing in Free Universal Childcare in South Africa, Turkey and Uruguay: A comparative analysis of costs, short-term employment effects and fiscal revenue. New York: UN Women. Demery, Lionel. 1996. “Gender and Public Social Spending: Disaggregating Benefit Incidence.” Poverty and Social Policy Department Discussion Paper, Washington DC: World Bank. Dinkelman, Taryn, and Rachel Ngai. 2022. “Time Use and Gender in Africa in Times of Structural Transforma- tion.” Journal of Economic Perspectives 36 (1). Winter. Donald, Athelia, Gayatri Koolwal, Jeannie Annan, Kathryn Falb, and Markus Goldstein. 2020. “Measuring Wom- en’s Agency.” Feminist Economics 26 (3): 200–226. Doorley, Karina, and Claire Keane. 2020. “Tax-Benefit Systems and the Gender Gap in Income.” IZA Discussion Paper No. 13786 (available at SSRN). Duflo, Esther. 2012. “Women Empowerment and Economic Development.” Journal of Economic Literature 50 (4): 1051-1079. Dunbar, Geoffrey R, Arthur Lewbel, and Krishna Pendakur. 2013. “Children’s Resources in Collective House- holds: Identification, Estimation, and an Application to Child Poverty in Malawi.” The American Economic Review 103 (1): 438–471. 40 Ferrant, Gaëlle, Luca Maria Pesando, and Keiko Nowacka. 2014. “Unpaid Care Work: The Missing Link in the Analysis of Gender Gaps in Labor Outcomes.” Organization for Economic Cooperation and Development (OECD) Discussion Paper. Filmer, Deon and Lant Pritchett. 1999. “The Impact of Public Spending on Health: does Money Matter.” Social Science and Medicine 49 (10): 1309-1323. Fragoso, Lucia Perez, and Francisco Gonzalez. 2010. “Gender Analysis of Taxation in Mexico.” In Taxation and gender equity: A comparative analysis of direct and indirect taxes in developing and developed countries, edited by Caren Grown and Imraan Valodia. Routledge International Studies in Money and Banking. Abingdon: Routledge. Fuchs, Alan, and Maria Fernanda Gonzalez Icaza. 2022. Assessing the effects of the fiscal system on gender dispari- ties in Armenia. World Bank: Washington, DC. Gaddis, Isis, and Lionel Demery. 2012. “Benefit incidence analysis, needs and demography: Measurement issues and an empirical study for Kenya.” Courant Research Centre: Poverty, Equity and Growth – Discussion Pa- pers 122, Courant Research Centre PEG. Garcia-Peña Bersh, Natalia. 2019. “A gendered fiscal incidence analysis for Barbados.” Master of Economics The- sis, Departamento de Economia, Universidad de San Andres, Buenos Aires, Argentina. Glick, Peter, Rumki Saha, and Stephen D. Younger. 2004. “Integrating Gender into Benefit Incidence and Demand Analysis.” Cornell Food and Nutrition Policy Program Working Paper No. 167. Greenspun, Samantha. 2019. “A gender sensitive fiscal incidence analysis for Latin America: Brazil, Colombia, The Dominican Republic, Mexico, and Uruguay.” PhD Dissertation, Tulane University. Grown, Caren, and Hitomi Komatsu. 2010. “Methodology and Comparative Analysis.” In Taxation and Gender Equity: A Comparative Analysis of Direct and Indirect Taxes in Developing and Developed Countries, edited by C. Grown and I. Valodia, Abingdon: Routledge. Grown, Caren, and Imraan Valodia, eds. 2010. “Taxation and gender equity: A comparative analysis of direct and indirect taxes in developing and developed countries.” Routledge International Studies in Money and Bank- ing. Abingdon: Routledge. Haddad, Lawrence, John Hoddinott, and Harold Alderman. 1997. “Intrahousehold Resource Allocation in Devel- oping Countries: Models, Methods, and Policies.” IFPRI Food Policy Statement 24. 41 Heckman, James, and Rasmus Landersø. 2021. “Lessons from Denmark about Inequality and Social Mobility.” NBER Working Papers 28543, National Bureau of Economic Research, Cambridge, Massachusetts. Ilkkaracan, Ipek, Kijong Kim, and Tolga Kaya. 2015. “The impact of public investment in social care services on employment, gender equality, and poverty: The Turkish case.” Research Project Report, Istanbul Technical University, Women’s Studies Center in Science, Engineering and Technology, and the Levy Economics In- stitute, in partnership with ILO and UNDP Turkey, and the UNDP and UN Women Regional Offices for Europe and Central Asia. ILO. 2023. “Labor Force Statistics (LFS, STLFS, RURBAN Databases).” Kalil, Ariel, Hope Corman, Dhaval M. Dave, Ofira Schwartz-Soicher, and Nancy Reichman. 2022. “Welfare Re- form and the Quality of Young Children’s Home Environments.” National Bureau of Economic Research Working Paper 30407. Cambridge, MA: National Bureau of Economic Research. Komatsu, Hitomi, Alemayehu Ambel, Gayatri Koolwal, and Manex Yonis. 2021. “Gender and Tax Incidence of Rural Land Use Fee and Agricultural Income Tax in Ethiopia.” Policy Research Working Paper WPS 9715, World Bank, Washington, DC. Lagarde, Christine, and Vito Gaspar. 2018. “Give Today’s Children a Chance.” IMF Blog, September 24, 2018. Lahey, Kathleen. 2017. “Australian Tax-transfer Policies and Taxing for Gender Equality: Comparative Perspec- tives and Reform Options.” In Tax, Social Policy and Gender: Rethinking Equality and Efficiency, edited by M. Stewart. Canberra: Australian National University Press. Lewbel, Arthur, and Krishna Pendakur. 2008. “Estimation of Collective Household Models with Engel Curves.” Journal of Econometrics 147 (2): 350-358. Lustig, Nora. 2018. “The Sustainable Development Goals (SDGs), Domestic Resource Mobilization and the Poor.” G20 Insights Policy Brief. Lustig, Nora, ed. 2021. Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty. Washington, DC: Brookings Institution Press. Molyneux, Maxine, and Marilyn Thomson. 2011. “Cash Transfers, Gender Equity and Women’s Empowerment in Peru, Ecuador and Bolivia.” Gender and Development 19 (2): 195–212. Pailhe, Ariane, Anne Solaz, and Maria Stanfors. 2020. The Great Convergence? Gender and Unpaid Work in Europe and the United States. hal-02537017. 42 Pritchett, Lant, and Deon Filmer. 1999. “What Education Production Functions Really Show: A Positive Theory of Education Expenditures.” Economics of Education Review 18 (2): 223-239. Quisumbing, Agnes R., and John A. Maluccio. 2000. “Intrahousehold Allocation and Gender Relations: New Em- pirical Evidence from Four Developing Countries.” Food Consumption and Nutrition Division (FCND), the International Food Policy Research Institute (IFPRI) Discussion Paper 84. Rodriguez-Enriquez, Corina, and Nicolas Aguila. 2017. “Gender Bias of Regressive Taxation in Latin America: Overview and Exploration of the Argentinean Case.” In Rethinking Taxation in Latin America, edited by J. Atria, C. Groll, and M. Valdes. London: Palgrave MacMillan. Rodriguez, Laura, Matthew Wai-Poi, and Jeffrey Woodham. 2022. Taxes, Transfers and Gender: Fiscal Policy Inci- dence Across Fiscal and Care Categories in Jordan. Washington, DC: World Bank. Rossignolo, Dario. 2018. Gender equity in the Argentine tax system: An estimation of tax burdens by household type. Santiago, Chile: Economic Commission for Latin America and the Caribbean (ECLAC). Rossouw, Laura, and Hana Ross. 2020. An Economic Assessment of Menstrual Hygiene Product Tax Cuts. Rubiano, Eliana, and Mariana Viollaz. 2019. “Gender Differences in Time Use: Allocating Time Between the Mar- ket and the Household.” Policy Research Working Paper No. 8981, World Bank, Washington, DC. Sahn, David E., and Stephen Younger. 2000. “Expenditure Incidence in Africa: Microeconomic Evidence.” Fiscal Studies 21 (3): 329–347. Selden, Thomas M., and Michael J. Wasylenko. 1995. “Measuring the Distributional Effects of Public Education in Peru.” World Bank Report. Spiess, Katharina C., and Katharina, Wrohlich. 2008. “Does Distance Determine Who Attends a University in Germany?”. German Socio-Economic Panel Study (SOEP) Paper 118. Statistical Committee of the Republic of Armenia (SCRA). 2019. Social Snapshot and Poverty in Armenia 2019. Stotsky, Janet. 1997. “How Tax Systems Treat Men and Women Differently.” Finance and Development Magazine. Washington, DC: International Monetary Fund. Stewart, Miranda, ed. 2017. Social Policy and Gender: Rethinking Equality and Efficiency. ANU Press. Thomas, Duncan. 1993. “The Distribution of Income and Expenditure within the Household.” Annales d’Écono- mie et de Statistique 29: 109–135. 43 USAID. 2019. Armenia Gender Analysis Report. Prepared by Banyan Global for the United States Agency for International Development. Washington DC, August. Vijil, Ana Lucia Alvarez. 2017. “Incidence of the Value-Added Tax on Men and Women in Nicaragua.” Master of Public Administration Thesis, Cornell University. World Bank. 2012. World Development Report 2012: Gender Equality and Development. Washington, DC. World Bank. 2020. State of the Mashreq Women 1:Women’s economic participation in Iraq, Jordan and Lebanon. World Bank. World Bank. 2023a. “World Development Indicators.” World Bank. 2023b. “World Bank Gender Data Portal.” Zwijnenburg, Jorrit, Sophie Bournot, David Grahn, and Emmanuelle Guidetti. 2021. “Distribution of household income, consumption and saving in line with national accounts: Methodology and results from the 2020 collection round.” OECD Statistics Working Papers No. 2021/01, OECD Publishing, Paris. 44