Gender bias in tax systems reflects prevailing social norms. Many industrial and some developing countries are reforming their tax systems to reflect changing attitudes about men’s and women’s roles in society.
Tax systems reflect a tapestry of decisions, made over many years. These decisions have been influenced by a variety of factors, including social attitudes about the respective roles of men and women. As a result, many tax systems exhibit gender bias—they treat men and women differently in ways that can negatively affect their decisions on whether and how much to work, their personal consumption habits, and their overall tax liability. While it is easier to detect gender bias in personal income tax arrangements, gender bias—explicit and implicit—may be present in other taxes such as consumption taxes and import duties. Gender bias may also be found in the way payments are linked to the receipt of benefits under social insurance programs.
Prompted by changing attitudes, many countries in recent years have reformed their tax systems to reduce gender bias (see Box 1). In the 1980s, for example, several Western European countries reformed their personal income tax laws to eliminate provisions that explicitly discriminated against women. More generally, in industrial countries the personal income tax system based on joint filing by members of the same family unit has given rise to a discussion over how the income tax treats the incomes of secondary earners (generally assumed to be women) and the incentives the income tax has on their work patterns, child bearing, and other behavior.
But reform has not been confined to the industrial countries. Although developing countries as a group have generally been slower to implement reforms than the industrial countries, some have also begun to change their tax systems to address gender bias. As in the industrial countries, these efforts have generally focused on the personal income tax.
Yet, in examining the issue of gender bias, we should not forget that apparent bias may be acceptable in some societies because it accommodates typical social arrangements or encourages certain generally desirable social behavior. In these societies, such provisions are not discriminatory but simply reflect prevailing norms. Moreover, there may be grounds for discriminating between men and women in the tax code that are related to, for example, life expectancy, such as in the tax treatment of pensions and annuities.
Men, women, and income tax
Gender bias may be both explicit and implicit. Explicit forms are specific provisions of the law that treat men and women differently. They are relatively easy to identify, since they depend largely on the language used in the tax code or tax regulations. Implicit forms of gender bias are provisions of the law that, because of typical social arrangements and economic behavior, tend to have different implications for men and women. It is much more difficult to identify implicit gender bias, because it depends in large part on value judgments as to desirable social and economic behavior, which may vary considerably from society to society and from one time period to another (see Box 2 for an example of the latter).
Box 1Eliminatmg gender bias
A number of countries have taken steps to eliminate gender bias in the personal income tax—for example:
• France (1983) moved from requiring only a husband’s signature on family tax returns to requiring that both spouses sign.
• Ireland (1993) moved from joint filing in the name of the husband with an option for separate assessment on labor income for the wife, to an option for the wife to be the “primary taxpayer.”
• Malaysia (1991) moved from a tax system in which the income of a married woman was attributed to her husband unless she elected separate assessment, to a system in which husbands and wives are treated as separate taxable units with an option for joint treatment.
• The Netherlands (1984) moved from granting a higher tax-free allowance to a married man than to a married woman, to an equal basic tax allowance.
• South Africa (1995) moved from applying a higher rate schedule to single persons and married women than to married men, to a unified schedule.
Explicit and implicit bias
Explicit gender discrimination is more typically found in the personal income tax than in other taxes, because the income tax applies to individuals or family units and therefore more easily accommodates different treatment for men and women. Explicit discrimination can be found in the rules governing the allocation of shared income (such as nonlabor income and income from a family business); the allocation of exemptions, deductions, and other tax preferences; in tax rates; in who files the tax return; and in who pays the tax. It is easiest to detect implicit gender discrimination in the personal income tax, compared with other taxes, since it directly affects labor supply and other behavior.
Income taxes may be global or schedular. Under a global income tax, income is aggregated and typically only one schedule of tax rates applies to it. In contrast, under a schedular system, each type of income faces a different schedule of tax rates. Global income taxes have typically been the source of gender bias and the focus of efforts to eliminate it, particularly in industrial countries. Explicit gender discrimination in a pure schedular income tax is not very common, because the tax liability is established with respect to a particular source of income rather than a particular taxpayer. In countries with a global personal income tax, gender bias may take several forms, including the way in which nonlabor or business income and tax preferences are allocated, tax rates, and the responsibility for complying with the law.
Types of discrimination
The way in which nonlabor income is allocated often entails explicit gender discrimination. Typically, under a system of individual filing, the wage income of married couples is attributed to the worker, while nonlabor income poses a more complex problem since it must be allocated to one or the other spouse. There are, however, several gender-neutral ways to allocate this income, such as attributing all of it to the higher earner, allocating the income equally between the spouses, allowing couples to allocate the income in whatever way they choose, or allocating the income to the spouse with legal ownership of property that is not jointly held.
Many tax codes using individual filing contain explicit gender bias in that they allocate all nonlabor income to the husband regardless of the circumstances. This practice is derived from the English common law tradition in which all income earned by a married couple was assumed to be the property of the husband. This contrasts with the civil law tradition, more prevalent in Latin countries, in which income earned by a married couple during the course of their marriage was considered “community property” (i.e., the property of both spouses). The allocation to the husband may raise the tax on this income, if the husband pays tax at a higher marginal tax rate.
Family business income is also attributed in many countries to the husband regardless of the respective roles of the spouses in the business. While shifting income from a family business to the spouse who pays tax at a lower marginal tax rate is a common means of tax avoidance, there are administrative solutions to this problem that do not require attributing all of the income to the husband.
The allocation of deductions, exemptions, and other tax preferences is another way in which explicit gender discrimination occurs in a system of individual filing. Typically, countries give exemptions or deductions for various purposes, for dependent children, a nonworking spouse, and so on. Under a system of individual filing, these exemptions and deductions must be allocated across spouses. In some countries, the nature of the exemptions and deductions that taxpayers may claim varies according to whether the taxpayer is a man or a woman. This allocation may also affect the tax liability.
Another form of gender discrimination is that countries may levy different tax rates on men and women, with a higher rate being applied to married women, as was done, for instance, in South Africa until 1995.
In some countries, a form of gender discrimination is a requirement that a joint tax return be submitted in the name of the husband; thus, a wife has no separate existence as a taxpayer. For example, until 1990, the British income tax provided for the husband alone to file the joint tax return, as did the French income tax until 1983. Switzerland continues this practice today.
A system of joint filing with a progressive marginal rate schedule, such as exists in the United States, may discourage secondary workers, because the tax on their income starts at the highest marginal tax rate of the primary income. This so-called “marriage tax” under a system of joint filing has been typically viewed as discriminating against women, although it would, of course, apply equally to a husband if he were the secondary earner.
Studies of income taxation have long taken gender into account by explicitly considering the differences in the labor supply behavior of men and women, and their implications for public policy. These studies suggest that the labor supply response of married women, who are often assumed to be the secondary earners, to the income tax is greater than that of married men. This implies that to minimize the efficiency cost of the income tax, everything else being equal, married women should be taxed at a lower rate than married men.
Steps to reform in industrial countries. In Europe, the issue of gender neutrality in tax systems rose to prominence in the 1980s. A 1984 report of the European Community (EC) (see reference) examined whether EC tax systems were neutral with respect to women’s labor force participation. The main concern was that secondary earners faced a high marginal tax rate under a system of joint filing that would create a disincentive for women to work.
The report concluded that European tax systems had an adverse effect upon married women’s tax burdens in several areas, including joint taxation; granting of allowances or tax reductions a priori to the husband; the lack of an allowance or deduction for the costs incurred in child care and domestic help when both spouses work outside the home; the inability of women to declare their own income for tax; the responsibility for the nonpayment of tax by the other spouse; and limitations on the amount of income that can be paid to an “assisting wife” by a husband.
The report recommended a system of totally independent taxation to achieve equal treatment of men and women and, at the very least, to allow separate assessment as an option. A number of European countries, including France, the Netherlands, and the United Kingdom, have reformed their tax systems to eliminate explicit gender discrimination.
To some, the idea that gender neutrality in personal income taxation requires independent taxation of husband and wife might seem radical. Certainly, explicit discrimination can be eliminated without independent taxation of spouses. But it is debatable whether the elimination of implicit discrimination requires independent taxation. In contrast to the conclusions of the EC report, many hold the view that the family is the preferable unit of taxation because of the nature of joint consumption within the household. They do not see any inherent conflict between gender equality and joint taxation. Some studies that have examined the treatment of the family under the income tax laws of industrial countries found that there are wide variations as well as areas of similarity. Presumably, these differences reflect value judgments of the different societies about the way they view the family, or they may simply reflect historical inertia.
Box 2Evolution of income tax reform in the United Kingdom
The British income tax offers an interesting perspective on the evolution of attitudes toward women. The income tax in the United Kingdom was first introduced in 1799. A couple’s income was attributed to the husband, reflecting the legal status of married women and the property laws of the time. In 1894—12 years after the enactment of the Married Women’s Property Act in 1882, which allowed women to retain management and control of their separate property and earnings—a tax provision was introduced that allowed a wife’s earnings to qualify for the same tax relief as those of a single person if a couple’s combined annual income was less than £500. In 1918, a married man’s allowance was introduced; it was increased in 1982 to 1.6 times a single person’s allowance. As recently as the mid-1970s, the Inland Revenue corresponded only with husbands, refused to reply directly to married women who sent in letters, and mailed any repayment due on over-withheld taxes from married women’s paychecks to their husbands. These practices led to criticism. The 1978 Finance Act gave married women the right to receive their own withholding repayments and the Inland Revenue began to reply directly to married women who had written to them.
However, these reforms were not sufficient to silence the critics. Several noted committees and academics began to recommend radical reforms in the system of personal income taxation. In 1988, the government legislated, effective in 1990, that all taxpayers should file returns on an individual basis on both earnings and investment income. The “married man’s” allowance for couples continued until 1993, when it was converted to an allowance transferable between the spouses.
For details, see Norma Briggs, 1985, “Individual Income Taxation and Social Benefits in Sweden, the United Kingdom, and the USA: A Study of Their Interrelationships and Their Effects on Lower-Income Couples and Single Heads of Household,” Bulletin of the International Bureau of Fiscal Documentation, Vol. 39 (June), pp. 243–61.
The US personal income tax has never contained any explicit gender discrimination (except for, at one time, a small difference in child care allowances). Nevertheless, the issue of the appropriate treatment of family income has frequently been the focus of debate, and the tax code has changed over time in its treatment of the couple vis-à-vis the individual. Today, the US tax code provides a clear advantage to a one-earner couple, compared with a two-earner couple. In some cases, the tax code is advantageous to married couples compared with single taxpayers, but, in other cases, is disadvantageous to the former, depending on how the earnings of the spouses are split.
Developing countries. Gender discrimination also appears in the tax systems of many developing countries. The most common form is to attribute the income of a married woman to her husband and to levy the tax in the husband’s name for any nonschedular income taxes, though a separate assessment of a wife’s employment income is allowed in many cases.
Reform has been taking place in some developing countries. For example, until recently, the South African tax system used different rate schedules for married persons, single persons, and married women, applying a higher rate to the latter two categories. These rates were unified in 1995. In Malaysia, the tax system was changed in 1991 from one in which the income of a married woman was attributed to her husband unless she elected for separate assessment, to a system in which husbands and wives are treated as separate taxable units. The wife’s income is still reported on the husband’s tax return, and joint assessment is still allowed.
Some developing countries have explicit provisions in the income tax code that distinguish men from women so as to accommodate typical social arrangements or to encourage certain social behaviors. Although some observers feel these provisions are not discriminatory and simply reflect prevailing societal norms, others believe these practices legitimize a secondary role for women.
Some developing countries explicitly discriminate in favor of women. For example, Singapore’s tax system is unique in the nature of explicit gender differentiation it builds into the income tax in the form of child relief. A basic child relief is available. In addition, a married woman is entitled to additional allowances for children if she has elected to be charged to tax in her own name and has passed at least three subjects in one sitting at the examinations for the General Certificate of Education or has obtained an equivalent or higher educational qualification—a provision whose objective appears to encourage well-educated women to have children. The tax code in Pakistan discriminates in favor of women by allowing a basic exemption that is higher for a working woman than a man, and the tax code in India also contains provisions favoring women.
Taxes on goods and services, such as the value-added tax and retail and excise taxes, tend not to show explicit gender bias in that the tax liability is established with respect to the purchase or production of a commodity. The same is true of taxes on foreign trade, such as customs duties. In practice, however, these taxes are not gender neutral and may have certain implicit biases.
There are many ways in which implicit bias could manifest itself in a consumption tax. One way is through the choice of goods and services covered by the tax. Implicit biases may result from different consumption patterns by men and women of these goods, though the exact nature of these implicit biases is difficult to establish. The issue is, in fact, more complicated, since such goods are usually purchased within the context of a household containing both men and women.
Excise taxes tend to have a more obvious implicit bias than broad-based consumption taxes. For instance, excise taxes are typically levied at a high rate on alcohol and tobacco, which are disproportionately consumed by men. Therefore, excises might be viewed as having an implicit bias against men, though, again, in a household context, if other members of the household reduce their consumption of other goods so that men can maintain their consumption of taxed goods, this might lessen the bias against men.
Preferential treatment of particular consumers or producers also might imply gender bias, but, as with preferential treatment of specific goods and services, a lack of neutrality would be difficult to establish. For instance, if nonprofit institutions typically provided services that benefited the poor disproportionately and the poor consisted mainly of female-headed households, then, implicitly, preferential treatment of nonprofit institutions would be biased toward women and children.
In many countries, import duties are a critical component of the tax system and clearly influence the pattern of economic development. One little-studied issue is whether typical patterns of import duties tend to favor industries that employ primarily men or women. Import duties in industrial countries that discriminate against low-technology goods may tend to be biased against women, because they often account for a large share of factory workers in low-income countries. Import duties in developing countries might shift development in ways that favor or disfavor women. Similar issues might arise with the corporate income tax.
Grounds for gender bias?
Are there legitimate grounds for discriminating between men and women in the tax code? Perhaps the most pervasive argument used to provide a justification for distinguishing between men and women in the tax code for some purposes is differences in average life expectancy between men and women. For instance, in the US tax code, a deduction is allowed for charitable contributions that take the form of a contribution of property that is effective upon the death of the taxpayer though the taxpayer retains the use of that property and its income until his/her death. At the time of the contribution, the taxpayer is allowed to take a deduction for the present discounted value of that contribution, based upon the taxpayer’s remaining life expectancy. For instance, a woman donating property at the age of 60 might have a life expectancy of 25 years while a man might have a life expectancy of 20 years. Thus, she would discount the donation over a 25-year horizon, and he would do so over a 20-year horizon. Similar issues arise in relation to incomes from pensions and annuities. Men could be required to receive a larger proportion of the total value each year, starting with the age of receipt, since their life expectancy is shorter. Differences in average life expectancy between men and women might also provide an argument in favor of applying different social security tax rates to men and women, providing different benefits for similar contribution histories, or using a different formula for linking tax payments to benefits.
Public finance has only recently fully acknowledged the importance of gender bias in public policies. Gender bias in tax systems is therefore a fruitful area of inquiry. Many nations have tried to eliminate explicit and implicit gender bias in tax systems while others have made little progress. Variation in cultural norms will undoubtedly continue to lead to differences of view as to what constitutes discrimination and the need for change.
This article draws on the author’s Working Paper (WP/96/99), “Gender Bias in Tax Systems,” IMF, Washington, 1996.