This is indeed a modest proposal. It involves no radical change in either the definition of the tax base or the progressivity of the rate schedule. It does not claim to solve all of the problems surrounding the taxation of families. But we cannot ignore the opportunity to make small but positive changes while waiting for a massive tax reform which may never come. The elimination of joint filing would be one step forward toward a more equitable and efficient tax structure. — Harvey S. Rosen1
A crucial deficiency in the model of the single utility-maximizing individual is the simple dichotomization of time into market work and leisure. To tackle questions of inequality, taxation, labor supply, and fertility, economic models must incorporate time spent on the production of goods and services within the household.2 In the past, it was women who exclusively devoted their time to household production, and standard models kept them hidden. Although the number of women employed in high-paid jobs has increased dramatically, women still disproportionately contribute their time to household production, sacrificing their economic success when they become mothers.3 Even now, “the discrepancy between investments and biology is a source of conflict and even agony for the biologically deviant.”4 Instead of directing policy efforts toward the problem of female labor force participation to reflect changing social attitudes about the respective roles of men and women, the United States has largely shirked any commitment to gender equality. A simple consideration of gender budgeting reveals that there is a clear “presumption of gender bias in fiscal budgetary processes.”5 As Joseph A. Schumpeter has suggested, “The budget is not just about spending money”; it is about deciding how an economy might be organized; it reflects “the aims and ambitions of the nation state.”6 Only with adequate models and appropriate policies can states address changing attitudes about men’s and women’s roles in society.
This article analyzes the effect of public policy — in particular the system of child support and taxation — on household labor supply and fertility decisions. It takes as its starting point the problem of the standard workhorse model of welfare economics, which specifies an individual decision-maker: a single consumer-worker who chooses to divide his time between leisure and market labor. While models of this type remain relevant, there are reasons to revise it. Its usefulness is limited because “it provides an inadequate basis for obtaining a satisfactory understanding of household decisions.”7 It does not sufficiently reflect the reality of multiperson households. This has major implications for the development of public economic policy.
Given the significantly improved collection of empirical evidence over the last four decades, providing more detailed information on household consumption, production, and time-use decisions, it is necessary to move beyond the black-box assumption of intra-household decision-making in order to advance alternative models.8 In this article, I target the theoretical basis of the approach to estimating the costs of having children as market consumption costs and the common assumption that household income is a reliable indicator of a household’s standard of living. Each point has serious implications for policy design. The first offers justification for the government to invest in child care facilities for children of all ages, while the second provides a critique of basing taxation on household income as opposed to individual income.
The analysis is provided in the next three sections. The next section expands on how the economics literature has typically defined child costs as the amount required to keep parental utility constant. I discuss the basic conceptual difficulties in defining child care and review several exceptions to the standard theoretical approach. This includes brief scrutiny of child benefit systems and family policy patterns. There are considerable negative effects of an inadequately developed child care system on female labor supply.9 The public provision of some private services — such as day care — could stimulate labor force participation and effectively increase the overall tax base.10
The following section evaluates common assumptions that underpin the traditional model of the household and clarifies why the specification of work and leisure must be reexamined. Using household income as a measure of welfare is especially problematic because it ignores how domestic production contributes to family welfare. Although policies that reflect this assumption “amount to a form of gender bias,” they still appear in systems of joint and even quasi-joint taxation: when the formal tax system is based on individual income but family benefits are withdrawn on the basis of household income.11 Inadequate consideration has been given to the issue of income taxation as a tax on monetized trade between the household and market sectors when there is identifiable segregation by sex. Given this distinct bias in the design of tax and transfer policies, which reveals strong disincentives for the second earner, I reference the growing literature that supports a revenue-neutral switch toward individual taxation.12
The final section concludes with a general analysis of the effect of public policy on birth rates and female labor supply, referring to the stark change from a negative to a positive correlation between female employment and fertility over the last 25 years. It is possible to encourage through both appropriate tax and family policies.13
The True Costs of Children
A sufficient measure of the costs of children is required in various policy contexts. Horizontal equity in taxation requires individuals in similar economic circumstances — with similar ability to pay — to pay similar taxes.14 This implies that the tax levied on a given income should take into consideration whether that income has to support dependent children. It also implies that the transfer a household receives should depend on how many children it contains.15 But some critics of deductions and credits for children insist that a decision to have a child is a consumption decision — that “the government should not favor one form of consumption over another.”16 Increased fairness and efficiency are major impetuses for tax reform. In the United States, there is a static view of horizontal equity — that joint income is the right measure of fairness. There is a long-standing belief that the benefits of redistributing on joint income outweigh the distortive effects of the higher tax on secondary earners that it creates. This claim is frequently made in defense of the earned income tax credit, a refundable tax credit to lower-income families.
The importance of analyzing the EITC is underscored by the fact that after supplanting welfare as the primary means through which the United States supplies economic support to disadvantaged families,17 it is now generally considered the most important cash transfer program for lower-income working families.18 Advocates suggest that while the EITC schedule is slightly larger for married couples, placing a higher marginal tax rate on secondary earners, this does not significantly reduce the distortion.19 But these reforms represent a shift in the wrong direction. The larger EITC schedule for married couples, the high marginal tax rate on secondary earners, and notable reduction in tax rates on higher incomes have increased inequality and decreased efficiency. This transition reflects a distinct ideological position, exaggerating the effects on top income earners’ “incentives to earn income of tax rates that place on them a fairer share of the overall tax burden.”20 This issue is expanded in the following section, with regards to the strong negative incentive effects for secondary earners.
Any insistence that universal family benefits are more costly, for example, because they provide part compensation for the direct and indirect costs of having children instead of tightly targeting family payments on the basis of reducing child poverty “reflects a misunderstanding of the trade-off between efficiency and equity in tax design.”21 There are grave consequences for inadequately assessing the costs of children. Children are incredibly costly in terms of parental time. If reforms do not support the reallocation of parental time from the home to the market, women will disproportionately suffer.
Considering that the costs of having children are generally perceived as market consumption costs, there is often little if any consideration of the element of child costs that takes the form of parental time. Upon the arrival of the first child, parents will inevitably divert their time from alternative uses to child care, and it is disproportionately women’s time “on the front line of the trade-off between production and procreation — the work-family interface.”22 Time spent on child care reflects resources devoted to children and can be used as input for determining the average total cost of children.23 Defining child costs in terms of a child’s physiologically determined “basic needs” is unfavorable among economists because it implicitly identifies “welfare” with a limited set of measures of physical well-being.24 Instead of accepting a definition of poverty in absolute physiological terms, the strategy has been to define it in relative terms. But this requires estimating the amount of time that households spend on their children. The approach adopted by Apps and Rees defines the cost of a child as “an outcome of the choice of an intra-family distribution of consumption,” meaning the amount of consumption of all goods — such as child care — a household allocates to its children at an equilibrium.25 In a world of perfect markets and complete information, the assumption is that parents will not incur child costs because they can use the capital market to smooth their consumption. In reality, it is the incompleteness in capital markets that necessitates intergenerational transfers to cover child costs and strengthens the case for public involvement to support these costs.26 The problems associated with defining the costs of children can have considerable negative effects on the formulation of policy. For this reason, the modeling approach to this estimation problem must include more than just consumption expenditure data.
Given that the modeling approach adopted in much of the literature fails in its empirical applications to extend beyond a simplified model of the household, as a single individual who maximizes a utility function subject to constraints on time and income, children still often feature as parameters. The household utility function is considered “that of the head” and child costs are defined as the amount of money required to hold parental utility constant.27 Policy studies that use this approach tend to generate specific assumptions about the intra-household distribution of welfare: “Common practice is to assume they are all equally well off.”28 Clearly, the household utility function approach is inadequate. It is insufficient in analyzing policy issues that involve the welfare of the individuals. An appropriate alternative involves retaining the individual as the basic unit of analysis. Apps and Rees discard the approach based on holding parental utility constant and present a model that treats all household members — including children — as distinct individuals with separate utility functions.29 The intention is to incorporate two key concerns: the issue of parental time and the differential incidence of child costs on the adult members of the household. Domestic production is crucial.30 Empirical household models must define the cost of a child in terms of its “full consumption,” taking into account “the sum of consumption market goods, domestically produced goods, and parental time in the form of child care.”31 Adults are expected to divvy up their time among market sector work, child care, domestic production, and pure leisure, as well as select an allocation of market income over bought-in inputs and consumption goods into the domestic production process.32 The goal is to avoid misleading results regarding the behavioral responses to economic policy and the distribution of income among family members within the household.
Even though the importance of including parental time as a major part of the costs of a child has been widely recognized in the theoretical literature, it has been generally overlooked in the empirical work. Still, there have been noteworthy exceptions. For instance, Ugo Colombino criticizes the general approach for its treatment of children as an exogenous characteristic of the household, insisting that “children bring (or may bring) utility to the household.”33 He emphasizes that the traditional concept of the cost of children, as well as the associated equivalence scale, is insufficient for the purpose of designing family policy because it can lead to “wrong or paradoxical policy prescriptions.”34The concern is that this has not been taken seriously in empirical research concerning social policy design. He presents a model of household resource allocation and accounts for child costs in terms of the consumption allocation. This includes parental time that the child receives. Although his approach does not estimate the incidence of child costs on respective parents and fails to distinguish among household types, it remains relevant. He clarifies that family policy should clearly target a social goal and compute a strategy that maximizes that goal under the corresponding constraints, such as household behavioral responses. For Colombino, the strategy should not involve compensating households with children.
Another approach, adopted by Björn Gustafsson and Urban Kjulin, draws on Swedish time-use data and net wages in to estimate the time costs of child care and other types of child-induced housework among families. The total cost of children is determined by the cost of time as well as the costs of goods, services, and housing. Their approach distinguishes between mother’s and father’s inputs and allows the effect of children on child care and housework to vary by age of the child, as well as whether the child is cared for outside the home. The results reveal that financing the time generated by children requires a substantial decrease in market labor supply, and as expected, most of the time that is used for child care is provided by women.35 They concede that the division between spouses of time devoted to child care and housework can be made more equal by providing adequate day care.36
Despite convincing evidence that exposes the incredible costs of children in terms of parental time, the United States has largely left the responsibility for children to their natural families. The system is composed of family-income-targeted child payments that impose substantial disincentives for women to participate in the labor force, and the few policies regarding the provision of child care services on the national level broadly assume that women are the primary caretakers of children.37 This is a stark example of government philosophy that favors leaving child care to the market, restricting any intrusion “for fear of reducing ‘efficiency.’”38 The logic is deeply flawed. Given that the United States has failed to assume adequate responsibility for child care and education sectors for children under school age, these sectors are “expensive, poorly developed, and frequently difficult to access.”39 Instituting reforms that allow for the reallocation of parental time — primarily female labor — from the home to the market has the potential to increase productivity and expand the tax base.40 The gender-based division of responsibilities for social reproduction effectively “constrains women’s lives”; most women never fully recover their initial market wage in the post-child phase.41 Olivier Thévenon confirms that the provision of accessible child care services for children under three years old plays a decisive role in allowing women to balance their family responsibilities and work commitments. Still, the “cross-national variations in public money invested in the provision of education and childcare services are wide.”42 Variation in child care policies reflects variation in social rights and protections across regime types, exposing the extent to which states assume familial care responsibilities. In terms of family policy, some countries are beginning to question the ability of welfare states to tackle “new social risks” — like that of reconciling market sector work and care — because of increased pressures for “market-driven” reforms.43 Still, an extensive and inexpensive child care provision serves two important goals: reducing social inequalities by enabling women to combine family life and work — effectively boosting labor supply — and providing essential social investment in children, who will then become “better workers and taxpayers.”44 It is essential to remedy the problems associated with defining the costs of children. An insufficient approach can have considerable negative effects on the formulation of policy.
The Case for Individual Taxation
Changes in female labor supply are determined by the method of substitution between market and household work that involve bringing up children. There is a high degree of heterogeneity in the labor supply of second-income earners after the first child. While the male partner usually continues to work full-time after the arrival of children and only about one-third of female partners still allocate their time exclusively to home production across OECD countries, there remains a high degree of across-household variation in the total hours that women contribute to market work. The incredible increase in female labor force participation since the early 1950s marks a distinctive trend and signals that the traditional model of the household is no longer appropriate. It is no longer assumed that the male head specializes in market work and the female in household work.45 Typically, the traditional model of the household defines the life cycle based on the age of the male partner — the “head” — in a household couple. The family is treated as a single person. Demographic characteristics are treated as control variables. While keeping constant the presence of a spouse or dependent children allows for closer scrutiny of other variables, there is a significant cost to this approach. Fertility decisions are not accounted for. The norm is to average income and consumption across couples with and without children, and that gives the impression that female earnings track male earnings — that the female can be overlooked. Because this conventional approach does not construct an empirically realistic model of the couple household, it is necessary to consider a new one: a model that accounts for fertility decisions on the life cycle consumption and saving profiles. Any change in household income, consumption, and saving is largely determined by these fertility decisions and their effect on female labor supply. Again, the model must adapt to reflect available data on time use and household expenditure. It must integrate the multiperson nature of the household to reflect two-earner couples with at least one young child and factor in the allocation of time to household production by the second earner.46
This highlights a primary issue in the design of tax and transfer policies: the choice of household income as a measure of welfare. It is especially problematic because it leaves out the contribution of domestic production to family welfare.47 The specification of work and leisure needs to be reexamined; the interpretation of work as labor supply to market production requires scrutiny.48 In a given production period, individuals have a fixed endowment of time that is allocated to production for trade and to production for personal consumption. Trade is not limited to the exchange of market goods or household services. It refers to any sort of exchange or social interaction between individuals. Households represent groups of trading individuals. Any division of activity into market and household sectors is “purely institutional.”49 All activity is in some measure work and in some measure leisure, regardless of whether it takes place in the market sector or in the household sector. Still, as Apps identifies, there is frequent confusion about the nature of household production for trade. She confirms that individuals produce for trade in the household sector even if they are not the sole consumers of their products. For example, a dependent spouse may allocate time to the production of household services for the primary earner of the household, who is employed in the market sector. Given the institutional constraints on female employment possibilities, women are typically limited to work in the household sector. Undoubtedly, “there is job segregation on the basis of sex” and “the two segregated groups trade.”50 The trade between women and men of household services is monetized given that women produce household services purchased by men and use their incomes to purchase market sector goods. But the household sector cannot avoid a tax on trade between the two sectors by producing similar goods as the market sector for equal production costs. The low-wage group employed in the household sector can pay absolutely more tax because the greater part of the income taxation burden can fall on the household sector.
While optimal tax theory considers the distortionary effects of the untaxability of leisure, not enough consideration has been given to the issue of income taxation as a tax on monetized trade between the household and market sectors when there is identifiable segregation by sex and increased opportunities for leisure in the company as wages rise. As Apps insists, in a first best world the rules for optimal taxation are inappropriate in the presence of institutional constraints on job choices, regardless of the possibility to tax production for own consumption. There remains distinct bias in the design of tax and transfer policies, resulting in strong negative incentive effects for the second earner. This is especially true in systems of joint and even quasi-joint taxation, in which the formal tax system is based on individual income but family benefits are withdrawn on the basis of household income and the implementation of family-income-targeted child payments, as well as notable reduction in tax rates on higher incomes over recent decades.51 The United States is one of only seven OECD countries to apply mandatory joint taxation.52 The transition from joint to individual taxation has been increasing among OECD countries since the 1970s, based on the conclusion that joint taxation lacks marriage neutrality. Sara LaLumia evaluates the effects of joint taxation of married couples on labor supply and non-wage income in the United States since joint taxation was introduced at the federal level in 1948.53 Joint taxation has the effect of equalizing the marginal tax rates imposed on a husband and wife in the same household. Given that men typically represent the primary income earner and women the secondary income earner, LaLumia asserts that the introduction of joint taxation effectively lowered the marginal tax rates on men and raised them on women. Her empirical results confirm the argument that the implementation of joint taxation reduces the labor supply of married women and that the effect on the labor supply of married men is theoretically ambiguous. This piece contributes to the growing literature on the appropriate tax treatment of families.
The insistence is that the tax rate on primary income earners should be higher than the tax rate on secondary income earners if the labor supply of the latter is more elastic — that the optimal tax rates on primary income earners can be, and should be, two to three times as large as the optimal tax rates on secondary income earners.54 In an evaluation of the effect of income tax on wage-earning women in Spain since the latest changes in the personal income tax (PIT), Paloma De Villota goes further to suggest that any tax reform that reduces the highest marginal rates in the tax scale will have a definite gender impact and undeniably benefit men.55 The implementation of the PIT in Spain occurred in the late 1970s during Spain’s transition to a democratic system of government. This large-scale tax reform happened in three phases that are linked to tax administration, direct taxation, and indirect taxation. The overarching goal was to implement a modern tax regime in harmony with existing systems throughout the European Economic Community. The PIT was initially considered a uniform tax that treats all incomes the same regardless of nature and origin. But it has undergone crucial changes since the Constitutional Court’s verdict declaring compulsory joint taxation unconstitutional. Now it is an optional tax and individuals can choose whether to fill out a joint tax return or an individual tax return. Essentially, family tax returns are still valid and about 30.19 percent of tax returns are still joint.
Assuming that women receive much lower incomes than men and that the total number of women participating in market sector work is substantially lower than the number of men, the first marginal rates of the PIT have affected a higher proportion of women than men. Any change in taxation that influences the lower bands of the tax scale will affect women to a larger extent.56 This has serious implications for the design of the EITC schedule in the United States, given that it is slightly larger for married couples and places a higher marginal tax rate on secondary earners. Considering that male labor supply is typically unresponsive to a change in the net wage, Apps maintains that implementing a strongly progressive rate scale and universal family payments has the potential to minimize the gender gap in labor supply.57
Birth Rates and Female Labor Supply
The previous two sections have examined how public policy influences female labor force participation. Appropriate family and tax policies have the power to effectively combat the seemingly intractable institutional constraints that encourage the formation of single-earner households and the dependency of married women. Some policies have the potential to promote and sustain a positive correlation between female employment and fertility. Cross-country differences likely result from the effects of tax and social security arrangements, as well as the cost and accessibility of child care outside the home. Data collected between 1970 and 1990 tracking female employment and the total fertility rate reveal starkly varied results.58 Before 1990, developed economies witnessed a negative relationship between female employment and fertility over time. But the results following 1990 signify that this negative correlation no longer holds. This comparison exposes various concerns that states must consider in the debate about how to reverse a decline in fertility rates. The most often proposed policy measure to combat fertility decline and induce women to exit the labor force and return to the household is that of increasing child-related cash transfers to households. Still, this is in no way the only option. Empirical work — such as that by Angela Luci-Greulich and Thévenon, Hans Fehr and Daniela Ujhelyiova, and Apps and Rees — implies that another strategy would be to increase the quality and accessibility of the child care provision to adjust the terms of the trade-off between female labor supply and family size.59 The reconciliation of family and work remains a crucial challenge for policy makers in developed countries. Among high-income countries, none are solidly above the fertility rate of 2.1 children per woman required to replace the population at a constant level.60
The aging process associated with the trend of low average birth rates across rich states tends to inflict various concerns for social security systems and labor markets.61 But the exclusive focus on potential budget deficits attributable to a rising aged dependency ratio (ADR) has curbed concern of the possible gains from a decrease in the child dependency ratio (CDR). Both the ADR and CDR must be weighted by cost: On average, a child is much more costly than a retiree. Children are incredibly costly in terms of parental time. Each child requires at least a decade of parental and public investment in her education, and each preschool child necessitates full-time care.62 Although the growth in the ADR is a direct consequence of the drop in fertility rates and family size among developed states, it is both problematic and costly to suggest a solution that relies on women exiting the labor force to provide child care services in the household. Given the issues of inappropriate tax policy and inadequate family policy identified in the previous sections, there is reason to suggest that a high-quality public child care and education system — especially for children under age 3 — can achieve an increase in female employment rates and birth rates simultaneously. For instance, Fehr and Ujhelyiova develop an overlapping generations equilibrium model for Germany to propose a set of reform options for child benefits and family taxation that would encourage a positive correlation between total fertility rates and female labor force participation.63 They maintain that — holding all else constant — any higher direct or indirect transfers to families may increase the average birth rate but reduce female labor supply; individual taxation may increase female labor force participation but reduce fertility rates of medium- and high-skilled women. Yet it is the provision of additional child care facilities that will generate a joint increase in female employment and fertility. The introduction of individual taxation to finance these additional outlays will extend the increase, and the long-term welfare gains will affect all households.64 It seems that the implementation of a more progressive income tax system based strictly on individual incomes could help fund improvements to the child care system; promote saving, investment, and growth; and restore “fairness to the income distribution in an age of increasing inequality.”65
There are considerable unrecognized gains from a reallocation of female labor supply from the household sector to the market sector. Any increase in aggregate income and the tax base has the potential to encourage an increase in public expenditure without the possibility of incurring budget deficits.66 But instead of directing policy efforts toward the problem of female labor force participation, to reflect a changing trend in social attitudes about the respective roles of men and women, the United States has continuously shirked genuine commitment to gender equality. The principal goal of this article has been to assess the problem of female labor force participation regarding child support and taxation and to determine the further implications of this problem on total fertility rates. My criticism has targeted the theoretical basis of the approach to estimating the costs of children as market consumption costs and the common assumption that household income is a reliable indicator of a household’s standard of living. Based on this analysis, the former offers justification for government investment in child care facilities for children of all ages, and the latter provides reason to support a highly progressive income tax system based strictly on individual incomes. Essentially, these adjustments to family and tax policy have the potential to affect and even reverse the seemingly intractable institutional constraints that encourage female dependency. But it is not just policy that needs to change. Both the theoretical literature and empirical work must adapt, as well. Any long-term solution to the problem of gender equality must involve a more equitable gender bargain and this requires broader social changes.
1 Rosen, “Is It Time to Abandon Joint Filing?” 30 Nat’l Tax J. 429 (1977).
2 Patricia Apps and Ray Rees, “Labour Supply, Household Production and Intra-Family Welfare Distribution,” 60 J. of Pub. Econ. 199-200 (1996) (demonstrating the importance of incorporating household production in models of labor supply to avoid misleading results concerning behavioral responses to economic policy).
3 Patricia Hewitt and Penelope Leach, Social Justice, Children and Families, ch. iv (1993) (referencing evidence that suggests having a child reduces a woman’s average lifetime earnings by two-thirds).
4 Gary Becker, A Treatise on the Family 24 (1981) (arguing that investment differences reinforce biological differences: “investments in children with ‘normal’ orientations reinforce their biology, and they become specialized to the usual sexual division of labor”).
5 Ann Mumford, Tax Policy, Women and the Law 3-4 (2010).
6 Id.; Joseph A. Schumpeter, “English Economists and the State-Managed Economy,” 57 J. Pol. Econ. 371 (1949).
7 Apps and Rees, Public Economics and the Household 1-19 (2009) (introducing the flawed economic model that underlies most of public economics and labor economics and outlining the purpose of the book: “to bring together the economics of multi-person households and those areas of labor supply and public economics for which this generalisation of the standard model seems to us to be most relevant and important”).
8 Laurens Cherchye, Bram De Rock, and Frederic Vermeulen, “Opening the Black Box of Intra-Household Decision-Making: Theory and Non-Parametric Empirical Tests of General Collective Consumption Models,” IZA Discussion Paper Series No. 1603, at 2-25 (2005) (expanding on the methodological argument that the standard unitary model fails to provide an adequate description of “observed multi-person household behavior”).
9 Apps, “Closing the Gender Gap in Labour Supply,” in CEDA (research report) in Australia’s Future Workforce? 168-178, section 3.2 (2015) (providing evidence from time use and household expenditure survey data).
10 Sören Blomquist and Vidar Christiansen, “Public Provision of Private Goods as a Redistributive Device in an Optimum Income Tax Model,” 97 The Scandinavian J. Econ. 547-567 (1995) (discussing attempts to explain why private goods are publicly provided — most notably, a study of the public provision of day care by Theodore Bergstrom and Blomquist).
11 Yuri Andrienko, Apps, and Rees, “Gender Bias in Tax Systems Based on Household Income,” IZA Discussion Paper Series No. 8676, at 2-13 (2014) (referring to Australia and the United Kingdom).
12 Hans Fehr and Daniela Ujhelyiova, “Fertility, Female Labour Supply, and Family Policy,” 14 German Econ. Rev. 138-165 (2012) (discussing the potential for individual taxation in Germany to increase female employment, as well as insisting that it is possible to increase birth rates and female employment rates simultaneously if the government increases investment in child care facilities).
13 Apps and Rees, “Fertility, Taxation and Family Policy,” 106 The Scandinavian J. Econ. 745 (2004) (looking at how a combination of individual taxation and family support through child care facilities supports both increased fertility and increased female labor supply).
14 Joseph Stiglitz, The Economics of the Public Sector 694-696 (2000).
15 Apps and Rees, “Household Production, Full Consumption and the Costs of Children,” 8 Labour Econ. 621-622 (2002).
16 Stiglitz, supra note 14, at 695.
17 Laura Tach and Sarah Halpern-Meekin, “Tax Code Knowledge and Behavioral Responses Among EITC Recipients: Policy Insights From Qualitative Data,” 33 J. Pol’y Analysis and Mgmt.413 (2014).
18 Marianne Bitler and Hilary W. Hoynes, “The State of the Safety Net in the Post-Welfare Reform Era,” National Bureau of Economic Research Working Paper No. 16504 (2010).
19 Email from Hilary Hoynes to Zoe Nutter (Apr. 5, 2017).
20 Apps and Rees, “Australian Family Tax Reform and the Targeting Fallacy,” 43 The Australian Econ. Rev. 172 (2010).
21 Id. at 154.
22 Heather Joshi, “The Opportunity Costs of Childbearing: More Than Mothers’ Business,” 11 J. Population Econ. 162 (1998) (considering the “entwined themes of men’s interests in parenthood, the sex division of labour and its evolution, policy for gender equity, and policy to support the level of social reproduction”).
23 Björn Gustafsson and Urban Kjulin, “Time Use in Child Care and Housework and the Total Cost of Children,” 7 J. Population Econ. 287 (1994) (investigating the use of time for child care and housework among Swedish families).
24 Apps and Rees, supra note 15, at 622.
26 Apps and Rees, supra note 7, at 96-101.
27 Id. at 99; Apps and Rees, “Taxation and the Household,” 35 J. Pub. Econ. 355 (1988).
28 Id. at 355-356.
29 Apps and Rees, supra note 15, at 623.
30 Apps and Rees, supra note 2, at 199-219 (outlining the importance, in conceptual and empirical terms, of including household production in models of labor supply).
31 Apps and Rees, supra note 7, at 99.
33 Colombino, “The Cost of Children When Children Are a Choice,” 14 Labour 79 (2000).
34 Id. at 91-92.
35 Gustafsson and Kjulin, supra note 23, at 295.
36 Id. at 287-306.
37 Leach, Children First! What Our Society Must Do and Is Not Doing for Children Today (1994) (arguing that this trend has been “a matter of individualist laissez-faire principles”); Heidi Gottfried, “In Defense of Socialized Child Care: A Comparison of Child-Care Policy in the United States and Sweden,” 1 NWSA J. 336 (1988-1989) (providing a historical overview of U.S. and Swedish national child care policy initiatives, comparing the experiences in the United States with the most progressive child care policy of any capitalist country).
38 Ruth Fincher, “The State and Childcare: An International Review From a Geographical Perspective,” in Kim England, Who Will Mind the Baby? Geographies of Childcare and Working Mothers 143-169 (1996).
39 Apps and Rees, supra note 7, at 10.
40 Apps, supra note 9, at 168-178 (providing evidence from time use and household expenditure survey data).
41 Commission of the European Communities, “Green Paper on European Social Policy: Options for the Union” (1993); Apps and Rees, supra note 7, at 11.
42 Thévenon, “Drivers of Female Labour Force Participation in the OECD,” OECD Social, Employment and Migration Working Paper No. 145, at 14-15 (2013).
43 Mara A. Yerkes, “Collective Protection for New Social Risks: Childcare and the Dutch Welfare State,” 43 J. Social Pol’y 811 (2014).
44 Jon Kvist et al., “Changing Social Equality and the Nordic Welfare Model,” in Changing Social Equality: The Nordic Welfare Model in the 21st Century 1-22 (2012).
45 Apps and Rees, supra note 7, at 3.
46 Apps and Rees, “Family Labour Supply, Taxation and Saving in an Imperfect Capital Market,” 8 Rev. of Econ. Household 297 (2010) (outlining the importance of this nuanced approach; the overall thesis that they develop has significant implications for the composition of the tax system and public policy more broadly).
47 Apps and Rees, “Optimal Taxation, Child Care and Models of the Household,” CESifo Working Paper No. 4578, at 2-35 (2014) (demonstrating that “a richer and more realistic specification of household time use widens the set of cases in which individual taxation is welfare-superior”).
48 Apps, A Theory of Inequality and Taxation 13-15 (1981) (discussing the effects of income taxation under institutional inequality).
49 Id. at 13.
50 Id. at 14.
51 Andrienko, Apps, and Rees, supra note 11, at 2-13 (referring to Australia and the United Kingdom).
52 James Alm and Mikhail Melnik, “Taxing the Family in the Individual Income Tax,” 5 Pub. Fin. and Mgmt. 56 (2005).
53 LaLumia, “The Effects of Joint Taxation of Married Couples on Labor Supply and Non-Wage Income,” 92 J. Pub. Econ. 1698 (2008).
54 Id. at 1700 (referencing a series of works published in 1999 by Apps and Rees, Gottfried and Richter, and Piggott and Whalley).
55 De Villota, “A Gender Perspective Approach Regarding the Impact of Income Tax on Wage-Earning Women in Spain,” in Challenging Gender Equality in Tax Policy Making 109-133 (2011).
56 Id. at 126 (superimposing the PIT taxation scale on a density function diagram to ascertain a complete picture of the effect of rates on men and women).
57 Apps, supra note 9, at 172.
58 Apps and Rees, supra note 13, at 746 (referring to Table 1, sourced from OECD Labour Force Statistics that track the participation and employment of women between the ages of 25 to 44; the total fertility rate indicates the number of children a woman would expect to have if she were to experience all of the age-specific birth rates occurring in that year).
59 Luci-Greulich and Thévenon, “The Impact of Family Policies on Fertility Trends in Developed Countries,” 29 European J. Population 387 (2013); Fehr and Ujhelyiova, supra note 12, at 138-165; Apps and Rees, supra note 13, at 745-763.
60 Robert Fenge and Lisa Stadler, “Three Family Policies to Reconcile Fertility and Labor Supply,” CESifo Working Paper No. 4922, at 2-35 (2014).
62 Apps, supra note 9, at 168.
63 Fehr and Ujhelyiova, supra note 12, at 138-165.
64 Id. at 161-162.
65 Apps, supra note 9, at 177.
66 Id. at 176-177.