The tax treatment of the family.
Taxing people as individuals would remove some of the important disparities between married couples and unmarried partners. While many other industrialized countries have individual-based income tax systems, the US system is family-based. (1) Family-based tax systems create a number of difficulties that both reduce prosperity and create inequality across households and individuals.
First, progressive family-based tax systems--like the current federal tax system--penalize income earned by the secondary earner. In a progressive tax system, the average tax rate rises with income. Primary earners and single people who earn small amounts of income face relatively low average tax rates. But because the progressive rate structure is imposed at the family, rather than the individual, level, a married couple's average tax rate rises when the secondary earner begins to work. In other words, secondary earners--even those who earn small amounts of income--typically face high average tax rates. This feature of the tax system is economically harmful because secondary earners are most often married women who are very likely to cut back on work because of high tax rates. High average tax rates therefore discourage married women from participating in the labor market, reducing economic output and increasing gender inequality in earnings.
Second, the differential treatment of married couples and unmarried partners raises concerns about fairness. In some cases, two individuals pay more in taxes as a married couple than they would as two singles; in other cases, they pay less. Whether marriage is penalized or rewarded depends on the couple's income level and how earnings are divided between spouses. Marriage is typically rewarded for single-earner couples and is often penalized for dual-earner couples with similar incomes.
While these marriage "bonuses" and "penalties" probably have little effect on the decision to marry, they raise important fairness concerns because unmarried partners do not receive the bonuses and penalties. These concerns have become more pronounced in recent decades as the number of nontraditional families, including unmarried partners and same-sex couples, has increased.
However, moving to an individual-based tax system also raises important concerns about fairness. First, just like the current tax system, many means-tested welfare programs are based on family income. Because these programs also create work disincentives for secondary earners and disparities between married and unmarried couples, it may be appropriate to shift them to an individual basis as well. However, basing welfare benefits on individual income could allow the nonworking spouse of a high earner to receive benefits, an outcome that is inconsistent with the purpose of these programs. This issue could be addressed either by continuing to base welfare programs on family income, as many countries with individual-based tax systems do, or by reducing the amount of means testing in welfare programs.
Second, an individual-based tax system creates disparities across families with similar incomes based on how income is distributed between the two spouses. For example, under progressive tax rates, a family in which both spouses earn roughly equal incomes would pay less in taxes than a family in which one spouse earns all the income. On the other hand, in an individual tax system, individuals in similar circumstances are always treated similarly. Therefore, we must consider whether fairness is best evaluated at the individual or family level.
We begin by examining the effects of individual--versus family-based taxation on the level of economic output, focusing on the work decisions of secondary earners. Next, we explore issues of fairness under family-based taxation compared with individual-based taxation. Finally, because it is important to think about the system of taxes and transfers as a whole, we discuss the issues involved in individual- versus family-based transfer programs. We argue that, despite the complexity of the issues involved, it is worth considering a switch to an individual-based system, particularly for taxes and universal transfer programs such as Social Security.
Secondary Earners' Work Incentives
A number of economists have pointed out that family-based taxation creates work disincentives for secondary earners. Consider, for example, the Jones family, a married couple with two children. Suppose Mr. Jones earns $50,000 per year. Mrs. Jones is not currently employed but is considering a job that pays $30,000. If she accepts the job, the family's income would less than double; however, the family's combined income tax bill would increase more than tenfold, from $438 to $4,938. (2) This amounts to a 15 percent tax on Mrs. Jones's entire earnings. (3)
Note that under a family-based tax system, both spouses face the same marginal tax rate. That is, they pay the same tax on an additional dollar of earnings, regardless of which spouse earns it. In particular, Mr. Jones's income puts the family in the 15 percent tax bracket, so both Mr. and Mrs. Jones would pay a 15 percent tax on any additional earnings. However, Mrs. Jones faces an average tax rate (in other words, total taxes as a share of earnings) of 15 percent, much higher than the average tax rate the couple faces when only Mr. Jones works or the average tax rate Mrs. Jones would face if she were single. (4)
Both marginal and average tax rates can affect work decisions. Generally, average tax rates affect an individual's decision to enter the labor force, while marginal tax rates influence an individual's decision to work more hours. Consider a worker who is deciding whether to earn an extra $100 per week by working five more hours per week (at $20 per hour). If her marginal tax rate is 20 percent, the increase in her take-home pay would be $80 per week. The worker compares this amount with the value she places on using the extra five hours for leisure or household chores instead. Her decision to work more hours and earn a small increase in income depends on the marginal tax rate she would face.
On the other hand, consider an individual who currently does not work but is considering a job that pays $40,000 per year. For this large change in income, the average tax rate is more relevant. Consider a simple tax system in which the first $ 10,000 is free of tax and the remaining income is taxed at a 20 percent rate. Under this system, the individual would pay a tax of $6,000 ($10,000 times 0 plus $30,000 times 0.20), leaving a take-home pay of $34,000. She compares this take-home pay with the value of the free time she would need to sacrifice to do the job. In this case, the average tax rate (6,000 divided by 40,000, which equals 15 percent) is more relevant to her decision than the marginal tax rate of 20 percent.
Taxes harmfully distort the value of work compared to the value of free time. If the worker considering the $40,000-per-year job values the free time she would have to sacrifice at $35,000, the tax would prevent her from working even though the value of her work to her employer--reflected in the $40,000 salary she is offered--is greater than the cost of giving up her free time.
When labor supply is very responsive to taxes, high taxes cause individuals to reduce their work substantially, imposing a large cost on society. Michael Keane's 201 1 survey presents evidence that women's work decisions are very responsive--more responsive than those of men--to take-home pay. This is especially true for the decision to work or not work, for which the average tax rate is most relevant. (5)
Because taxes do more harm when work decisions are highly responsive to them, economic output is maximized when individuals who are the most sensitive to taxes pay low rates. (6) In an early and widely cited 1983 study, Michael Boskin and Eytan Sheshinski argued that because secondary earners are more responsive to tax incentives than primary earners, they should face lower tax rates than primary earners. (7) Because secondary earners by definition have lower incomes than primary earners, such an outcome can be partially achieved through an individual-based tax system in which tax rates increase with income. (8)
In 2011, Alberto Alesina, Andrea Ichino, and Loukas Karabarbounis argued for tax rates that differ explicitly by gender, again citing the difference in male and female responsiveness to taxes. (9) Concerns about fairness are likely to rule out gender-based taxation. (10) However, it is clear that the current system, in which married women typically face high average tax rates on even small amounts of income, causes economic harm by discouraging women from working. An individual-based tax system equalizes average tax rates between men and women with the same earnings and can therefore increase prosperity.
Economists often cite three criteria for evaluating the fairness of family-based tax systems:
* Horizontal equity: Married couples with the same income should pay the same amount in total taxes.
* Progressivity: The average tax rate should increase with family income.
* Marriage neutrality: Two individuals' total tax bill should not change if they get married.
Despite the intuitive appeal of each criterion, achieving all three goals is impossible. Suppose there are four single individuals: individuals A and B each earn $40,000, individual C earns $80,000, and individual D earns nothing. Consider a very simple progressive income tax under which the first $ 10,000 of income is tax free for individuals, and the remaining income is subject to a 20 percent tax. Under this system, A and B would each pay $6,000 in taxes, C would pay $14,000, and D would pay nothing.
Now suppose A and B get married, and C and D get married. To make the tax system marriage neutral, A and B's total tax as a married couple should be $12,000, while C and D's total tax as a married couple should be $14,000. In other words, marriage should not change the total tax paid by either couple. This is exactly what happens under an individual-based tax system. However, this outcome violates horizontal equity because both couples have the same income ($80,000) but pay different amounts in taxes.
To achieve horizontal equity across married couples, one would need to implement a family-based tax rate schedule that applies to all couples. If married couples were to face the same rate schedule as singles, then both couples would pay $14,000 in taxes on their $80,000 income. In this case, however, A and B would face a marriage penalty. Alternatively, suppose the tax brackets for married couples were twice as large; that is, suppose married couples were to pay no tax on the first $20,000 of income and then 20 percent on any additional income. Under this system, both couples would pay $12,000 in taxes, with C and D receiving a marriage bonus.
Therefore, any household-based system with progressive rates will create marriage bonuses and penalties. Under the current system, most couples receive marriage bonuses, with one-earner couples benefiting more than two-earner couples. Before 2001, the majority of couples already received marriage bonuses. (11) The 2001 tax cuts further reduced marriage penalties and increased marriage bonuses. (12) Marriage penalties generally apply to couples with similar incomes and are caused by the fact that tax brackets for married couples are sometimes not twice as large as brackets for singles. In addition, the means-testing of benefits such as the Earned Income Tax Credit and Medicaid can create large marriage penalties because one spouse's income can cause the other spouse to lose benefits. (13)
Despite their prevalence, marriage bonuses and penalties are unlikely to have a large effect on marriage decisions. For example, a 2000 study by Nada Eissa and Hilary Hoynes estimated that a $1,000 increase in the marriage penalty reduced the probability of marriage by 0.4 percentage points. (14) In their 1999 and 2003 studies, James Alm and Leslie Whittington similarly found that the effect of taxes on marriage is small. (15) This means that marriage neutrality is mainly an issue of fairness.
A key reason why it is impossible to achieve all three fairness criteria is that horizontal equity is defined based on families rather than individuals. Both horizontal equity and marriage neutrality are possible in a progressive, individual-based system if horizontal equity is redefined to require that two individuals with identical incomes should face the same tax bill. Of course, such a system will not result in horizontal equity across families. In our earlier example, under individual filing, A and B face a lower household tax bill than C and D because income is more equally divided across the two spouses.
The main argument for family-based horizontal equity is that families share resources, making it appropriate to consider a family's total resources in determining its ability to pay to fund the government. However, one could argue that living arrangements are a consumption choice like any other. Individuals who choose to save on living expenses by forming families should not necessarily be treated any differently than those who choose to save on living expenses by living with roommates or buying less expensive cars.
In addition, because only legally married couples are considered to be a "household," our household-based system creates horizontal inequity between married couples and unmarried partners, as married couples face marriage penalties or bonuses while unmarried partners do not. This problem has become more pronounced in recent decades as the number of unmarried partners has grown. Table 1 shows that the fraction of unmarried partner households, including same-sex couples, has almost doubled since 1990, while the fraction of married-couple households has fallen.
An additional source of horizontal inequity is the disparate treatment of same-sex and opposite-sex married couples. While same-sex partners may marry or form domestic partnerships in some US states, the Defense of Marriage Act requires that they file their federal taxes as singles. As a result, same-sex married couples may be subjected to additional tax complexity. (16)
While individual-based taxation deals quite well with these fairness concerns, it also raises some issues of its own. One question regards the allocation of asset or other nonlabor income between spouses. If spouses can transfer assets between themselves, they may be able to avoid paying taxes by assigning ownership of assets to the spouse in the lower tax bracket. Melvin Stephens Jr. and Jennifer Ward-Batts found that this avoidance strategy is common under the individual tax system in the United Kingdom. (17)
One solution to this problem is to divide all asset income equally between spouses, or to assign all asset income to the spouse with the higher earnings, regardless of legal ownership. Of course, that approach would undermine marriage neutrality to some extent because the tax system would reassign asset income if and only if the individuals in question are married.
A similar question regards dependent exemptions and other child-related tax provisions. Again, several approaches are possible. For example, a couple could choose which spouse takes the dependent exemption for a particular child, or the dependent exemption could be assigned to a particular spouse, such as the one with the higher income.
Depending on the treatment of state-level community property laws, individual taxation may also create inequality with respect to the treatment of wages. In community property states, each spouse legally owns half of the couple's property, including earnings from their jobs. Under individual taxation, when two individuals with unequal incomes marry in a community property state, their total tax bill will generally fall because earnings are reassigned from the high-earning spouse to the low-earning spouse, who will be in a lower tax bracket. Indeed, the perceived unfairness of marriage bonuses in community property states was a main motivation for switching from an individual-based to a family-based tax system in 1948. (18) One way to deal with this issue is to have an individual-based tax system disregard community property laws and in all states treat wages as belonging to the spouse who earned them.
Dealing with Transfer Programs
Looking at taxes alone provides an incomplete picture of how different kinds of households are treated. The phase-out of means-tested transfer programs can also discourage secondary earners from working and create disparities between married and unmarried couples. For example, a single parent who qualifies for Medicaid may lose coverage if he or she marries a spouse with even a modest income. (19) Even in countries with individual taxation, welfare benefits are often based on household income, resulting in high taxes on secondary earners in low-income households. (20)
In some cases, welfare benefits can be large enough to influence marriage decisions. For example, some observers have noted that when one member of an older couple needs long-term care (which is generally not covered by Medicare), it may be in the couple's interest to get divorced so that the spouse who needs care can qualify for Medicaid without depleting the family's assets. (21)
In addition, Social Security can impose high implicit tax rates on secondary earners. For singles and primary earners, paying the Social Security payroll tax has an offsetting benefit: it earns them a higher Social Security benefit in retirement. It therefore does not discourage work as much as an income tax. But many secondary earners expect to collect benefits based on their spouse's earnings record rather than their own, so their Social Security benefit does not increase with additional work. For these individuals, working and paying the payroll tax has no offsetting benefit because they will receive their spousal benefit regardless of whether they work.
One way to address these issues would be to treat people as individuals for both taxes and transfers. This is fairly straightforward for universal programs like Social Security. Workers could be treated as individuals or married couples could use earnings sharing, in which each spouse receives credit for half of the couple's total earnings. However, switching to an individual-based system is much more difficult for means-tested programs. The main objection to an individual-based welfare system is that nonworking or low-earning spouses from high-income households may qualify for means-tested benefits. Of course, the same objection applies to allowing the low-earning spouses of millionaires to benefit from low tax rates under an individual tax system.
Another alternative would be to restructure the welfare system to rely less on means-tested programs. The most radical version of this reform would provide each individual with a lump-sum payment regardless of income. Some research suggests that such a system, coupled with a flat income tax, may be a good way to deal with the tradeoff between work incentives and progressivity. (22) Such a system would also be more transparent than the current system of overlapping and complex transfer programs.
We have shown that a family-based tax system discourages work by secondary earners and treats married couples differently than unmarried partners. While an individual-based tax system eliminates these distortions, it raises issues of its own. For example, it raises the question of how to treat income on shared assets, and it creates situations in which households with similar incomes face very different tax liabilities. Furthermore, moving to a fully individual tax and transfer system may require rethinking the structure of welfare programs.
Even with these drawbacks, we think an individual tax system is worthy of careful consideration. In comparison to 1948, when the family-based federal tax system first went into effect, there are now far more women in the workforce and a greater number of unmarried and same-sex partner households. Switching to an individual tax system can create large gains for society by improving work incentives for secondary earners and treating different kinds of families fairly.
We thank Alan Viard for his helpful discussion and Brittany Pineros for her excellent research assistance.
(1.) For details, see James Alm and Mikhail Melnik, "Taxing the 'Family' in the Individual Income Tax" (working paper, Andrew Young School of Policy Studies Georgia State University, Marietta, GA, 2004), table 3; and Herwig Immervoll et al., "An Evaluation of the Tax-Transfer Treatment of Married Couples in European Countries" (Institute for the Study of War, discussion paper No. 3965, Bonn, Germany, 2009), table A4.
(2.) See the calculator at http://taxpolicycenter.org/taxfacts /marriagepenaltycalculator.cfm. For these calculations, it is assumed that the couple has two children between the ages of 13 and 17, no income other than earnings, and no itemized deductions.
(3.) Payroll taxes, which are not included in this calculation, are likely to impose an additional penalty. We discuss the effects of payroll taxes and the Social Security benefit formula later in this paper.
(4.) Of course, the average tax rate on Mr. Jones is also higher than the average rate the couple faces if only Mrs. Jones works. But the difference is less dramatic because of Mr. Jones's higher income. Also, as discussed in the text, Mr. Jones's work decision may be less influenced by taxes.
(5.) Michael P. Keane, "Labor Suppy and Taxes: A Suivey," Journal of Economic Literature 49, no. 4 (2011): 961-1075. See also Richard Blundell and Thomas MaCurdy, "Labor Suppy: A Review of Alternative Approaches," in Handbook of Labor Economics Volume 3, ed. Orley Ashenfelter and David Card (Elsevier Science, 1999), 1559-695.
(6.) Frank P. Ramsey, "A Contribution to the Theory of Taxation," Economic Journal37, no.145 (1927): 47-61.
(7.) Michael J. Boskin and Eytan Sheshinski, "Optimal Tax Treatment of the Family: Married Couples," Journal of Public Economics 20 (1983): 281-97. Other economists have countered that a household-based tax may be preferable to an individual-based tax because taxing spouses at different marginal rates can result in household chores being assigned to the spouse who is less efficient at doing them, simply because that spouse faces a higher marginal tax rate on paid work. For a debate on this issue, see John Piggott and John Whalley, "The Tax Unit and Household Production," Journal of Political Economy 104, no. 2 (1996): 398-418; Patricia F. Apps and Ray Rees, "Individual versus Joint Taxation in Models with Household Production," Journal of Political Economy 107, no. 2 (1999): 393-403; Peter Gottfried and Wolfram F. Richter, "The Tax Unit and Household Production: A Comment," Journal of Political Economy 107, no. 2 (1999): 404-09; and John Piggott and John Whalley, "Reply to Apps /Rees and Gottfried/Richter," Journal of Political Economy 107, no. 2 (1999): 410-18.
(8.) Immervoll et al., "An Evaluation of the Tax-Transfer Treatment of Married Couples in European Countries."
(9.) Alberto Alesina, Andrea Ichino, and Loukas Karabarbounis, "Gender-Based Taxation and the Division of Family Chores," American Economic Journal: Economic Policy 3, no. 2 (2011): 1-40.
(10.) For discussions of "tagging," or basing taxes on characteristics such as race, gender, height, or age, see Matthew Weinzierl, "Why Do We Redistribute so Much but Tag so Little? The Principle of Equal Sacrifice and Optimal Taxation" (National Bureau of Economic Research Working Paper 18045, Cambridge, MA, 2012); and N. Gregory Mankiw and Matthew Weinzierl, "The Optimal Taxation of Height: A Case Study of Utilitarian Income Redistribution," American Economic Journal: Economic Policy 2, no. 1 (2010): 155-76.
(11.) US Congressional Budget Office, "For Better or for Worse: Marriage and the Federal Income Tax" (June 1997), www.cbo.gov/sites/default/files/cbofiles/ftpdocs/0xx/doc7 /marriage.pdf.
(12.) Adam Carasso and C. Eugene Steuerle, How Marriage Penalties Change under the 2001 Tax Bill (Urban Institute, 2002), www.urban.org/UploadedPDF/410491.pdf.
(13.) See Adam Carasso and C. Eugene Steuerle, "The Hefty Penalty on Marriage Facing Many Households with Children," Urban-Brookings Tax Policy Center 15, no. 2 (2005): 157-75.
(14.) Nada Eissa and Hilary Hoynes, "Tax and Transfer Policy, and Family Formation: Marriage and Cohabitation" (University of California, Davis, working paper, December 2000).
(15.) James Alm and Leslie A. Whittington, "For Love or Money? The Impact of Income Taxes on Marriage," Economica 66, no. 263 (1999): 297-316; and James Alm and Leslie A. Whittington, "Shacking Up or Shelling Out: Income Taxes, Marriage, and Cohabitation," Review of Economics of the Household 1 (2003): 169-86.
(16.) For further discussion, see Robertson Williams, "DOMA's Tax Hassles for Same-Sex Couples," Tax Vox, March 26, 2013, http://taxvox.taxpolicycenter.org/201 3/03/26/domas-taxhassles-for-same-sex- couples/.
(17.) Melvin Stephens Jr. and Jennifer Ward-Batts, "The Impact of Separate Taxation on the Intra-Household Allocation of Assets: Evidence from the UK," Journal of Public Economics 88, no. 9-10 (2004): 1989-2007.
(18.) See Sara LaLumia, "The Effects of Joint Taxation of Married Couples on Labor Supply and Non-wage Income," Journal of Public Economics 92 (2008): 1698-719.
(19.) See Carasso and Steuerle, "The Hefty Penalty on Marriage Facing."
(20.) Immervoll et al., "An Evaluation of the Tax-Transfer Treatment of Married Couples in European Countries."
(21.) See Michael Farley, "When 'I Do' Becomes 'I Don't': Eliminating the Divorce Loophole to Medicaid Eligibility," The Elder Law Journal 9 (2001): 27-52.
(22.) For a discussion, see Aspen Gorry and Sita Nataraj Slavov, "Marginal and Average Tax Rates in Optimal Tax Policy," Tax Notes (April 1, 2013): 83-88.
About the Author
Aspen Gorry is a research fellow at AEI and an assistant professor at the University of California, Santa Cruz, and Sita Nataraj Slavov is a resident scholar at AEI.
Table 1 Distribution of Household Type 1990 Millions of Percent Households of Total Total Households 92.0 100.00% Total Coupled Households 55.0 59.8% Married-Couple 51.7 56.20% Unmarried-Partner 3.3 3.60% Opposite-Sex Same-Sex Nonpartnered 37.0 40.2% 2000 Millions of Percent Households of Total Total Households 105.5 100.00% Total Coupled Households 60.0 56.90% Married-Couple 54.5 51.70% Unmarried-Partner 5.5 5.20% Opposite-Sex 4.9 4.60% Same-Sex 0.6 0.60% Nonpartnered 45.5 43.10% 2010 Millions of Percent Households of Total Total Households 116.7 100.00% Total Coupled Households 64.3 55.10% Married-Couple 56.5 48.40% Unmarried-Partner 7.7 6.60% Opposite-Sex 6.8 5.90% Same-Sex 0.9 0.80% Nonpartnered 52.5 44.90% Notes: A breakdown of unmarried partner couples is unavailable for 1990. In 1990, same/sex partners who indicated they were married were reclassified as either opposite/sex married couples, siblings, roommates, parent and child (if their age difference was large), or, in a few cases, unmarried partners. From 2000 onward, same/sex partners who indicated they were married were reclassified as "unmarried/partners." (See www2.lib.uchicago.edu//lar1/LIS424/samesex/ unmarried_partners.html for more information.) Sources: US Census Bureau, Examining American Household Composition: 1990 and 2000, table A/9, August 2005, www.census.gov/prod /2005pubs/censr-24.pdf; and US Census Bureau, Households and Families: 2010, table 2, April 2012, www.census.gov/prod/cen2010 /briefs/ c2010br-14.pdf.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ECONOMIC PERSPECTIVES|
|Author:||Gorry, Aspen; Slavov, Sita Nataraj|
|Publication:||AEI Paper & Studies|
|Date:||May 1, 2013|
|Previous Article:||The impact of Dodd-Frank on community banks.|
|Next Article:||Understanding the argument for market valuation of public pension liabilities.|