Where’s the Sex in Fiscal Sociology? Taxation and Gender in Comparative Perspective

1. Introduction: Beyond War

Getting “sex” into the title of a chapter in book on a subject as daunting as “fiscal sociology” makes obvious marketing sense. Yet in fact something important is missing from the other chapters, each valuable and interesting in its own right.

The field of “fiscal sociology,” born in Schumpeter’s stirring invocation of the “thunder of history” (Schumpeter 1991 [1918]), has dwelt, as many of the other chapters show, mainly with how government tax (or “extraction,” in Charles Tilly’s preferred word choice (Tilly, this volume)) schemes interact with large-scale issues of war, crisis, state construction and destruction, and the like. These are important subjects, situated at the macro-level of society. But a closer and more detailed look at tax systems show that they have deep, persistent consequences on the micro-level, as well. Fiscal policies affect patterns of marriage, child-bearing, work, savings, education, charity, home ownership, and more. “Fiscal sociology” is thus an essential element of “cultural sociology.”

Social norms and biases are reflected in fiscal—tax and transfer—systems, and such systems in turn exercise coercive force, tending to entrench patterns of social life, in an endless feedback loop, a point also emphasized in Beverly Moran’s chapter. This is rich and important subject matter for multi-disciplinary scholars to explore.

The taxation of households illustrates the point. Tax codes and other aspects of fiscal systems have often been explicitly sexist (Stotsky 1996; Teck 2004). In Britain, for instance, income taxation was instituted in 1799, with all income attributed to the husband, reflecting the laws of the time whereby all of the wife’s property became the husband’s at marriage. The Married Women’s Property Act in 1882 allowed women for the first time to retain management and control of their separate property and earnings. (For an excellent analysis of the movement for women’s property rights in the United States, see Siegel 1995)). After a delay, the tax code was amended in 1894 to allow the wife’s earnings the same tax relief as a single person if the couple’s combined income was less than the then significant sum of 500 pounds; above that threshold, an “aggregation” model persisted, whereby the wife’s income was added atop the husband’s in his taxable income. In 1918, a ”married man’s allowance” was introduced and was increased to 1.6 times a single man’s allowance in 1982 (the 1.6 multiplier clearly deriving from U.S. law, as we shall see below); the extra deduction amount giving a man a tax savings for having a wife, whether she had market earnings or not. Well into the 1970’s the Inland Revenue, the United Kingdom’s tax agency, refused to correspond or send tax withholding refunds to married women, dealing only with their husbands. The law had to change to compel bureaucratic sensitivity to modern norms: In 1978, the Finance Act required the Inland Revenue to send married women their refunds. Still, as a matter of substance, households with wives earning above a minimal amount of market earnings faced a marriage tax created by the aggregation model of joint filing, discussed below, which was eliminated by legislation in 1988, effective in 1990, requiring individual returns for earnings and investment income. The married man’s allowance became an allowance that could be transferred between spouses in 1993. (Briggs 1985; Stotsky 1996). Just about the entire history of this household tax policy was written in explicitly sexist language, of “men” and “women,” “wives” and “husbands,” and the English system had clear effects throughout the British Empire, such as in Canada (Kesselman 2007), New Zealand (Jones 2005), Australia (Cass and Brennan 2003), and Malaysia and Singapore (Teck 2004). Now it is one thing to note, in honest language, the inevitably sexist and gendered effects of tax laws; it is quite another thing to make as a matter of positive law a rule whereby taxing authorities will not communicate with married women.

Sexist effects run far deeper than expressive and dignitarian harms, however severe these may be. In the United States, to take an important example, tax systems, mainly created in the 20th Century—and hence generally free of the openly sexist language found in England and many Commonwealth societies— are not facially gendered. Yet major features of the American tax system were put in place in the 1930s, 40s, and 50s, periods when the one-earner family with a stay-at-home wife was both the descriptive fact and the normative ideal for most Americans. These structural elements persist, unmitigated and indeed often increased in their magnitudes, well into the 21st Century.

Fiscal policies in the U.S. now make it difficult to be a household featuring two wage-earners, married with children—even though this has long been the dominant model among married couples with children in the U.S.: In 2005, 65.1% of all married couples with children under the age of 18 fit it (U.S. Census Bureau 2008 at Table No. 582). It is important to understand that the burdens on two- earner households are analytically distinct from the “marriage penalties” that have dominated political discussion and rhetoric (Zelenak 2007 at 1140); indeed, as marriage penalties for the middle and upper classes have been lessened, the analytically distinct secondary-erarner bias has increased (Richards 2008). In any event, the deep bias against two-earner households has different effects along class lines, and affects patterns of the sexual division of labor and hierarchy throughout the wider society:

At lower-income levels, marriage becomes the choice variable, as working for pay is not: Tax and other fiscal laws make it hard to be a two-earner married family, hence children are raised by unmarried, single parents. One-third of American children live in households without two parents, most with their mother, and most of these households are poor (U.S. Census Bureau 2008 at Tables No. 65 and 676). Women in these settings are trapped by economic needs, social stigma, and the relentless constraints of too little time. And yet, ironically, marriage penalties among the poor persist in full force (Richards 2008, McCaffery 2003).

At the upper-income levels, in contrast, the same bias against two- earner households leads to the choice variable being whether the “second” earner works outside the home, for pay, or not: a bias against two-earner households pushes towards “traditional,” single-earner households with a stay-at-home spouse, typically the wife, making this model predominate among the economic elite. See for example (Tahmincioglu 2006).

In the vast middle-income classes, the bias pushes towards stress and instability, as most married couples are forced to swim upstream, as it were, living out a life of work-family “balance” in a socio-economic structure designed to discourage that very life. (Crittenden, 2001; McCaffery, 1999a). Here the issue is not marriage penalties—which ironically have been reduced—but relief for child-care, which has not been increased.

Fiscal policies are both the causes of and the absence of cures for the attendant social stresses and ills. These effects come from choices, consciously and subconsciously made, over nearly a century of comprehensive tax policy. And America’s choices meanwhile have had echoes and resonances throughout the developed and even the developing world, just as the British answers have had throughout their empire (and see the Brownlee chapter in this volume for another example of American influence—for the worse?—in matters of tax policy).

The American and British stories are just two among many. While the U.S. tax system provides a strong orientation for a “traditional,” two-parent, single earner family—a model that has had perverse and perhaps unintended consequences on those households who do not, or cannot, fit it—other countries have addressed the issues differently. After a period of time, in the wake of World War II, in which many developed countries around the world more or less blindly mimicked the American approach (see again the Brownlee chapter, on the Shoup Mission, for an interesting parallel), the vast majority of developed and many developing countries are now following a “world-wide trend” towards individual, rather than joint, filing (Pechman and Englehardt 1991; see also OECD 2006 at 54-56). A prime mover for the reforms seems to have been the desire to encourage greater female labor force participation: In 1984, the European Community (EC) examined the impact of EC tax systems on women’s participation in the workforce, a major concern being the high marginal rate secondary earners face under joint filing systems, such as the current U.S. one, and concluded that a disincentive for women to enter the labor force resulted:

Indeed, the continuing increase in the number of married women entering the job market, and the consequent rise in the number of two earner families, coupled with the continued application of tax systems which benefit the traditional family, produces the anomalous situation in some Member States that a growing number of couples will be financially better off remaining unmarried, and this particularly where there are children, owing to benefits granted for one parent families. (European Communities 1985).

The EC analysis takes working women as a social fact, and hence concludes that a bias against two-earner families will lead to non-marriage; the United States in its policy, in contrast, has never so acknowledged the fact of working wives, and is clearly ambivalent about it. In any event, the United Kingdom, France, and the Netherlands, among others, reformed their tax systems after the EC Memorandum to mitigate the effects. (Stotsky 1997). South Africa did so as well in 1995. (Smith). Today, most developed nations have separate or individual filing, although the presence of other features in the tax systems, such as deductions allowed to the earner in a one-earner household, continue to make it difficult to discern just what the net effects in any one country are. (Pechman and Englehardt 1991; OECD 2005, Ryrstedt 2006).

But yet again, there is great variation. The Czech Republic, for example, bucked the world-wide trend by opting for joint taxation in 2005, apparently using an optional separate filing option (OECD 2006 at 54). Some societies, like

Germany, following a model developed in West Germany and in marked contrast to the incentives prevailing in East Germany prior to unification, continue to use tax and other fiscal policies to encourage stay-at-home mothers (Duggan 2003); Germany is often singled out as an example of tax and fiscal policies using a “male breadwinner model.” (Palme 2005). France uses a complex system of family taxation and generous state-sponsored child-care programs, both to encourage large families and to facilitate women staying in the paid workforce after giving birth (Henneck 2003; see also Pechman and Englehardt 1991). The only parallel to the family’s being the appropriate taxable unit seems to be the “Hindu joint family,” (HJF) as used in Malaysia and Singapore, where the HJF files and pays taxes as a consolidated unit, provided that the “karta,” or head of the family, typically the eldest male, is resident in the nation. (Teck 2004). Japan has a system of individual filing that features an “Allowance for Spouses” (AS) and “Special Allowance for Spouses” (SAS) that appears to have been designed with the express intent of encouraging married women to work outside the home—but only on a part-time or limited basis. (Akabayashi 2006). Scandanavian countries boast of employing a “dual earner model” that “supports female labour force participation and to a large extent also male participation in care of children.” (Palme 2005).

Back to America, much of today’s law and effects harkens back to a policy implemented in 1948, when a major piece of the “peace dividend” attendant on the end of World War II was spent on a tax law change designed to get women back into the homes that they had left during the war effort. (McCaffery 1999a). And so here is another relevance of the macro-scale issue of war to fiscal sociology: the perceived need for the central government to accommodate and even actively shape the demographics of the workforce, during and post wartime. Nations need women workers, to fuel war efforts or growth, or they do not want women workers to compete with men during periods of high unemployment (see, e.g., discussion in McCaffery 1999a at 76-78); depending on the need, nations change their tax and other fiscal policies. So, too, countries seem to change tax policies in response to natalist concerns, about too many or too few children. And so on: In each case, fiscal tools—government tax and transfer policies—have strong effects, often surprisingly actively argued for and supported, on the character of family life, and the division of sexual roles and labor throughout the society.

In all of this, some common and interconnected themes emerge, three of which I comment on in this chapter.

One, causes and effects are often hard to see in the dizzying complexity of tax and fiscal policy: there is a “fog of tax,” akin to the “fog of war” (Morris 2004), making it hard even to understand what is going on. It can be difficult, for example, to compare general fiscal political regimes across countries because an effect evident in one system, such as the encouragement of married woman’s paid labor force participation that comes about under a system of “separate filing” under a comprehensive income tax, can be undercut by some incentive effect generated by another system, such as the absence of subsidized child-care options, or special tax and other fiscal benefits for traditional one-earner couples. (European Communities 1985; Pechman and Englehardt 1991; OECD 2005).

Two, in the complexity and haze, there is much room for rhetorical manipulation and even cognitive error. Terms like “penalties” and “bonuses” are often used and abused; competing conceptions of “neutrality” or “horizontal equity” vie with one another; abstractions such as “imputed income” lose out to more tangible realities. Underlying this phenomenon is a simple analytic fact that has dramatic cognitive consequences: What is a subsidy or a bonus viewed from one perspective or baseline is a tax or a penalty viewed from another. (Thaler 1980; Schelling 1981; McCaffery and Baron 2004). Child credits are childless surcharges; marriage bonuses are singles penalties; subsidies for stay-at-home spouses are taxes on working ones. Given the many degrees of freedom to characterize equivalent policies, cognitive confusion bred by systematic complexity, and a status quo bias in any event, change is in general hard. Thus, tax and other fiscal policies that shape and constrain familial arrangements tend to persist, as in the British and American contexts, well after the underlying norms of family have changed. (McCaffery 1999a; Ventry 2007).

Three, when change does come, it more often than not favors the elite— as shown by the United States (McCaffery 1999a; Ventry 2007), Canadian (Kesselman 2007) and New Zealand (Jones 2005) case studies we shall consider below—and/or is predicated on macro-economic concerns, such as the need for more (or less) labor-force participation (European Communities 1985), or more or fewer children (Henneck, 2003). Absent from the domain of fiscal politics, by and large, is a thicker substantive conception of rights or fundamental fairness, especially one looking to the dynamic effects of micro-level decisions about work and family on systemic patterns of discrimination and entrenchment. (McCaffery

1993b; McCaffery 1996). By dynamic, I mean a concern with the effects of static rules, such as tax-law structures, over time: in how they affect decisions such as human capital formation, on the “supply” side, and hiring and promotion decisions, on the “demand” side, that endure through time. And even where such a concern does seem to be present, as in Scandanavia (Palme 2005), and, perhaps, France (Stotsky 1997), changes in the direction of helping emergent family models do not necessarily obtain that goal, because of the complexity of the whole, and the tendency of competing political groups and ideas to get rewarded elsewhere in the system (European Communities 1985; Pechman and Englehardt 1991). Further, a concern with fiscal consequences, and/or social norms that suggest that working mothers are more appropriate or more needed among the poor has led to many child and family oriented policies being means- tested, but then the withdrawal of these benefits as households attempt to emerge from the lower-income realms works like a tax, creating various forms of “poverty” and “family structure” traps. (McCaffery 1999b; Palme 2005). Just as scholars concerned with “fiscal sociology” can fail to see the sexist and other socially constructed biases enmeshed in tax and other fiscal systems, so too can those persons and movements concerned with egalitarian and progressive goals fail to look to fiscal policies for their particular gendered and other social constructions (as opposed to their broad and often crude distributive properties), or fail accurately to integrate the divergent and often conflicting elements of a single system into a coherent whole.

These various themes lead to a strong conclusion that more detailed work, on a country-by-country basis, is needed, simply to ascertain what is going on and why, and to a more tentative conclusion that the game may no longer be worth the candle: that there is good reason to be skeptical of complex tax and fiscal systems consciously or unconsciously aimed at “social engineering,” even if we accept the inevitability of some non-neutral effects from any set of rules.

This chapter explores these themes with an historical and comparative focus. Like the other chapters in this volume, the topics raised here are large ones, and the sweep, in both time and space, is wide, so I can do little more than note the prevalence and interest of the themes. The limits of time and space are compounded by the simple fact that the analytics of the sexual aspects of tax policy are highly complex—this is one of the themes, after all. Any meaningful exploration of tax-law biases in any historical or comparative context should proceed from a sound analytic understanding of the facts of the matter. In the next Section, therefore, I lay out some of the basic terms and concepts, illustrating both the analytic points and the broader themes with comparative and historical observations. Section 3 then circles back to discuss briefly how the analytics and certain case studies illustrate the three themes noted above.

2. The Analytics of Taxing Households

Comprehensive tax systems as found in the United States and other developed countries—and increasingly mimicked in developing countries, for better or, largely, worse (Bird and Zolt 2005)—are complex. Sexist and other socially contingent effects lie hidden beneath the surface—this is part of the story, for only certain actors go looking for social construction through tax codes, and these actors are often rewarded for the effort. It is important to define terms clearly and understand the analytics well. This section gives an abbreviated but necessary guide to key concepts and vocabulary.

a. The Filing Unit, Marriage Penalties and Bonuses

We begin with some basics. But as we do so, it is important to bear in mind that marriage penalties and bonuses—which easily grab one’s attention— are not the main issue of concern to the sexist division of labor at home and in the market. That comes with the secondary-earner bias, which comes later.

i. Choices of Filing Unit

Any system of comprehensive individuated taxation must answer the question of attribution, or the appropriate filing unit, that is, in short, the question of who pays taxes. In the first instance, this typically has meant deciding what to do about households with two adults living together. In the United States today, this question still comes down to how to tax married couples, for the federal tax law has refused to consider any accommodation for unmarried couples, including same-sex ones generally denied the legal right to marry under federal law. (Ventry 2006). As of 1991, the Netherlands was the only state to allow unmarried couples living together to obtain the same benefits as heterosexual married ones. (Pechman and Englehardt 1991).

One answer, the initial default one in America (more or less), is to ignore marriage, and to treat all individuals like individuals. This is known as separate filing. Most countries around the world, many of which tried the more typically American system of joint filing for a while, have now chosen or reverted to separate filing. (Pechman and Englehardt 1991; Stotsky 1997; OECD 2005).

A second answer, which has been the American one since 1948, is to treat spouses as a unit, which is known as joint filing.

A compromise or hybrid answer is to countenance both answers, a system of optional separate filing, in which couples can choose to file their taxes together or apart. As explained below, this is not the same system as “married, filing separately” under U.S. law. The Czech Republic apparently adopted this solution in 2005. (OECD 2006).

There are also other possible filing units. One of relevance to the story of taxing households is known in the United States as “Head of Household” status (Pechman and Englehardt 1991), which refers to a single parent or other adult raising children, a common feature across the globe. But the main options I consider below are separate filing, joint filing, and optional separate filing.

ii. Why Does it Matter?

Why does any of this matter? Because most developed nations’ principal tax systems depend on a pattern of progressive marginal rates. This mechanism, commonly misunderstood, works like a step function or a metaphoric ladder. One pays different rates of tax on different levels of income, moving into a higher rate bracket, on the margin, as one’s income rises. This new rate does not apply to all of one’s income, just the amount that is in the new bracket.

iii. Two Neutralities and One Impossibility Theorem

Under a system of progressive marginal rates, two seemingly attractive ideals, that of marriage neutrality and couples neutrality, come into tension, posing a certain impossibility theorem. Marriage neutrality holds that a couple’s taxes should not be affected by marriage. Couples neutrality holds that “equal-earning couples” should bear equal total tax burdens. The impossibility theorem is that, in a system with progressive marginal rates, there cannot be both types of neutrality at once.

iv. A Numerical Example

This all sounds complicated, but fortunately the main ideas can be illustrated with a simple example, using a simplified rate schedule.

Suppose that the income tax had just two rate brackets—as proponents of a “flat” tax in fact desire. There is an initial “0 bracket,” followed by a 20% one. We begin with a rate structure for all individuals, as would exist in a world of universal separate filing:

Income Marginal Rate
$0 to $20,000 0%
Above $20,000 20%

Table 1: Individual Rate Schedule

Now imagine two couples, the more traditional Ozzie, who earns $40,000, and

Harriet, who earns $0, and the more “modern,” egalitarian Harry, who earns $20,000, and Sally, who also earns $20,000. Under the individualized, separate filing system, Ozzie pays $4,000 in taxes (0 on his first $20,000 plus 20%, or $4,000, on his next $20,000); Harriet pays $0, and Harry and Sally each also pay

$0, because all of their income falls within the 0 bracket. As the rate structure applies to both unmarried individuals and married individuals as individuals (“filing separately,” that is), there is marriage neutrality: The Ozzie and Harriet couple pay $4,000 in taxes between them, whether they are married or not, and the Harry and Sally couple pay $0 in taxes, whether they are married or not. Marriage is irrelevant. But this example violates couples neutrality, because the two couples, each earning $40,000, combined, pay different total taxes. In the United States, however, this resolution came to be seen as unacceptable. In essence, “couples neutrality” has been held to be a higher value than “marriage neutrality.”

After the War, the U.S. revisited the issue, and now moved to a system of joint filing with full income splitting, at double the singles’ rates, as discussed below. This plan not only rendered moot the incentives for spouses to income split, or pretend to income split, on their own, but it also benefited all married households, especially high earning one-earner ones. It was thus widely approved. (McCaffery 1999a at 51-54). Ever since this fateful moment in 1948, U.S. law has made provision for joint filing, which means that all married couples are taxed as a unit, and thus, by design, all equal-earning married couples bear the same aggregate tax. And, even though only the Czech Republic, France, Germany, Ireland, Luxembourg, Poland, Portugal, Switzerland and the U.S have some form of joint filing today (OECD 2006 at 54-56), the norm of “couples neutrality” still seems to haunt world-wide analysis of tax systems, prompting calls for reform or tax and fiscal breaks for one-earner families who, like Ozzie and Harriet in our running example, pay higher taxes than two-earner ones under separate filing. (European Communities 1985; Pechman and Englehardt 1991; OECD 2005; Kesselman 2007).

v. Moving to Joint Filing, and Adding in Rates

Once a system has gone to some form of joint filing, the question becomes what to do with the rate schedule. Suppose, in the first instance, that the law maintained exactly the same rate structure for married couples as for individuals. This is often called aggregation in the literature (European Communities 1985; Jones 2005; Kesselman 2007) because the two spouses’ incomes are simply aggregated together. This was the British way, for most of their taxing history, as noted above. New Zealand had this system in place until 1960 (Jones 2005), and it is what was proposed—and widely opposed—in the United States in the midst of World War II, as discussed above.

What would happen to our two hypothetical couples under aggregation? Table 1 would still apply, now to married couples as well as to individuals. Ozzie and Harriet, if unmarried, would still pay $4,000 in combined taxes; Harry and Sally would still pay nothing. But now let us marry the two couples. Ozzie and Harriet stay at $4,000 in taxes, but Harry and Sally now must pay this, too—as a couple, their $40,000 bears no tax on the first $20,000, and a 20% tax on the “second” $20,000. Hence the marriage penalty: the taxes of the equal-earning couple increase on marriage. Note that any couple with just about any kind of split in their incomes—even those couples who could split their capital income only, or those who used legal manipulations to appear to split their labor income, or yet again those who could rely on state community property law to do the splitting for them—would see their taxes increased under joint filing with aggregation. This is why the system was so vociferously opposed—by wealthy traditional households—in the U.S. in 1941. But the aggregation answer does have couples neutrality, because the two married couples, each earning a combined $40,000, pay the same total tax.

Suppose, finally (for now) that the law doubles the rate brackets for married persons, allowing, that is, a couple to have a $40,000 zero bracket in the running example. This is often called income splitting in the literature, because the effect of doubling the rate brackets is to treat each spouse as if she or he had earned one-half of the total (combined) spousal incomes—the law does the splitting for couples, in effect, so that they do not have to, or pretend to, do so themselves. The rate schedule would now have different rates for unmarried individuals and married couples, as follows:

Unmarried Individuals Married Couples
Income Marginal Tax Rate Income Marginal Tax Rate
$0 – $20,000 0% $0 – $40,000 0%
Above $20,000 20% Above $40,000 20%

Table 2: Rate Schedule with Individual and Joint Filing

If our two couples are unmarried, the analysis continues as above: Ozzie pays $4,000, none of the other individuals pays any tax, using the left-hand rate schedules. But if and when the couples get married, both couples see their taxes go down to $0; both couples, as couples, now have a $40,000 “0 bracket,” using the right-hand side. Hence arose the marriage bonus, which comes only to the more traditional, single-earner Ozzie and Harriet household, because their aggregate tax burden falls on marriage; recall that Harry and Sally paid no taxes when they were unmarried. The effects of “income splitting” are evident here, because the combined, doubled-up rate schedule for married persons has the same effect as treating each spouse—Harriet as well as Ozzie—as if she or he were a single person earning half the combined spousal income; viewed from this perspective, Harriet gets a 0 bracket, too, just like Sally has always had by virtue of her own work. Note, too, that we once again have couples neutrality, because all married couples pay the same tax, using the right-hand rate schedule in Table 2.

vi. The Math of it All, and a Third Way

All of the math of marriage penalties and bonuses can be understood from this simple example, although the magnitudes get higher as rates and rate brackets increase, and also as we consider other factors, such as tax credits and non-tax benefit programs, that also have marriage penalties and bonuses within them (McCaffery 1999b; OECD 2005; Palme 2005). The question, under joint filing, is by how much, if at all, the law multiplies the individual rate brackets to accommodate married couples. In the first example, featuring aggregation, we had no accommodation, which is the same thing as multiplying the singles rate brackets by 1. Such an answer yields only penalties and has no bonuses. In the second example, featuring full income splitting, we doubled the rate brackets, that is multiplied them by 2. Such an answer yields only bonuses, no penalties.

This mathematical view of the matter helps to explain additional options, such as those that have prevailed in the U.S. starting in 1969. The problem at that time was considered the “singles penalty.” (Groves 1963; McCaffery 1999a at 59-67). A singles penalty is simply the absence of a marriage bonus. Consider that, under our running example and the doubled-up rate brackets put in place in 1948, if Ozzie did not marry Harriet, he would be paying $4,000 in taxes; if he got married, this tax would go away. It is simply an analytic fact that “bonuses” can be converted into “penalties” by semantic manipulation of the baseline: One could fairly say that there was a “penalty” on not being married—the absence of a bonus can always be restated as a penalty. (Thaler 1980; Schelling 1981; McCaffery and Baron 2004). In 1969, the singles penalty came to be seen as especially unfair, and so the American rate structure changed, by multiplying the individual rate schedule by 1.6—a value between 1 and 2 (and the same value used for the expansion of the British “married man’s allowance” in 1982, as discussed above)—for married couples. This created both marriage penalties and marriage bonuses in the same law.

Consider Table 3, with a rate schedule such as has obtained in the U.S. since 1969:

Unmarried Individuals Married Couples
Income Marginal Tax Rate Income Marginal Tax Rate
$0 – $20,000 0% $0-$32,000 0%
Above $20,000 20% Above $32,000 20%

Table 3: Rate Schedule with Joint Filing but not Full Income Splitting

Notice that now the right-hand rates, for married couples filing jointly, have a 0 bracket of just $32,000, less than double the singles 0 bracket of $20,000: the amount is 1.6x the singles bracket, or (equivalently) 80% of double it.

Consider, for one final time, Ozzie and Harriet, and Harry and Sally. Each couple still earns, as a couple, $40,000, but now they must pay a tax: of $1600, or 20% of the amount in excess of $32,000. Ozzie and Harriet still get a marriage bonus, as their taxes are reduced from $4,000 pre-marriage, but this bonus has been reduced, from the full $4000 to $2400. Harry and Sally, in contrast, now see their taxes increase on marriage, in a marriage penalty, from $0 to $1600. Couples neutrality is preserved. But there are both penalties and bonuses on marriage.

It is important to note who pays penalties and who gets bonuses: the more equal-earning a couple is, the higher their marriage penalty, the more one earner earns as a percentage of the household’s total earnings, the higher their marriage bonus. Hence Harry and Sally, at a 50-50 split, see a maximal marriage penalty, and Ozzie and Harriet, at a 100-0 split, see a maximal bonus. It has been consistently true in America that one-earner families, a distinct minority of households, are the big winners in the story.

As a final wrinkle, since this change in 1969, American married couples have had the option of being married, but filing separately. But this is not, however, the same thing as “optional separate filing,” as apparently the Czech Republic has now adopted. Consider that there are two effects evident in Tables 2 and 3, with their joint filing. One is the attributional matter of joint filing. The other is the rate schedule. In Table 2, the rate structure is set at a doubling of the individual rate schedules, making for full income splitting. This cannot raise any couple’s taxes, and will lower them for all but those couples making the same income. In contrast, in Table 3 the rate schedules are set at double 80% of the singles rate schedules (1.6x or 160%); both bonuses and penalties result, depending on the pretax split of income between the spouses. Under the “married filing separately” rate schedules, one pays tax under one-half this amount, in other words, 80% of the true singles’ bracket, for a 0 bracket of $16,000, not $20,000, in our running example, as illustrated in Table 4:

Unmarried Individuals Married Filing Jointly Married Filing Separately
Income Marginal

Tax Rate

Income Marginal

Tax Rate

Income Marginal

Tax Rate

$0 –


0% $0-$32,000 0% $0-$16,000 0%


20% Above


20% Above



Table 4: Prior Rate Schedule with Married, Filing Separately Status

If Sally and Harry are married and decide to file separately, they continue to pay $1600, as a couple: each pays $800, or 20% of the $4000 between their incomes, $20,000, and the 0 bracket of $16,000. Since their incomes are already split, they Ozzie and Harriet, in contrast, would pay higher taxes if they were married and filed separately; Harriet would pay 0, as per usual for her, but Ozzie would pay $4,800, or 20% of his $24,000 of income over the 0 bracket: a significant increase over the $1600 Ozzie and Harriet pay filing jointly. Income splitting by means of joint filing is a large benefit for traditional couples, like Ozzie and Harriet, who have not split their economic incomes on their own.

A system of truly optional separate filing, in contrast, allows married couples to file as if they were unmarried, in which case couples who suffer marriage penalties, like Harry and Sally, do not need joint filing, and the decision to file separately is neutral for them would file separately, using the left- hand rate schedule, and those who got marriage bonuses, like Ozzie and Harriet, would file jointly, using the middle schedule. America has soundly rejected the option of optional separate filing. Simply to illustrate the confusion in all this— and to make you, dear readers, feel better about working it all through—note that the OECD, in a lengthy published report on “fundamental reform of personal income taxes,” showed its misunderstanding, noting without explication that the U.S. gives couples the option to file jointly or separately. (OECD 2006 at 55). But it is not much of an option in the United States, and it certainly does not remove marriage penalties or the secondary-earner bias.

vii. A Look at the Norms

Table 5 summarize the five different filing unit-rate schedule systems we have considered: separate filing; joint filing with the singles rate schedule, 1x (aggregation); joint filing with double the singles rate schedule, 2x (full income splitting); the compromised American-style system with joint filing and 1.6x the singles rate bracket; and optional separate filing. It indicates whether each system meets the norm of “couples neutrality” and “marriage neutrality,” and whether it has marriage bonuses, penalties, or both.

Norm/effect Filing Unit/Rate Schedule Structure


Joint Filing Optional Separate Filing
At 1x Singles Schedule (aggregation) At 1.6x Singles Schedule (U.S. post


At 2x

Singles Schedule (full income splitting)










Table 5: Filing Unit, Rate Schedule Structures and Norms and Effects

I trust it is now more than apparent that this is all a tangled web, indeed. The United States, ever since 1948, has only had options that feature couples neutralities and marriage bonuses; further, ever since 1986, no tax law change has benefited two-earner couples without also benefiting one-earner ones. (McCaffery 1999a; Ventry 2007). Most of the rest of the developed world, in contrast, has chosen or reverted to separate filing, with its marriage neutrality. (Stotsky 1997; OECD 2005). Yet because of the perceived unfairness of more heavily taxing one-earner couples with the same combined income than two- earner couples—because of the violation of couples neutrality, that is—most countries that have separate filing give some additional tax benefits to one-earner couples. (Pechman and Englehardt 1991; OECD 2005; Kesselman 2007).

The political and psychological complexity of this all is only about to increase, however, as we move on to consider that these very norms of couples and marriage neutrality are not the only effects at work, and are almost certainly not the most important ones. I have put them first in the analysis because this is where they have tended to be in the political and intellectual historical development of tax policy. But now it is time to enrich and complicate the analysis. First let us note that the very measurement of these norms, as reflected in Table 5 and political and economic discourse world-wide, misses a very important point.

viii. A (Not Just) Semantic Note

Before leaving the narrow but important subject of taxable units and rate schedules, I want to draw attention to one semantic matter that is far more than semantic. We have seen that in the United States, the ideal of “couples neutrality” has won out over that of “marriage neutrality”—there is strong norm of taxing “equal-earning couples equally.” This norm has had influence world- wide, even in countries that reject the American style “male breadwinner model.” Canada, for example, is considering a reform even as I write to allow greater income splitting for married couples. (Kesselman 2007). Canada has long had a system of separate filing, although its nuances, like most countries’ tax systems, are complex.

Yet there is a sleight of hand in this analysis, which only the most sensitive readers might have caught: it lies in the assumption that Harriet in our running example is not really doing anything valuable at all. In other words, how are we even thinking about “earnings” or “income”? Are Ozzie and Harriet, and Harry and Sally, really “equal-earning” couples?

As all married couples with children—and hopefully all people, period—know full well, the “stay-at-home” spouse and parent is providing tremendously valuable services, including child-rearing and home care. (Staudt 1996). Tax experts refer to the market value of such services as imputed income. This reflects the amount of money that one would have to pay to have a third party do what one does for him or herself. The critical point is that the American, as all other comprehensive “income” tax systems, simply ignores imputed income. Tax falls on monetary income, in cash or cash-equivalents, and this is the metric by which the “equal-earning couple norm” was measured. This means that “traditional” one-earner households already get a benefit, which a system with marriage bonuses or other special relief for them only compounds.

Looked at another way, consider how easily manipulated conceptions of “neutrality” are. “Couples neutrality” sounds like an appealing norm and, in fact, in psychological experiments I have conducted with Jonathan Baron, we found that ordinary citizens were very attracted to it. (McCaffery and Baron 2004). Yet the “equality” in “couples neutrality” is one in observed, monetized income, which is easy to see as an arbitrary variable. Suppose instead that we posited a norm of “child-care neutrality.” This would hold that the value of child-care services, whether provided by one’s self or by a paid third-party provider, should be taxed the same. I suspect that many readers would recoil at the first intuition as to how to meet this goal: the idea of putting a value on the services of the stay-at-home parent and care-giver, and taxing her for it. There would of course be significant complexities in administering a scheme that taxed non market house-work. But consider that imputed income is also the absence of the out-of-pocket costs that come when one does not have self-provided capital or labor. The two-earner couple, Harry and Sally, in other words, have child-care costs, which they must pay in cash, that one-earner households like Ozzie and Harriet typically do not have. Thus the norm of “child-care neutrality” could (simply) mean a general deduction for paid child-care from the income tax. Neither America nor any other country has such a provision; indeed, many countries give tax breaks for one-earner households. (Pechman and Englehardt 1991; OECD 2005). (It is true, however, that many countries, including France and the Scandanavain ones, have very generous state-sponsored child-care provision outside of the tax system, which does have the effect, typically intended, of facilitating “dual earner” households. (Duggan 2003; Henneck 2003; Palme 2005). Meantime, working parents must earn twice as much as they need to pay for child care because their distant Uncle Sam (or other sovereign “relative”) must get paid, first, before the nanny.

In a final irony, scholars (Staudt 1996) and many political movements sometimes point out that the services of stay-at-home spouses are not properly valued, which relates to the fact that they are not cashed out in monetary terms. I do not deny that, in many important expressive and dignitarian ways, not putting a cash value on something means not valuing it fully. But in tax systems that tax monetized transactions, this “non-valuing” is a benefit. Nonetheless, the psychological perception that one-earner families are suffering under a tax system with separate filing, perhaps aided by conscious or unconscious social engineering and construction, has led many countries to give benefits to couples with stay-at-home spouses. (OECD 2005).

b. Not Just the Marriage Penalty: Secondary-Earner Bias

Alas, the plot must thicken, as anticipated above. We have just explored, at some length, the questions of the filing unit and rate schedules, which lead to marriage penalties, bonuses, or both, under a system of progressive marginal tax rates. These analytic facts and various rhetorical characterizations of them have played a role in the sexual fiscal politics of many countries world-wide. But the marriage penalty and its cognates, a powerful psycho-political element in the United States and elsewhere, has almost nothing to do with another effect, which exerts a real coercive force on women world-wide: the secondary-earner bias.

Once you have joint filing, the question—an accounting question in the first instance—arises as to whose income to put “first” in the calculation of household income and taxes. Suppose, for example, that Harriet in the running example considers going to work outside the home. Given that Ozzie is already working and most likely not thinking of taking any significant time off, Harriet’s first dollar will be taxed at 20%. And here is a place where the simplification of this chapter matters, because many U.S. spouses will see their first dollar taxed at 50% or more, and the number can approach 100% in some lower-income households, when one considers the costs of the loss of various means-tested benefits, or programs that depend on a putative recipient’s other available means of support. (McCaffery 1993a; Shaviro 1999; McCaffery 1999b). Contrast that with the “first” or “primary” earner, who always enters the workforce in the zero bracket—and for whom this zero bracket gets bigger, in the case of joint filing (or, indeed, in the case of most separate filing systems, historically and world- wide, for primary earners with non wage-earning spouses). Note that this effect does not depend on the precise rate accommodation for married couples, nor does it depend on whether there are marriage penalties, bonuses, or both. The secondary-earner bias comes about simply and strictly because of joint filing. In a system of joint filing with aggregation, there can be only marriage penalties; in one with income splitting, only bonuses; but in both—and indeed in all—cases where husbands and wives are added together by the tax system, the secondary or “marginal” worker will see her labor income taxed at a rate bracket dictated by the “primary” worker’s income. If Harriet goes to work, given that Ozzie is earning $40,000, her first dollar will be taxed as the couple’s 40,001st one. The effects can be dramatic, as we shall see below, and there is plenty of evidence that the discouragement of second workers, in whole or in part, in at least certain countries—the Unites States, Germany, and Japan foremost among them—is and has been quite deliberate.

Of course, the language and rhetoric of “primary” and “secondary” can be offensive, and it is increasingly common for wives to have higher incomes than husbands. But any attempt to avoid a sexist language—putting the woman “second”—can get in the way of understanding and opposing a sexist reality. “Marriage penalties” and “bonuses” have a great deal of salience, and arguments for valuing stay-at-home spouses have clouded the problems of secondary- earner bias. Yet looks abroad and through time indicate that powerful decision- makers have been generally aware of the inducements for second earners within households to enter the paid workforce, or not, and have willingly used fiscal policy as a tool to shape society in their preferred ways.

c. A Note on the Kids: Child versus Child-Care Credits

A key distinction is between tax relief designed to lower the burden of taxation for the presence of children—relief that benefits all households with children, whether or not both parents work outside the home—and provisions for child care, which benefit just the families with needs for paid third party care, typically meaning dual earner households. A large and intensifying theme in the U.S. has been the tendency to benefit children, as by per child credits, and thereby to encourage stay-at-home spouses, and the reluctance to do anything distinct or special for households using paid care, thereby not helping specifically paid market working mothers. (McCaffery 1999a). Where there is child-care relief, it tends to be geared towards lower-income households, introducing an important class dimension into the analysis: a thought that poor wives and mothers ought to work, but rich ones ought not.

Germany has a similar pattern to the American of benefits for one- earner households, child benefits not tied to dual workers, and means-tested programs, (Duggan 2003; Henneck 2003), all earning it a reputation as a strong “male breadwinner model” society. (Palme 2005).

France, in contrast, which has a family-based tax system to reward and encourage large families, also has a series of provisions meant to enable dual- earner households, especially working mothers. For example, France offers generous parental leave options and family allowances. These family allowances are intended to replace wages, and thus reward a baseline salary for both spouses (or at least encourage the wage-earning spouse to take time off and care for his children), rather than serving as payment for motherhood, per se, as in Germany or the U.S.. A larger portion of French women are employed full time than American women, apparently due to the availability of high quality and affordable child care: 25% of 0-2 year olds and 95% of 3-5 year olds are in public child-care (the lower number for younger children related to the generous leave policies), while 78.8% of women in the peak child-bearing years, between 25 and

39, are employed. Still, in recent years in France, means-testing has increased, which may actually and ironically cause some low-income families to decide to have a stay-at-home parent so as not to not forego benefits—once again, the loss of such benefits acting like a tax on the working poor. Additionally, France offers either parent paid parental leave until the child is three years old, although women are virtually the only ones to take such leave. (Henneck 2003). Thus, from a Scandanavian vantage point, even with the generous state subsidized child care and the prevalence of market working mothers, France gets labeled as a “male breadwinner model,” just like Germany. (Palme 2005).

Italy has perhaps the strongest approach to keeping mothers at home: parental leave for working women can last up to 8 years, but child care for infants and toddlers is locally controlled and highly inconsistent, thus incentivizing women to stay home. Additionally, maternity leave is mandatory for the first five months after childbirth at 80% of prior earnings. Thereafter, there is parental leave of 30% of earnings until the child is 8. (Henneck 2003).

Through slightly different fiscal mechanisms, the United States, France, Italy and Germany each manifest a goal of facilitating parental care in the home. Many German mothers drop out of the labor force after the leave period expired; for instance, of West German mothers taking leave in 1990 and 1992, about half did not return to work. (Henneck 2003). Tax matters.

d. Not Just the Income Tax

As if all of the above were not enough, one cannot consider the full effects of fiscal policy on child-bearing, household formation, and the sexual division of labor by looking at the income tax alone, even supplemented with reference to child-care mechanisms inside and outside the tax system. In the Unites States, once again, an extremely large and important tax system is the payroll tax that funds social security and medicare. And, here, yet again, one- earner families are large “winners” while two-earner ones are large “losers.” Non-income-earning spouses receive benefits based on their spouses’ paid work efforts, a policy dating back to 1939. But this then means that, in the majority of two-earner households, the “second” earner is paying a pure tax, with no offsetting benefit in her benefits profile, because she would get these anyway, as a spouse. (McCaffery 1999a).

The analysis could continue. Employer-provided “non cash” or “fringe” benefits, which now account for a large share of compensation in America (and world-wide), in large part because of favorable tax provisions, typically are built

on and reward a traditional, one-earner model, by extending free or low cost benefits to stay-at-home spouses, while requiring all workers to accept benefits, even if, on account of their having a working spouse, they do not need them. And so on: Governmental benefits and burdens of all sorts must be examined with some considerable care to unearth all of the biases and hidden coercive elements involved.

e. The Sum Total

Notwithstanding the considerable complexity touched on in the prior pages, this summary only suffices to scratch the surface of the role for sex and other cultural norms and biases in fiscal systems. In sum, in the U.S., there are severe marriage penalties affecting the working poor, on account of the loss of benefits as one enters the lower middle class. Marriage is very costly among the poor, who can least afford the price. In the middle-income classes, the secondary earner bias is severe. And among the upper-income classes, the incentives for a traditional “male breadwinner” model are clear.

There is plenty of evidence that these biases have been noticed world- wide. In the wake of World War II, the United States was influential in setting global tax policy (see Brownlee chapter, this volume, again), and so many countries adopted the American solutions. Yet after a few decades, other nations discovered that they could not afford them: the pressure on women to work was great, and this revealed the biases in the American way. (European Communities 1985). Hence, countries reverted to separate filing, which mitigated the secondary-earner bias. Still, these nations have tended to feel some remorse over the higher taxes paid by one-earner families—although this is the flip side, and necessary correlate, of eliminating the secondary earner bias—and so they have instituted one-earner tax relief. (Pechman and Englehardt 1991; OECD 2005). Some countries, however, seem to relish the traditional male breadwinner model, such as Germany and Japan. Others try to back off of it, but with unclear effects, like France and Italy. Still others try to create brave new worlds, like the Scandanavian countries, but, absent a full and deep exploration, there is little reason to have confidence as to the net effects. Everywhere, countries seem haunted by the ghosts of the past, and by norms with superficial appeal but little connection to anything that seems really to matter.

3. Three Themes

In this Section, I briefly return to the three themes noted at the outset, each of which has been advanced in the prior analytic section.

a. The Fog of Tax

I take it that there is little reason to expound any more on the “fog of tax;” I suspect that most readers came into this chapter with some sense of the dizzying complexity of the subject matter, and that that attitude has only deepened. And, as noted throughout, we have only scratched the surface. The EC Report in 1984 commented tersely that “[t]here are a bewildering variety of methods for taking a taxpayer’s family and personal circumstances into account in the calculation of tax liability.” (European Communities 1985). Indeed. No two countries seem to approach the matter in quite the same way, and there is n doubt, in the scholarly and political commentaries on point, that confusion abounds.

b. Rhetorical Manipulation and Cognitive Error

The complexity of the subject matter combined with some difficult abstractions, and perhaps political motivations, has meant that there has been a great deal of room for rhetorical manipulation as well as flat-out error. The OECD, for example, seems simply to misunderstand separate filing, confusing the U.S. style “married, filing separately” with “optional separate filing,” as detailed above. “Neutralities” and “equities” are explored using observable, monetary income, consistently missing the facts that “traditional” households with stay-at-home spouses have a large and important store of invisible, imputed income—often, indeed, encouraged by the tax system—which is not being taxed. Consider three quick case studies.

One comes again from the United States. The conservative Contract with America and other events in the early 1990s put the subject matter of “marriage penalties” in the tax system on the front page of many newspapers and high in the public political consciousness. Yet the only “solutions” ever enacted, under both the Democrat President Clinton and the Republican President George W.

Bush, have been in the direction of child, not child-care, relief and a doubling of the rate brackets: solutions that not only eliminate marriage penalties, but also increase marriage bonuses and exacerbate the secondary-earner bias. Child not child-care tax credits also help all couples with children, with or without working spouses or the need for paid third party care. In the wake of Taxing Women’s publication, I was honored to see that the book led to some consideration of optional separate filing, which, as we have seen, would eliminate marriage penalties without also increasing marriage bonuses. My honor was short lived: conservative forces, led by Phyliss Schlafly of the Eagle Forum (“Leading the pro-family movement since 1972,” is their motto) mounted a fierce counterattack, protesting that to do anything that helped only two-earner families would encourage working mothers and thus harm children. Once again, we see the effects of a baseline: the biases against two-earner families are deep and entrenched, but are now taken for granted. Schlafly was able to isolate a political and rhetorical focus on the change, and argue that it was unfair not to help traditional one-earner families. And so the U.S. has continued to do nothing particular to reverse decades of bias against the most common model of married households today—America is no Scandanavia.

Another interesting example comes from New Zealand in 1960. (Jones 2005). New Zealand had retained the British, overtly sexist “aggregation” model, in which wives’ income was simply added atop their husbands’, filing together under a singles’ rate schedule. In part because this is a harsh model with a severe secondary earner bias, the tax system evolved long ago to have a generous “working spouse” exemption that meant that only a small percentage of New Zealanders were impacted by aggregation: high-income ones with spouses who also had high incomes. Still, there was a broadly popular movement to repeal aggregation and move to a purer separate filing system. Proponents of change were able to exploit the rhetoric of hurting modern, two-earner families, and women, in order to benefit a comparative handful of the wealthiest.

A similar drama seems to be playing out today in Canada. (Kesselman 2007). This country has long had a system of separate filing. But of late the bias against one-earner families has become more noticeable, and there is a movement afoot to move towards a full or partial system of income-splitting— just like the U.S. has. Although it is easy to see that such a move would create secondary-earner biases, perhaps creating or exacerbating patterns of sexual discrimination in the workplace (McCaffery 1993b), the proposal seems to have attracted a great deal of support north of the border, including from the Green Party. Politics, fiscal sociology, and flat-out confusion make for strange bedfellows.

c. Elites and Efficiency

The three case studies just noted illustrate my third and final theme. In the fog of tax, with confusion swirling about and multiple characterizations possible, change is hard. Reforms responding to salient rhetorical characterizations often work at cross-purposes. Predicting the direction of change or the net effects of any one element of the system in isolation becomes all but impossible. And in the swirl of confusion and complexity, it is most often the case that one of two types of “winners” prevail.

One is elites, who always have advantages when it comes to manipulating the political process, and who have the time, resources and incentives to exploit them. For example, changes in America in the family arena, ever since 1948 and arguably before, have helped traditional, high wealth, one- earner families over all competing models, even those this is a model that fits fewer than 10% of American households. The lower classes are left to fend for themselves, on scraps of “workfare,” existing in a growing universe of single- parent households altogether outside the vaunted American “dream.” (McCaffery 1999b). Meanwhile, the middle classes are left to scrape on by, while the wealthy traditional one-earner families continue to prosper. This seems also to be true in Germany and Japan, two wealthy nations. The New Zealand and Canadian examples also illustrate how wealthy actors can use rhetoric of family and fairness simply to reduce their own taxes.

The chief counter-weight to elite private actors seems to be the macro- economic needs of the state. Thus, where it seems important to get women into the workforce, to drive up national income or to maintain wages at times of low unemployment, tax policy wearing a family-friendly face can come to the rescue.

So, too, with natalist concerns, which seem to form an explicit backdrop in

Germany and France at least.

Absent far more and far deeper exploration into individual country’s stories, it is hard to say much more, or to conclude that the apparent counter- examples of the Scandanavian countries in fact reflect richer substantive notions of equality between the sexes and accommodation of different household forms. But it does appear to be the case that, given the complexity, confusion, and relative advantages of the elites in finding ways to further their private ends, that fiscal policy is very often a hodgepodge, with different mechanisms pointing in different directions. Thus, for a large set of examples, while separate filing eliminates the secondary earner bias and encourages a model in which men and women share time both at home and in the paid workforce, the panoply of benefits to one-earner families in countries with separate filing (Pechman and Englehardt 1991; OECD 2005)—which bonuses to one-earner households are, of course, penalties on two-earner ones—reintroduce some if not all of the biases. Countries world-wide, in embracing fiscal tools to manage and shape households, have woven tangled webs.

4. Conclusion

We are born free, but are everywhere in chains, as Rousseau taught us. And the strongest of chains are those emanating from our households, where we were born and raised, as Freud taught us. States and state actors, who have

featured so prominently in most of the pages of this volume, have figured all this out. They have taken their power, where they have found it—and a more enduring, frequent source of power than even that of musket and missile has been the power of the fisc, to tax and spend—and used it to shape social life, consciously and unconsciously. Once in place, the fiscal choices have been hard to change: indeed, they are hard even to see or understand. Social choices and constructions linger, shaping lives in ways barely seen. Confusion and complexity form a breeding ground for narrow interest group politics, continued incremental reform, and the lingering choke-hold of the status quo.

Given what nations have done with the tools of fiscal power, it is hard to be sanguine about keeping those tools in their hands. Although there can be no naïve neutrality, perhaps it is time to get the sex out of fiscal policy. But an indispensable step towards that lofty goal is to get the sex into fiscal sociology, as a legitimate and deeply important tool for understanding where we are, now, and where we could be, tomorrow, with a little more thought and enlightenment.


Akabayashi, H. (2006). “The Labor Supply of Married Women and Spousal Tax Deductions in Japan—A Structural Estimation.” Rev Econ Household 4: 349-78.

Bankman, J. and T. Griffith (1987). “Social Welfare and the Rate Structure: A New Look at Progressive Taxation.” California Law Review 75: 1905-67.

Berliant, M. and P. Rothstein (2003). “Possibility, Impossibility and History in the Origins of the Marriage Tax.” National Tax Journal 56: 303-317.

Bird, R. M. and E. M. Zolt (2005). “Redistribution via Taxation: The Limited Role of the Persona Income Tax in Developing Countries.” UCLA Law Review 52: 1627-95.

Briggs, N. (1985). “Individual Income Taxation and Social Benefits in Sweden, the United Kingdom, and the U.S.A.: A Study of their Interrelationships and Their Effects on Lower-Income Couples and Single Heads of Household.” Bulletin of the International Bureau of Fiscal Documents: 243-61.

Cass, B. and D. Brennan (2003) “Taxing Women: The Politics of Gender in the Tax/Transfer System.” eJournal of Tax Research Volume, 1-22 DOI:

Crittenden, A. (2001). The Price of Motherhood: Why the Most Important Job in the World is Still the Least Valued. New York, Henry Holt and Company.

Duggan, L. (2003). “East and West German Family Policy Compared: The Distribution of Childrearing Costs.” Comparative Economic Studies 45: 63- 86.

European Communities (1985). “The EC Commission on Income Taxation and Equal Treatment for Men and Women: Memorandum of December 14, 1984 presented to the EC Council.” Bulletin of the International Bureau of Fiscal Documentation 39 (June): 262-66.

Groves, H. M., Ed. (1963). Federal Tax Treatment of the Family. Washington D.C., Brokkings Institution.

Henneck, R. (2003). Family Policy in the US, Japan, Germany, Italy and France: Parental Leave, Child Benefits/Family Allowances, Chid Care, Marriage/Cohabitation, and Divorce. A Briefing Paper prepared by the Council on Contemporary Families.

Jones, C. C. (1988). “Split Income and Separate Spheres: Tax Law and Gender Roles in the 1940s.” Law and History Review 6: 259-310.

Jones, N. (2005). “A History of Taxing Married Wome in New Zealand.” Auckland University Law Review 11: 147-173.

Kesselman, J. R. (2007). Taxing Couples: Is Income Splitting Fair?, Institute for Research on Public Policy, Montreal, Canada.

McCaffery, E. J. (1993a). “Taxation and the Family: A Fresh Look at Behavioral gender Bias in the Code.” UCLA Law Review 40: 983-1060.

McCaffery, E. J. (1993b). “Slouching Towards Equality: Gender Discrimination, Market Efficiency and Social Change.” Yale Law Journal 103: 595.

McCaffery, E. J. (1996). “Equality, of the Right Sort.” UCLA Women’s Law Journal 6: 289.

McCaffery, E. J. (1999a). Taxing Women. Chicago, University of Chicago Press. McCaffery, E. J. (1999b). “The Burdens of Benefits.” Villanova Law Review 44: 445.

McCaffery, E.J. (2003). “Marriage Penalty Relief in the New Tax Law. National Center for Policy Analysis. Brief Analysis No. 445.

McCaffery, E. J. and J. Baron (2004). “Framing and Taxation: Evaluation of Tax Policies Involving Household Composition.” Journal of Economic Psychology 25: 679-705.

McCaffery, E. J. and J. Baron (2005). “The Political Psychology of Redistribution,.” UCLA Law Review 52: 1745-92.

Mirrlees, J. (1971). “An Exploration in the Theory of Optimum Taxes.” Review of Economic Studies 38: 175-208.

Morris, E. (2004). The Fog of War — Eleven Lessons from the Life of Robert S. McNamara, Sony Pictures.

OECD (2005). Taxing Working Families, Organisation for Economic Co- Operation and Development.

OECD (2006). Fundamental Reform of the Personal Income Tax, Organisation for Economic Co-Operation and Development.

Palme, J. (2005). Why the Scandanavian Experinece is Relevant for The Reform of ESM. Stockholm, Institute for Future Studies.

Pechman, J. A. and G. v. Englehardt (1991). “The Income Tax Treatment of the Family: An International Perspective.” National Tax Journal 43(1): 1-22.

Richards, Wendy. 2008. “An Analysis of Recent Tax Reforms from a Marital-Bias Perspective: It is Time to Oust Marriage from the Code.” Wisconsin Law Review. 611-654.

Ryrstedt, Eva. “Family Law and Social Law: Reciprocal Dependnecy In a Comparative Perspective.” Inmternational Journal of Law, Policy and Family. 20:44-53.

Schelling, T. C. (1981). “Economic Reasoning and the Ethics of Policy.” Public Interest 63: 37-61.

Schumpeter, J. (1991 [1918]). The Crises of the Tax State. The Economics and Sociology of Capitalism. R. Swedberg. Princeton, Princeton University Press: 99-140.

Shaviro, D. (1999). Effective Marginal Tax Rates Facing Low-Income Households. Washington DC, Employment Policies Institute.

Siegel, R. B. (1995). “The Modernization of Marital Status Law: Adjudicating Wives’ Rights to Earnings, 1860 – 1930.” Georgetown Law Journal 82: 2127- 2211.

Smith,        T. “Women and Tax in South Africa.” From http://www.internationalbudget.org/resources/library/WomenSA.pdf.

Staudt, N. C. (1996). “Taxing Housework.” Georgetown Law Journal 84: 1571- 1647.

Stotsky, J. G. (1996). Gender Bias in Tax Systems. IMF Working Paper, International Monetary
Fund: 1-18.

Stotsky, J. G. (1997). “How Tax Systems Treat Men and Women Differently.” Finance and
Development(March): 30-33.

Tahxincioglu, E. (2006). From the Sandbox to the Corner Office: Lessons Learned on the Journey
to the Top. New York, Wiley.

Teck, T. H. (2004). “Income Taxation of Husband and Wife in Singapore and Malaysia.” Bulletin of the International Bureau of Fiscal Documentation: 520-524.

Thaler, R. H. (1980). “Toward a Positive Theory of Consumer Choice.” Journal of Economic Behavior and Organization 1: 39-60.

U.S.        Census Bureau. (2007). “Statistical Abstract.” from http://www.census.gov/compendia/statab/.

Ventry, D. J. (2006). “No Income Splitting for Domestic Partners: How the IRS Erred.” Tax Notes 110: 1221.

Ventry, D. J. (2007). The Politics and Economics of Gender Norms and Competing Family Forms under the U.S. Federal Income Tax, 1969-2006.

Zelenak, Lawrence. 2007. “The Tehory and Practice ofTax Reform.” Michigan Law Review 105: 1133-1150.

∗ Robert C. Packard Trustee Chair in Law and Professor of Law, Political Science, and Economics, University of Southern California Law School, Visiting Professor of Law and Economics, California Institute of Technology. This paper is being prepared as a chapter in a book project, Taxation in Perspective: Comparative and Historical Approaches to Fiscal Sociology, Issac Martin, Ajay K. Mehrota, and Monica Prasad, editors, and a version was given at a talk to the Tax Program students at Northwestern Law School. I thank Ted Dillman for excellent research assistance.

Previously published by the University of Southern California Law School

Robert C. Packard Trustee Chair in Law and Political Science, USC Gould School of Law, Los Angeles, USA