SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

In this paper we estimate the impacts on income tax collections of legalizing same-sex marriage. We utilize new individual-level data sources to estimate the federal income tax consequences of legalizing same-sex marriages. These data sources also allow us to estimate the impact of legalization on state income tax collections. We find that 23 states would realize a net fiscal benefit from legalization, while 21 states would experience a decline in revenue. The potential (annual) changes in state tax revenue range from negative $29 million in California to positive $16 million in New York. At the federal level, our estimates suggest an overall reduction in revenues, ranging from a potential loss of $187 million to $580 million. Overall, we find that the federal and state impacts are quite modest. We also find that our estimates are only marginally affected by alternative assumptions about how many same-sex couples will choose to marry and which partner will claim any children for tax deduction purposes.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

Gays and lesbians have made major efforts in recent years to be accepted into all aspects of mainstream American life. Many of these efforts have centered on winning the right to marry, and same-sex couples have gone to court in several states seeking this legal right. Indeed, public support for same-sex marriage has increased considerably in the last ten years, and most recent polls suggest that a slight majority of Americans now favor legal recognition, even though groups have also organized to vote down or preemptively to ban state recognition. With the repeal of Section 3 of the Defense of Marriage Act (DOMA) by the U.S. Supreme Court in June 2013 (United States v. Windsor), the federal government now defers to the states regarding the classification of married and spouse. Consequently, the federal government now recognizes same-sex marriages for couples residing in states that have legalized same-sex marriage for federal tax and benefit purposes. By the same token, the federal government does not confer legal recognition of unions for couples in states that do not recognize same-sex marriage. Currently, 13 states and the District of Columbia recognize same-sex marriages;1 many more states continue to explicitly define marriage as occurring between one man and one woman only, and most states do not recognize marriages of same-sex couples from other states. In this paper, we estimate the impacts on federal and state income tax collections of legalizing same-sex marriage.

There are many unresolved—and likely unresolvable—controversies surrounding same-sex marriages, regarding such issues as the definition of marriage, the meaning of family, the notion of morality, the right of privacy, the influence of religion, and the scope of civil rights. There are also various economic issues related to marriage. One such economic consequence of allowing same-sex couples to marry is the potential impact on federal and state tax revenue.2 A change in an individual's filing status from either single or head-of-household to married can sometimes trigger an increase in tax liability, referred to as a marriage penalty or a marriage tax; under other circumstances, it can lead to a decrease in tax liability (a marriage bonus or a marriage subsidy). There have been several attempts to estimate these income tax effects, notably by Alm, Badgett, and Whittington (2000) and the Congressional Budget Office (CBO; 2004), and more recently by Stevenson (2012). Alm, Badgett, and Whittington (2000) constructed a representative household to estimate the federal tax consequences in 1997, and concluded that legalization of same-sex marriage may generate additional revenues to the federal government of as much as $1 billion annually. The CBO (2004) revised this estimate to $400 million annually to account for the Economic Growth and Tax Relief Reconciliation Act (EGTRRA) of 2001, which reduced the marriage tax for many couples. More recently, Stevenson (2012) used individual-level data rather than a representative household to estimate the annual federal revenue impact of legalization before and after the Jobs and Growth Tax Relief Reconciliation Act (JGTRRA) of 2003. Pre-JGTRRA estimates showed a positive impact of legalization that ranged from $118 million to $231 million when labor responsive is considered; post-JGTRRA estimates were reduced to a range of $19 million to $39 million.3

The differences between these various estimates are due to a number of features in the calculations: the differences in time period and the corresponding changes in incomes and tax features over these periods (e.g., 1997 vs. 2001 vs. 2003 vs. 2009); the detail of income and other tax-related information (e.g., a representative household vs. individual-level data, standard deductions vs. itemized deductions); and, especially, the assumptions about the numbers of same-sex couples and about their likely marriage behavior following legalization. Given the sensitivity of any calculations to income, to exemptions and other tax preferences, to the specific features of a tax code that has changed considerably over time, and to the size of the affected population, it is essential that estimates be based on current and accurate measures of these various factors.

Of some note, these previous studies have largely focused on the impact of legalization on changes in federal individual income tax revenues assuming same-sex marriage is legalized in all states. However, an individual state can choose independently to legalize same-sex marriage, changing the extent to which federal government tax revenues are influenced. Given the large differences in political support for legalization across states, it is likely that state legalization, if it does occur, will be done on a case-by-case basis. Given also that state income tax codes vary significantly, coupled with regional variation of same-sex couple characteristics that also varies significantly, the fiscal impact of same-sex marriage on state budgets seems likely to vary dramatically across states. Also, studies that have examined the overall economic impact within a state have typically found a net benefit to the state.4

Calculating the marriage tax/subsidy for heterosexual couples is a surprisingly difficult exercise (Alm & Whttington, 1996). Calculating the income tax consequences for same-sex couples is even more difficult. The calculation requires information on such variables as the number of gay and lesbian individuals, the number in same-sex relationships, the number in relationships who would marry, and, especially, their income and other tax-related characteristics (e.g., deductions, exclusions, and exemptions). Perhaps because of these difficulties, many previous estimates have tended to rely upon average tendencies generated from various surveys, as captured in a representative household. Such estimates are useful and make the calculations straightforward. Even so, this approach makes it difficult to generalize beyond these stylized taxpayers.

However, much has changed in the tax landscape and beyond since many of these estimates were first made, and there are now more reliable and more recent data sources that make calculation of the tax/subsidy at the household-level possible. We utilize these new data sources to estimate the federal income tax consequences of legalizing same-sex marriages. These data sources also allow us to estimate the impact of legalization on state income tax collections. Our estimates therefore provide, for the first time, a comprehensive and current look at the federal and state income tax consequences of legalizing same-sex marriage.

We find that 23 states would realize a net fiscal benefit from legalization, while 21 states would experience a decline in revenue; the remaining seven states do not levy income taxes and consequently would not be affected. Of the 13 states that currently recognize same-sex marriage, four are estimated to experience a tax revenue decline, seven are estimated to experience an increase, and the remaining two states do not collect state income taxes. The potential (annual) changes in state tax revenue for all states range from negative $29 million in California to positive $16 million in New York. At the federal level, our estimates suggest a reduction in federal revenue of $187 million to $580 million. Overall, we find that the federal and state income tax effects are very small. We also find that our estimates are only marginally affected by alternative assumptions about how many same-sex couples will choose to marry and which partner will claim any children for tax deduction purposes.

A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

The individual income tax in the United States was established in 1913, and its treatment of the family has varied over time. In its early years, the basic unit of taxation was the individual, in which each individual was taxed on the basis of his or her income independently of marital status. Because the tax liability did not change much with marriage, the income tax was largely marriage neutral. However, the Revenue Act of 1948 changed the unit of taxation from the individual to the family. With the adoption of income splitting for married couples, couples were now allowed to aggregate and to divide in half their income for federal tax purposes. This change meant that families with equal incomes paid equal taxes. However, because of the progressive tax rates in the income tax, it also meant that a couple's joint tax liability could fall when they married. The Revenue Act of 1948 therefore created the potential for a couple to receive a marriage subsidy.

It was not until the Tax Reform Act of 1969 that a widespread and significant marriage penalty was created for many married couples. The 1969 act established (effective in 1971) a separate tax schedule for single persons that insured a single person would incur a maximum tax liability of 120 percent of a married couple with equal income. Although the tax schedule for married persons was not changed, the act effectively increased the tax liability of some married tax filers relative to single filers, especially for couples whose partners had similar earnings. Marriage now could lead to a substantial increase in income tax liabilities for many couples. The 1969 act therefore generated the potential for a significant marriage tax, even though a potential marriage subsidy still existed for some couples.

Over the last 40 years or so, various tax changes have markedly affected the potential for a marriage tax or subsidy. In particular, the EGTRRA of 2001 and the JGTRRA of 2003 both cut taxes in such a way as to effectively eliminate the marriage penalty for low-income households (Gale & Potter, 2002).5 These measures, combined with the remaining potential marriage subsidies/penalties resulting from the tax system, imply that tax liability changes as a result of marriage are extremely sensitive to assumptions about individual earnings and taxable income. If, for example, same-sex couples comprised one earner or two earners with very unequal incomes, it is likely that they will gain from being able to file as married, resulting in a decrease in government revenues. If instead same-sex couples are comprised largely of equal-earning couples, it is likely that they will pay more when married, resulting in an increase in government revenues.6 However, recent research has found that many same-sex couples do specialize similarly to heterosexual couples, at least when children are present (Black, Sanders, & Taylor, 2007).

DATA AND METHODOLOGY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

We use data from the 2010 American Community Survey (ACS) to estimate the tax consequences of legalization.7 The ACS is an annual survey conducted by the Census Bureau that is a 1-in-100 random representative sample of households of the overall population in the United States. The census interviews all members of the household, obtaining information on each individual's annual income from work, social security, public welfare, investments, and other income sources. The ACS also provides information on the age, race, marital status, gender, and number of dependents for each of the persons living under the same household.

Individuals identify themselves by providing their relationship to the head of household. The ACS compiles information on 2 million individuals, single and married, coupled and living alone. Given that the sample provides a household identifier for each sampled individual, we can identify who in the household is a spouse, a child, a father-in-law, or an unmarried partner. Of special note, when a head of household and an unmarried partner in the same household are identified with the same gender, then the census suggests that they are a same-sex couple living in cohabitation, so same-sex couples can be identified from head of households who claim an unmarried same-sex partner living with them (Black, Sanders, & Taylor, 2000). Note that both the 1990 census and the 2000 census suffered from a biased estimate of the number of same-sex cohabiting couples, resulting from the census miscoding individuals who identified themselves as married to a same-sex individual. This miscoding was corrected in the 2010 census following a procedure suggested by Gates and Steinberger (2010); see also Gates (2010). Note also that we are able to identify who in the ACS has an opposite sex unmarried partner, which means they are heterosexual couples living in cohabitation.

Once these same-sex cohabitating couples are identified, we can use the household weights provided by the ACS to calculate the approximate number of cohabitating same-sex couples in the United States. Similar to the uncorrected census estimates, there is the possibility of measurement error. If couples who appear to be same-sex couples are actually heterosexual couples who have marked the wrong gender when answering the questionnaire, our estimates will be biased. However, the corrected estimation of number of same-sex couples by the census is larger than the raw estimation obtained from the ACS data. If all couples whose marital status has been allocated by the census are dropped from the ACS estimates, the estimations are roughly halved relative to the original estimates and more than halved relative to the census-corrected summary estimates. Given this, it is likely that the estimates obtained without dropping these observations are more appropriate, although we do provide these alternative estimates in our sensitivity analysis, as discussed later.8

Using the household weights provided by the census, we estimate that there are approximately 526,452 same-sex couples living together in the United States, an estimate that is in line with previous estimates (Graham & Barr, 2008). Table 1 describes the estimated (weighted) number of same-sex couples in each state, as well as the percentage of couples in one-earner versus two-earner households. An average of 24 percent of all same-sex couples are single-earner couples, ranging from 5 percent in North Dakota to 41 percent in Arkansas.

Table 1. Number and percent of one- and two-earner same-sex households
 One-earnerTwo-earnerTotal household
 household coupleshousehold couplescouples
StateWeighted numberPercentWeighted numberPercentWeighted numbersize
Alabama876193,847814,72360
Alaska747955931,02911
Arizona3,803299,3347113,137174
Arkansas1,491412,112593,60337
California22,2422857,7937280,0351,088
Colorado1,719159,8158511,534123
Connecticut2,127295,315717,44291
Delaware147101,277901,42431
District of Columbia24382,679922,92240
Florida10,9313222,8296833,760507
Georgia4,0152710,8457314,860190
Hawaii702291,723712,42526
Idaho34929872711,22120
Illinois4,2842314,2067718,490229
Indiana1,674187,720829,394120
Iowa724163,727844,45142
Kansas890242,771763,66144
Kentucky1,125194,744815,86979
Louisiana1,676303,922705,59875
Maine659182,919823,57844
Maryland2,013217,766799,779124
Massachusetts3,7891915,9468119,735246
Michigan3,183269,1997412,382145
Minnesota2,229209,0888011,317110
Mississippi702321,471682,17331
Missouri1,463158,6138510,07698
Montana186151,076851,26212
Nebraska582261,634742,21627
Nevada1,304234,350775,65473
New Hampshire345122,563882,90832
New Jersey3,6492610,4667414,115179
New Mexico1,192263,441744,63362
New York9,8192529,2147539,033476
North Carolina3,7022312,2207715,922188
North Dakota345717957519
Ohio3,7952412,3467616,141208
Oklahoma738163,929844,66750
Oregon1,712216,534798,246101
Pennsylvania4,5152315,2697719,784220
Rhode Island699252,058752,75740
South Carolina1,226273,376734,60268
South Dakota10520414805195
Tennessee2,246285,771728,017107
Texas7,8812031,3518039,232501
Utah464123,410883,87445
Vermont627321,318681,94524
Virginia2,183208,8428011,025137
Washington2,7081812,3638215,071171
West Virginia379251,117751,49622
Wisconsin1,154156,683857,83785
Wyoming282299781273
Nationally124,40324402,04976526,4526,630

Estimating the number of same-sex couples who would marry if same-sex marriage is legalized by the states is not a straightforward procedure. The assumed marriage rate will affect our projected tax revenue/loss. Our initial estimation assumes that the number of same-sex couples who would marry if legally allowed to do so is based simply on the total estimated number of same-sex couples currently living together (or 526,452 couples in our household sample), without considering those who are cohabiting but who choose not to marry; this initial estimation also does not consider those who are not cohabitating and thus not captured by our household sample.

Of course, the number of same-sex couples who would marry following legalization cannot be known. Some recent studies for individual states (Badgett, 2010; Badgett, Sears, & Kukura, 2005a, 2005b) assume a more modest 50 percent marriage rate, based upon observed marriage rates in states following legalization in these states. Accordingly, we test the sensitivity of our initial results by using this alternative assumption of a 50 percent marriage rate.

However, it seems plausible that the federal recognition of same-sex marriages will result in higher marriage rates than when only states recognize same-sex marriage. Federal benefits are more comprehensive, and in some cases result in significantly higher tax and legal benefits. Additionally, the 50 percent estimates are based on the number of couples who marry immediately following legalization or soon after. Given the current uncertainty of state recognition, it is likely that a higher percent of same-sex couples will marry over time as more states legalize same-sex marriages. Furthermore, the overall marriage rate of heterosexuals in the United States is estimated to be between 51 percent (currently married) and 72 percent (ever married; Cohn et al., 2011). If we make similar assumptions with respect to the homosexual population, this suggests that over 1 million same-sex couples will marry, far higher than our assumption that 526,452 couples will marry.

Regardless of these considerations, we include estimates that assume a more conservative 50 percent marriage rate, as noted. We also include estimates that assume that even more couples than those living together choose to marry and estimates that assume that same-sex couples marry at the same rate as heterosexual couples. These alternative scenarios are discussed in our sensitivity analysis.

In order to obtain the most accurate estimate of each person's income tax liability, we gather individual-level information on wage income, business and farm income, social security income, dividends, number of children, property taxes, and mortgage payments.9 Information on annual mean and median earned income by state is in Table 2. The mean and median wages are separated by the higher earner and lower earner in the case of the two-earner households with unequal incomes. Note that the mean earned incomes are considerably higher than the median, suggesting influential outliers and also suggesting that researchers must be wary of estimates based on a single representative agent obtained from national or state averages.10

Table 2. State mean and median annual personal earned income
 Mean earned income ($)Median earned income ($)
  Two-earner couples Two-earner couples
  Unequal earners  Unequal earners 
 One-earnerHighLowEqualOne-earnerHighLowEqual
Statecouplesearnerearnerearnerscouplesearnerearnerearners
Alabama38,60653,13926,82549,0009,50046,00021,00049,000
Alaska100,00081,40537,3160100,00060,00031,4000
Arizona39,44363,84734,48413,85527,10059,00031,30016,000
Arkansas18,96350,97618,357123,00019,00026,20018,000123,000
California70,39190,16739,76533,08650,00068,00031,00025,000
Colorado55,87565,60932,23614,58046,00060,00030,00015,000
Connecticut65,52593,41943,884080,00075,00033,7000
District of Columbia199,749113,29362,52316,500100,000100,00048,50016,500
Delaware6,30584,56532,74131,00086068,00032,00031,000
Florida38,26272,98729,53050,40728,20052,00025,00020,000
Georgia42,49271,54431,83321,07535,00055,00026,00020,000
Hawaii40,321114,87377,42025,00052,00070,00038,90025,000
Idaho21,48136,33618,370012,10031,30017,0000
Illinois56,07091,71438,84630,06640,00068,00033,00020,000
Indiana49,82754,21124,57719,00038,90049,00023,80020,000
Iowa34,56945,98829,37618,00040,00042,00030,00018,000
Kansas26,46351,43823,01918,50023,50042,00019,00018,500
Kentucky29,87752,42127,34614,71625,00042,00025,00014,000
Louisiana43,41574,27927,10115,37536,90055,00021,00020,000
Maine123,09948,28222,51719,73958,00045,00023,00020,000
Maryland57,07892,70542,65320,00050,00083,00046,00020,000
Massachusetts66,75677,42236,03255,19645,00070,00035,00020,000
Michigan34,72673,73934,12125,00016,00052,00028,00025,000
Minnesota55,57472,91837,03717,32124,00060,00034,00019,500
Mississippi65,54552,26220,67215,32315,00048,00017,00016,500
Missouri50,39157,82827,76415,78329,00050,00025,90017,500
Montana2,93232,21219,158090021,6003,5000
Nebraska7,55654,93225,92325,0004,00035,10017,80025,000
Nevada35,94156,19633,17719,00021,90042,00030,00019,000
New Hampshire24,00064,32636,81250,00024,00050,00033,20050,000
New Jersey106,54886,00635,63823,55249,00069,00035,00028,500
New Mexico67,71850,03019,594043,90040,00010,0000
New York77,443109,25738,45736,86740,00075,00030,00020,500
North Carolina34,50759,81925,49217,54230,00049,00025,00023,000
North Dakota33,50051,70321,348033,50050,00021,0000
Ohio48,09057,77527,96915,53430,10050,00022,90014,000
Oklahoma20,25879,76932,99715,49930,00043,00030,00014,500
Oregon37,05970,83831,67431,84830,00052,00028,00037,500
Pennsylvania52,31876,61533,95615,36035,00060,00030,00014,500
Rhode Island66,49685,14030,19833,50050,00064,00022,00033,500
South Carolina36,85856,24821,408032,00034,00014,0000
South Dakota9,00030,18127,28109,00025,00024,0000
Tennessee43,98957,68327,71120,00035,00055,00025,00020,000
Texas50,69273,14334,15121,26936,00059,20030,00022,500
Utah41,47456,96125,741025,80053,00020,0000
Vermont14,71650,19622,344012,80036,50011,0000
Virginia64,79674,29536,78711,38727,00065,00025,0008,000
Washington46,04874,99633,05125,09330,00058,00030,00018,000
West Virginia30,29060,29526,51817,50027,00050,00022,00017,500
Wisconsin34,59452,08435,7614,00021,70043,00028,0004,000
Wyoming25,00024,0003,278025,00024,0003,5000
National54,90275,82933,96226,61735,00060,00028,00019,000

Each person's tax liability in the sample is calculated using the NBER's TAXSIM simulation tool.11 TAXSIM allows us to input data on individuals’ income and other assets, number of dependents, assumed filing status (i.e., single, married, or head of household), whether a filer is 65 years of age or older, state of residence, and the desired tax year. In return, TAXSIM provides the estimated federal and state tax liabilities and their corresponding marginal tax rates for each person for the year specified. When the filing status is set to married, the simulator also includes the spouse's earned income as part of the couple's tax liability estimation.

We calculate each couple's potential marriage tax/subsidy in three steps. First, we use each individual's information to calculate the individual liability before. In cases where the couple reported having children, we allocated those to the person who claimed the child on the census (we also estimate the tax revenue changes assuming the higher earner and, separately, the lower earner claims any children in our sensitivity analysis).12 We then estimate each couple's total liability by adding the liabilities of the two individual filers. Note that we also assume a head of household filing status when possible (e.g., single individuals with dependents), allowing some individuals to lower their potential tax liability even further.

Second, we estimate the each couple's tax liability assuming they were allowed to file as a married couple. Instead of leaving the spouse's income and wages TAXIM field empty, we add it to the calculation and change the filing status to married.

Third, each couple's marriage tax/subsidy is then calculated as the difference between the tax liability as a married couple and the aggregated tax liability found when adding the liabilities as separate filers; a positive number indicates an increase in tax liabilities (e.g., a marriage penalty/tax), and a negative number indicates a marriage bonus/subsidy. The tax revenue consequences at the federal and state levels are obtained by using the corresponding household weights.

As with our sensitivity analysis that examines the impact of different assumptions about the marriage rate of same-sex couples, we also conduct sensitivity analysis on the impact of alternative assumptions about which partner claims any children for tax deduction purposes (e.g., the higher earning partner vs. the lower earning partner). All results are discussed in the next section.

ESTIMATIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

Main Results

We first consider the estimates obtained when we assume that all states legalize same-sex marriage and, separately, the estimated tax revenue implications of the states that have already legalized same-sex marriage. We present estimates at the state level and at the federal level; we also present estimates for one-earner couples separately from two-earner couples. Later we examine several alternative scenarios regarding the sensitivity of child allocation for tax deduction purposes and several in which we make different assumptions about marriage patterns of same-sex couples.

Our state-level estimates using individual-level data when all states legalize same-sex marriage are provided in Table 3 (aggregate estimates) and Table 4 (weighted average across households); we also present estimates when individuals are assumed to take the standard deduction, even when this increases their tax liability.13 New York is estimated to gain the most revenue from legalization, ranging from an increase in tax revenue of $10.8 to $15.8 million; in contrast, California is estimated to lose $28.8 million. In total, state governments experience a decrease in income tax collections, but the aggregate impact is small and negative (−$2.6 million to −$18 million). See also Figure 1 for the state patterns.

Table 3. State income tax revenue effects of same-sex marriage
 One-earnerTwo-earner 
 households ($)households ($)All households ($)
StateMinimizingStandardMinimizingStandardMinimizingStandard
Note
  1. State numbers over $10 million have been rounded to the nearest $100,000.

Alabama12,35224,550222,483292,585195,223277,522
Alaska000000
Arizona−525,967−525,967429,088429,088−486,346−486,346
Arkansas−291,124−320,63477,572−19,098−249,962−376,142
California−21,100,000−21,100,000−3,643,648−3,643,648−28,800,000−28,800,000
Colorado−225,012−421,662890,800−206,320603,156−690,616
Connecticut−80,291−80,291232,736232,736−83,164−83,164
Delaware−61,696−61,69671,50171,501−40,324−40,324
District of Columbia3,4455,370957,816950,046961,260955,416
Florida000000
Georgia−609,512−759,7471,464,2561,898,476557,100810,872
Hawaii−256,576−256,576−105,626−105,626−1,065,178−1,065,178
Idaho−228,579−228,57988,50988,509−142,175−142,175
Illinois−466,711−466,711−49,724−49,724−607,348−607,348
Indiana−127,785−127,785197,698197,69865,94865,948
Iowa−4,186−43,328359,605150,546309,72862,986
Kansas−117,207−165,189166,963−68,275263,577−19,643
Kentucky−182,330−182,33071,12871,128−200,854−200,854
Louisiana−300,152−397,56048,247−360,595−295,318−801,569
Maine−405,758−459,947502,697173,685−74,390−455,932
Maryland−322,100−398,928953,484436,904569,964−23,440
Massachusetts−558,583−558,583654,032654,032−238,464−238,464
Michigan−480,258−480,258−277,388−277,388−481,736−481,736
Minnesota395,067171,2253,512,6641,877,7003,898,3802,039,580
Mississippi−118,521−126,68684,30184,301−77,298−85,463
Missouri−393,825−466,230337,458−366,976280,680−496,160
Montana3,6603,660−11,324−11,324−227,573−233,290
Nebraska−12,424.9−12,424.9358,631183,190272,94397,502
Nevada000000
New Hampshire00−3,120−3,120−3,120−3,120
New Jersey−2,092,298−2,092,2981,363,5321,363,532−835,240−835,240
New Mexico4,937−92,425830,166513,036745,883344,481
New York−4,694,600−5,538,09620,700,00016,600,00015,800,00010,800,000
North Carolina−1,441,207−1,628,9161,240,176−288,264−513,688−2,270,992
North Dakota−3,300−3,300−53,502−53,502−56,802−56,802
Ohio−27,912−27,9124,984,2544,984,2544,933,7724,933,772
Oklahoma−214,865−214,865688,340282,836509,476103,972
Oregon−732,211−724,9921,551,1241,722,856726,684905,640
Pennsylvania−744,449−744,449−266,884−266,884−1,109,220−1,109,220
Rhode Island−133,053−133,053563,032563,03274,30974,309
South Carolina−266,798−310,093700,781337,807328,461−77,808
South Dakota000000
Tennessee−19,077−19,077−25,194−25,194−63,738−63,738
Texas000000
Utah84,04939,204503,114113,872595,347161,259
Vermont−115,988−155,989385,896205,969237,93936,033
Virginia−252,261−478,846543,440−397,420220,460−946,988
Washington000000
West Virginia83,46983,469520,562520,562569,650569,650
Wisconsin−351,396−303,885255,220255,220281,142328,654
Wyoming000000
Total−37,371,034−39,781,83042,074,89529,111,741−2,641,152−18,124,157
Table 4. State income tax revenue effects of same-sex marriage averaged across same-sex couples
 One-earnerTwo-earner 
 households ($)households ($)All households ($)
StateMinimizingStandardMinimizingStandardMinimizingStandard
Alabama14.128.040.055.335.850.9
Alaska0.00.00.00.00.00.0
Arizona−138.3−138.33.53.5−32.0−32.0
Arkansas−195.3−215.016.6−22.3−62.9−94.6
California−950.2−950.2−114.2−114.2−322.9−322.9
Colorado−130.9−245.377.5−25.248.6−55.7
Connecticut−37.7−37.7−0.4−0.4−9.4−9.4
Delaware−419.7−419.713.013.0−22.5−22.5
District of Columbia14.222.1357.5354.6329.0327.0
Florida0.00.00.00.00.00.0
Georgia−151.8−189.295.7128.934.450.0
Hawaii−365.5−365.5−386.5−386.5−381.2−381.2
Idaho−655.0−655.058.658.6−78.0−78.0
Illinois−108.9−108.9−8.8−8.8−30.0−30.0
Indiana−76.3−76.321.721.76.26.2
Iowa−5.8−59.877.726.365.013.2
Kansas−131.7−185.6120.045.964.9−4.8
Kentucky−162.1−162.1−2.8−2.8−26.2−26.2
Louisiana−179.1−237.21.1−90.3−48.0−130.4
Maine−615.7−697.987.31.1−16.7−102.4
Maryland−160.0−198.2104.343.953.9−2.2
Massachusetts−147.4−147.418.518.5−11.3−11.3
Michigan−150.9−150.9−0.1−0.1−33.5−33.5
Minnesota177.276.8357.6190.7324.2169.6
Mississippi−168.8−180.523.723.7−31.7−35.0
Missouri−269.2−318.773.9−3.326.5−46.8
Montana19.719.7−182.4−186.9−156.5−160.4
Nebraska−21.3−21.3132.250.999.635.6
Nevada0.00.00.00.00.00.0
New Hampshire0.00.0−1.1−1.1−1.0−1.0
New Jersey−573.4−573.4104.7104.7−53.4−53.4
New Mexico4.1−77.5168.499.3133.461.6
New York−478.1−564.0629.9503.1372.8255.6
North Carolina−389.3−440.069.3−48.0−30.1−132.9
North Dakota−97.1−97.1−52.9−52.9−54.4−54.4
Ohio−7.4−7.4361.8361.8281.8281.8
Oklahoma−291.1−291.1162.671.698.120.0
Oregon−427.7−423.5201.0224.781.0101.0
Pennsylvania−164.9−164.9−20.3−20.3−49.4−49.4
Rhode Island−190.3−190.372.872.820.920.9
South Carolina−217.6−252.9142.755.760.9−14.4
South Dakota0.00.00.00.00.00.0
Tennessee−8.5−8.5−6.7−6.7−7.1−7.1
Texas0.00.00.00.00.00.0
Utah181.184.5139.433.3144.139.0
Vermont−185.0−248.8232.2126.0110.616.8
Virginia−115.6−219.448.7−48.218.5−79.6
Washington0.00.00.00.00.00.0
West Virginia220.2220.2351.0351.0322.9322.9
Wisconsin−304.5−263.386.086.033.138.6
image

Figure 1. State Income Tax Revenue Effects of Same-Sex Marriage Assuming Individuals Claim Their Own Children and Minimize Their Tax Liabilities.

Download figure to PowerPoint

At the per household level (Table 4), one-earner couples in California on average see the largest decrease in their per household state tax liability upon marriage (or about $950), while those in West Virginia and Utah see the largest increase in tax liability (roughly $200). Two-earner couples fare the best in Hawaii with a reduction in estimated liability of $380, while those in New York would see the largest increase ($630).

The federal impact of legalization is significantly larger (Tables 5 and 6 and Figure 2). These federal tax revenue consequences are broken down by state, but remember that these estimates represent only the federal income tax liabilities, not the state tax liabilities. Our estimates indicate a reduction in federal tax revenue that ranges from −$187 million to −$475 million (Table 5).14 At the household level (Table 6), we estimate an average decrease for same-sex couples to be −$316 when individuals are assumed to minimize their tax liabilities; when individuals are assumed to take the standard deduction, the average change in tax liability becomes −$718.15

Table 5. Federal income tax revenue effects of same-sex marriage
 One-earnerTwo-earner 
 households ($)households ($)All households ($)
StateMinimizingStandardMinimizingStandardMinimizingStandard
Alabama−1,013,807−1,257,773−912,636−2,314,604−1,925,428−3,542,764
Alaska−674,603−674,603556,394−48,260−118,206−722,862
Arizona−4,340,127−5,992,9324,431,928−1,103,896−2,427,408−9,595,008
Arkansas−251,293−251,293547,264−444,896164,774−827,388
California−47,300,000−63,000,00023,700,000−11,500,000−32,000,000−83,000,000
Colorado−3,605,823−4,357,294−168,144−4,538,648−4,129,296−9,251,248
Connecticut−5,488,960−6,758,5843,074,872768,384−4,305,520−7,930,496
Delaware−72,893−72,893787,634162,998963,750324,838
District of Columbia−737,232−959,8683,801,7521,905,1123,064,528945,248
Florida−16,400,000−19,600,0002,480,000−11,900,000−17,500,000−35,500,000
Georgia−5,646,754−7,539,712−852,832−5,467,952−8,767,408−15,200,000
Hawaii−1,271,141−1,350,6313,937,2243,179,208173,680−906,632
Idaho−505,961−660,684−588,048−824,427−1,098,504−1,489,606
Illinois−8,657,170−10,900,0002,863,328−6,564,192−6,414,144−18,100,000
Indiana−1,910,435−2,483,473−536,616−3,381,624−2,524,944−5,942,984
Iowa−2,262,163−2,609,5511,246,942−143,084−1,225,226−2,978,886
Kansas−770,053−1,101,616−441,122−1,333,598−1,180,680−2,404,718
Kentucky−1,745,991−1,868,892396,500−1,247,704−2,412,296−4,179,404
Louisiana−2,851,427−3,458,070−1,018,976−3,216,152−4,116,500−6,920,320
Maine−1,216,686−1,800,496−432,732−1,769,548−2,072,496−3,997,256
Maryland−3,511,675−4,702,2063,522,736−1,326,848−205,600−6,252,256
Massachusetts−3,108,872−5,547,9402,191,648−8,186,208−2,928,768−15,800,000
Michigan−4,337,233−4,993,3513,117,488−2,234,592−2,343,104−8,312,496
Minnesota−2,080,880−3,056,8642,943,552−1,334,064658,224−4,595,376
Mississippi−1,403,765−1,561,470−179,769−562,715−1,793,834−2,334,486
Missouri−2,289,694−2,594,563−773,296−4,995,488−3,985,984−8,513,048
Montana−12,469−12,469−379,333−474,775−904,150−1,089,525
Nebraska−10,441−10,441234,68225,556−32,888−243,682
Nevada−1,614,054−2,103,770811,440−1,698,792−1,113,416−4,115,948
New Hampshire−475,669−475,6691,551,65699,6081,022,880−429,168
New Jersey−6,474,320−13,100,0004,348,768−3,780,496−3,582,240−18,300,000
New Mexico−2,194,540−2,743,318−355,204−1,666,386−3,315,892−5,207,072
New York−20,600,000−28,900,00011,300,000−8,173,376−11,400,000−39,300,000
North Carolina−4,642,740−5,477,600−4,315,144−10,600,000−9,638,480−16,900,000
North Dakota−37,740−37,740−622,702−622,702−660,441−660,441
Ohio−9,294,116−10,100,00027,032−5,829,360−10,800,000−17,700,000
Oklahoma−455,418−455,418103,072−1,531,364−359,624−1,994,060
Oregon−3,696,378−4,374,708−2,129,904−4,125,488−6,168,344−8,806,512
Pennsylvania−7,473,128−9,874,6423,394,368−7,129,968−5,311,520−18,200,000
Rhode Island−1,319,872−1,637,549−236,996−1,816,104−2,912,992−4,796,740
South Carolina−839,057−1,175,631−1,199,896−2,326,496−3,211,672−4,674,848
South Dakota00−301,292−415,290−301,292−415,290
Tennessee−3,599,748−4,182,304−1,745,600−4,103,204−6,858,224−9,798,384
Texas−14,800,000−18,100,0004,694,944−13,800,000−9,945,216−31,800,000
Utah−910,478−1,194,224−1,602,712−3,252,420−2,551,866−4,485,316
Vermont−746,905−770,32861,708−735,982−929,622−1,669,719
Virginia−4,735,366−5,769,244−1,293,120−5,619,872−6,056,560−11,400,000
Washington−3,672,090−4,615,600760,624−7,519,248−4,971,712−13,900,000
West Virginia−462,124−706,259173,377−642,062−449,767−1,509,341
Wisconsin−1,024,843−1,219,5152,814,864−11,3601,919,272−1,101,624
Wyoming−42,805−42,805−126,341−130,176−169,146−172,981
Total−212,588,938−276,233,99269,663,383−154,302,554−187,153,302−475,697,799
Table 6. Federal income tax revenue effects of same-sex marriage averaged across same-sex couples
 One-earnerTwo-earner 
 households ($)households ($)All households ($)
StateMinimizingStandardMinimizingStandardMinimizingStandard
Alabama−1,157.3−1,435.8−199.2−499.2−353.1−649.7
Alaska−9,116.3−9,116.3500.8−43.4−99.8−610.0
Arizona−1,141.2−1,575.8167.7−315.8−159.6−630.8
Arkansas−168.5−168.5167.4−231.741.4−208.0
California−2,127.8−2,830.3228.6−300.1−359.5−931.6
Colorado−2,097.6−2,534.8−49.0−458.1−333.0−746.0
Connecticut−2,580.6−3,177.5175.9−174.21,048.8−895.8
Delaware−495.9−495.9631.3242.2−486.3181.6
District of Columbia−3,033.9−3,950.11,419.1711.1538.7323.5
Florida−1,496.4−1,788.5−40.0−543.5−435.2−881.4
Georgia−1,406.4−1,877.9−256.1−630.4−541.1−939.5
Hawaii−1,810.7−1,924.0690.6212.262.2−324.5
Idaho−1,449.7−1,893.1−402.0−562.4−602.6-817.1
Illinois−2,020.8−2,551.6140.4−450.8−316.5−894.9
Indiana−1,141.2−1,483.6−68.8−387.4−238.1−560.4
Iowa−3,124.5−3,604.4256.5−91.4−257.1−625.0
Kansas−865.2−1,237.8−129.4−410.7−290.6−591.9
Kentucky−1,552.0−1,661.2−102.0−353.7−315.0−545.8
Louisiana−1,701.3−2,063.3−282.8−774.0−669.5−1,125.4
Maine−1,846.3−2,732.2−225.5−578.9−465.3−897.5
Maryland−1,744.5−2,335.9386.6−181.2−19.5−591.8
Massachusetts−820.5−1,464.210.4−590.1−138.7−746.9
Michigan−1,362.6−1,568.8178.3−296.8−163.1−578.6
Minnesota−933.5−1,371.4279.6−157.154.7−382.2
Mississippi−1,999.7−2,224.3−224.3−444.5−734.9−956.4
Missouri−1,565.1−1,773.5−185.7−648.0−376.2−803.4
Montana−67.0−67.0−703.2−849.4−621.8−749.3
Nebraska−17.9−17.9−10.4−108.1−12.0−88.9
Nevada−1,237.8−1,613.3103.5−416.1−181.3−670.3
New Hampshire−1,378.8−1,378.8525.816.3320.1−134.3
New Jersey−1,774.3−3,600.3240.9−433.4−228.8−1,171.5
New Mexico−1,841.1−2,301.4−254.8−559.8−592.9−931.0
New York−2,100.0−2,948.0284.8−317.7−268.5−927.9
North Carolina−1,254.1−1,479.6−373.3−856.6−564.2−991.6
North Dakota−1,110.0−1,110.0−615.9−615.9−632.0−632.0
Ohio−2,449.0−2,656.3−111.4−556.1−618.1−1,011.4
Oklahoma−617.1−617.121.5−345.3−69.2−383.9
Oregon−2,159.1−2,555.3−340.6−610.6−687.7−981.8
Pennsylvania−1,655.2−2,187.1120.4−465.9−236.4−811.8
Rhode Island−1,888.2−2,342.7−559.0−1,108.5−820.8−1,351.6
South Carolina−684.4−958.9−569.0−839.1−595.2−866.4
South Dakota0.00.0−564.2−777.7−471.5−649.9
Tennessee−1,602.7−1,862.1−486.0−837.7−766.3−1,094.8
Texas−1,883.0−2,299.8140.0−390.1−232.1−741.4
Utah−1,962.2−2,573.8−447.6−897.5−617.7−1,085.8
Vermont−1,191.2−1,228.6−119.9−590.2−432.2−776.3
Virginia−2,169.2−2,642.8−136.0−581.4−509.1−959.7
Washington−1,356.0−1,704.4−94.1−674.5−301.1−843.4
West Virginia−1,219.3−1,863.58.9−579.8−255.0−855.6
Wisconsin−888.1−1,056.8400.516.0225.7−129.5
Wyoming−1,528.8−1,528.8−1,276.2−1,314.9−1,331.9−1,362.1
image

Figure 2. Federal Income Tax Revenue Effects of Same-Sex Marriage Assuming Individuals Claim Their Own Children and Minimize Their Tax Liabilities.

Download figure to PowerPoint

There is considerable variation across states in the federal income tax consequences. On average, same-sex couples in Connecticut will pay slightly over $1,000 more in federal taxes if they choose to wed, while couples in Pennsylvania will pay about $800 less. If we consider one-earner and two-earner couples separately, one-earner couples will pay as much as $9,100 less in Alaska, but they will pay roughly the same amount in Nebraska before and after marriage.16 Similarly, two-earner couples will see the largest drop in federal tax liability in South Carolina (−$570) and the largest increase in the District of Columbia ($1,400)

Unlike differences in state-imposed state income taxes, federal taxes must treat all individuals the same, so that differences in the per household estimates are a result of differences in family structure, income, and deductions. For same-sex couples in particular, the differential appears to be largely driven by differences in income. In Alaska, the average one-earner household earns $100,000 while the average one-earner household in Nebraska earns only $7,500. Consequently, we estimate that one-earner couples in Nebraska will not experience the large reduction in tax liability upon marriage that one-earner same-sex couples in Alaska will experience. Similarly, in South Carolina, the higher earner of a two-earner household earns $56,000 on average while the lower earner earns an average of $21,000. In the District of Columbia, on the other hand, the higher earner earns $113,000 on average while the lower earner earns an average of $62,000. Averaged across same-sex households, those in South Carolina will face a lower tax burden upon marriage due to the low wages of the lower earning spouse.

Note that some care must be taken when we interpret tax revenue changes in states with few (unweighted) observations. For example, in Wyoming there are only three same-sex couples that are used to estimate the general population of same-sex couples. States with low numbers of same-sex couple samples include Alaska, North and South Dakota, Montana, and Wyoming. Eliminating the influence of these states does not change our general findings. However, in most states, the number of unweighted observations is quite large. For example, in California there are 2,176 observations, in Texas there are 1,002 observations, and in Florida there are 1,104 observations; the average number of observations (other than in the outlier states) is 276.17

The estimated spatial distribution of the marriage tax for same-sex couples in the contiguous 48 states is given in Figures 3 and 4, which depict the weighted average change in households’ state income tax liability and federal income tax liability, respectively (averaged across households in each state).

image

Figure 3. State Income Tax Revenue Effects of Same-Sex Marriage Averaged Across Same-Sex Couples.

Download figure to PowerPoint

image

Figure 4. Federal Income Tax Revenue Effects of Same-Sex Marriage Averaged Across Same-Sex Couples.

Download figure to PowerPoint

Given the recent Supreme Court ruling that grants same-sex couples who are in marriages recognized by their state to have their marriage recognized by the federal government, we also estimate the tax liability changes for states that currently recognize same-sex marriage or civil unions. These are a subset of our previous estimations, and are found in Table 7. Our estimations suggest that the federal income tax revenues will decline by approximately $74.2 million as a result of the recent ruling.

Table 7. Federal and state income tax revenue effects of same-sex marriage for states that currently recognize same-sex unions
 Federal tax revenueState tax revenue
StateMinimizingStandardMinimizingStandard
California−32,000,000−83,000,000−28,800,000−28,800,000
Colorado−4,129,296−9,251,248603,156−690,616
Connecticut−4,305,520−7,930,496−83,164−83,164
Delaware963,750324,838−40,324−40,324
Hawaii173,680−906,632−1,065,178−1,065,178
Illinois−6,414,144−18,100,000−607,348−607,348
Iowa−1,225,226−2,978,886309,72862,986
Maine−2,072,496−3,997,256−74,390−455,932
Maryland−205,600−6,252,256569,964−23,440
Massachusetts−2,928,768−15,800,000−238,464−238,464
Minnesota658,224−4,595,3763,898,3802,039,580
New Hampshire1,022,880−429,168−3,120−3,120
New Jersey−3,582,240−18,300,000−835,240−835,240
New York−11,400,000−39,300,00015,800,00010,800,000
Rhode Island−2,912,992−4,796,74074,30974,309
Vermont−929,622−1,669,719237,93936,033
Washington−4,971,712−13,900,00000

Sensitivity Analysis

We consider several potential scenarios that may alter our aggregate estimates. It is possible that a couples’ tax liability may be lowered if someone other than the person who claimed a child on the census claims the child for tax deduction purposes.18 We re-estimate the tax consequences assuming that the higher earning partner and, separately, the lower earning partner claims any present children as dependents. The changes in aggregate state and federal tax liabilities are quite modest and can be found in Tables 8 and 9. The estimated state income tax change increases a bit when the higher earner claims any children (ranging from $2.7 million to −$11.7 million), and decreases when the lower earner claims any children (−$10.3 million to −$26.3 million).

Table 8. State income tax revenue effects of same-sex marriage under different assumptions on allocation of children/dependents
 Claiming ownHighest earnerLowest earner
 childrenclaims dependentsclaims dependents
StateMinimizingStandardMinimizingStandardMinimizingStandard
Alabama195,223277,522176,112248,390158,421244,589
Alaska000000
Arizona−486,346−486,346−292,644−292,644−671,044−671,044
Arkansas−249,962−376,142−191,887−318,068−340,859−467,039
California−28,800,000−28,800,000−25,800,000−25,800,000−33,400,000−33,400,000
Colorado603,156−690,616604,676−651,220651,300−802,992
Connecticut−83,164−83,164−33,904−33,90495,18095,180
Delaware−40,324−40,324−40,324−40,324−40,550−40,550
District of Columbia961,260955,416961,260955,416884,228878,380
Florida000000
Georgia557,100810,872688,416994,016161,316361,752
Hawaii−1,065,178−1,065,178−1,047,788−1,047,788−1,060,412−1,060,412
Idaho−142,175−142,175−105,242−105,242−187,648−187,648
Illinois−607,348−607,348−576,060−576,060−650,876−650,876
Indiana65,94865,948−18,566−18,56691,12291,122
Iowa309,72862,986183,4929,846314,67155,698
Kansas263,577−19,643271,179−3,237557,080234,554
Kentucky−200,854−200,854−191,814−191,814−189,232−189,232
Louisiana−295,318−801,569−383,061−850,178−292,955−808,877
Maine−74,390−455,932−74,390−455,932−289,048−738,768
Maryland569,964−23,440387,612−94,556612,944−51,712
Massachusetts−238,464−238,464−287,808−287,808−744,528−744,528
Michigan−481,736−481,736−517,748−517,748−389,636−389,636
Minnesota3,898,3802,039,5803,801,6801,968,3643,628,1001,648,024
Mississippi−77,298−85,463−54,796−62,961−118,043−126,958
Missouri280,680−496,160226,652−467,330214,390−595,488
Montana−227,573−233,290−227,573−233,290−227,573−233,290
Nebraska272,94397,502271,702130,668347,84579,452
Nevada000000
New Hampshire−3,120−3,120−3,120−3,120−3,120−3,120
New Jersey−835,240−835,240−503,240−503,240−2,161,616−2,161,616
New Mexico745,883344,481801,825415,770655,430184,056
New York15,800,00010,800,00016,700,00012,000,00016,200,00011,600,000
North Carolina−513,688−2,270,992−125,816−1,744,664−823,592−2,724,888
North Dakota−56,802−56,802−42,161−42,161−80,731−81,103
Ohio4,933,7724,933,7725,044,7045,044,7044,836,8524,836,852
Oklahoma509,476103,972536,996185,312412,308−48,650
Oregon726,684905,640721,904919,416546,128750,084
Pennsylvania−1,109,220−1,109,220−472,500−472,500−1,063,332−1,063,332
Rhode Island74,30974,30947,81347,81331,16631,166
South Carolina328,461−77,808389,533−16,73678,852−343,621
South Dakota000000
Tennessee−63,738−63,738−63,738−63,738−63,738−63,738
Texas000000
Utah595,347161,259621,176161,259641,716222,564
Vermont237,93936,033199,08358,730237,93936,033
Virginia220,460−946,988220,128−947,312296,756−891,804
Washington000000
West Virginia569,650569,650569,650569,650597,893597,893
Wisconsin281,142328,654343,888391,400208,606256,118
Wyoming000000
Total−2,641,152−18,124,1572,719,616−11,741,387−10,300,000−26,337,406
Table 9. Federal income tax revenue effects of same-sex marriage under different assumptions regarding allocation of children/dependents
 Claiming ownHighest earnerLowest earner
 childrenclaims dependentsclaims dependents
StateMinimizingStandardMinimizingStandardMinimizingStandard
Alabama−1,925,428−3,542,764−1,771,828−3,188,728−2,318,088−4,012,808
Alaska−118,206−722,862116,280−488,374−118,206−722,862
Arizona−2,427,408−9,595,008−1,070,112−7,812,176−3,806,800−11,400,000
Arkansas164,774−827,388208,956−719,424−455,072−1,441,352
California−32,000,000−83,000,000−30,000,000−73,900,000−40,100,000−97,600,000
Colorado−4,129,296−9,251,248−3,840,576−8,775,872−5,873,200−11,900,000
Connecticut−4,305,520−7,930,496−2,802,960−5,757,792−4,780,544−9,643,392
Delaware963,750324,838850,040311,1381,337,874573,202
District of Columbia3,064,528945,2482,961,288895,6723,064,528945,248
Florida−17,500,000−35,500,000−14,500,000−31,000,000−22,200,000−41,200,000
Georgia−8,767,408−15,200,000−7,148,448−12,800,000−11,600,000−18,900,000
Hawaii173,680−906,632470,520−345,896−6,560−1,086,872
Idaho−1,098,504−1,489,606−979,219−1,291,072−1,252,199−1,643,301
Illinois−6,414,144−18,100,000−6,018,464−16,500,000−8,934,272−21,500,000
Indiana−2,524,944−5,942,984−1,437,792−4,274,928−4,029,848−8,226,672
Iowa−1,225,226−2,978,886−1,106,894−2,161,714−1,624,210−3,378,814
Kansas−1,180,680−2,404,718−1,116,284−2,232,502−2,876,266−4,255,318
Kentucky−2,412,296−4,179,404−2,536,344−4,263,764−4,040,588−6,354,832
Louisiana−4,116,500−6,920,320−3,263,772−5,814,080−4,874,452−7,756,256
Maine−2,072,496−3,997,256−2,072,496−3,997,256−2,790,728−5,776,404
Maryland−205,600−6,252,256−258,336−5,103,296−3,335,056−10,100,000
Massachusetts−2,928,768−15,800,000−1,622,176−13,200,000−5,227,328−20,700,000
Michigan−2,343,104−8,312,496−2,193,280−6,448,512278,848−6,235,888
Minnesota658,224−4,595,376748,688−4,440,0801,104,112−4,823,056
Mississippi−1,793,834−2,334,486−1,605,544−2,126,666−2,990,526−3,587,040
Missouri−3,985,984−8,513,048−3,476,272−7,319,616−5,483,456−10,100,000
Montana−904,150−1,089,525−904,150−1,089,525−904,150−1,089,525
Nebraska−32,888−243,68216,122−51,726400,636−205,328
Nevada−1,113,416−4,115,948−231,728−3,085,232−1,884,716−5,114,272
New Hampshire1,022,880−429,1681,132,436−313,692880,524−748,412
New Jersey−3,582,240−18,300,000−2,313,184−15,900,000−5,837,488−22,400,000
New Mexico−3,315,892−5,207,072−2,980,720−4,761,484−4,156,032−6,704,376
New York−11,400,000−39,300,000−7,707,136−33,400,000−14,200,000−47,300,000
North Carolina−9,638,480−16,900,000−7,016,640−13,600,000−12,400,000−20,600,000
North Dakota−660,441−660,441−323,086−323,086−962,519−965,233
Ohio−10,800,000−17,700,000−8,873,632−15,300,000−12,400,000−19,900,000
Oklahoma−359,624−1,994,060172,044−1,226,460−551,100−2,528,588
Oregon−6,168,344−8,806,512−5,940,256−8,428,344−6,036,264−9,709,648
Pennsylvania−5,311,520−18,200,000−3,079,072−15,400,000−7,382,848−21,100,000
Rhode Island−2,912,992−4,796,740−2,937,592−4,706,168−2,881,184−5,004,568
South Carolina−3,211,672−4,674,848−2,858,000−4,321,180−4,136,660−5,703,008
South Dakota−301,292−415,290−282,070−278,942−341,942−455,940
Tennessee−6,858,224−9,798,384−5,322,936−8,159,512−9,091,024−12,500,000
Texas−9,945,216−31,800,000−7,108,608−27,100,000−18,000,000−42,300,000
Utah−2,551,866−4,485,316−1,940,782−3,738,600−3,572,504−6,169,084
Vermont−929,622−1,669,719−413,921−942,105−929,622−1,669,719
Virginia−6,056,560−11,400,000−5,736,656−10,900,000−9,587,552−16,000,000
Washington−4,971,712−13,900,000−2,676,656−11,000,000−8,060,016−17,600,000
West Virginia−449,767−1,509,341−491,732−1,454,341−488,761−1,568,971
Wisconsin1,919,272−1,101,6242,455,448−240,4961,538,824−1,579,512
Wyoming−169,146−172,981−169,146−172,981−169,146−172,981
Total−187,153,302−475,697,799−149,041,152−408,601,088−254,085,581−579,915,582

As previously noted, the assumed marriage rate following legalization is unknown. Similar to Badgett, Sears, and Kukura (2005a, 2005b) and Badgett (2010), we have also estimated the revenue impacts when we assume that only 50 percent of cohabiting same-sex couples marry if allowed (rather than assuming that all cohabiting couples marry if allowed). This alternative assumption halves our estimates to an even more modest reduction of $1.3 million to $9 million for states and −$95 million to −$237 million for the federal government. Similarly, we have estimated the revenue impacts if even more couples than those living together choose to marry, say, 75 percent of couples choose to marry, again with minor impacts on our estimates. Other scenarios are of course possible. It is estimated that there are 4 million gays and lesbians in the United States (Gates & Steinberger, 2010). If same-sex couples marry at the same rate as heterosexuals, or 51 percent, according to the Census Bureau (2011), then there would be 960,000 same-sex married couples. This is almost double our current estimates, suggesting that state revenues may decrease by as much as $36 million and that federal revenues would have a potential annual loss of almost $1 billion.19

It is also plausible that couples who face large penalties may choose to marry at a relatively lower rate than couples who do not face large penalties. The range of subsidies/penalties faced by households at the state level and federal level are provided in Table 10. At the high end, some couples face a federal tax liability increase of over $15,000 in Massachusetts while some couples in Delaware would face a federal tax liability reduction of almost $19,000. At the state level, the largest reduction in state tax liability is almost $6,000 for some couples in Hawaii, and the largest increase in state tax liability is over $12,000 for some couples in New York.

Table 10. Summary statistics of federal and state marriage tax/subsidy
 Federal marriage tax/State marriage tax/
 subsidy by statesubsidy by state
 WeightedStandard  WeightedStandard  
StatemeandeviationMinimumMaximummeandeviationMinimumMaximum
Alabama−353.11,314.7−7,468.42,994.835.8125.0−252.8388.5
Alaska−99.82,622.7−9,116.33,830.20.00.00.00.0
Arizona−159.61,970.0−8,670.09,135.9−32.0201.6−1,204.8427.8
Arkansas41.42,434.0−4,790.012,643.0−62.9209.5−479.5371.7
California−359.53,069.7−12,906.613,745.4−322.9903.6−4,161.42,004.7
Colorado−333.01,867.6−7,335.09,272.148.6247.0−1,549.9388.9
Connecticut1,048.84,163.8−7,533.314,115.0−9.4576.2−3,035.01,050.0
Delaware−486.33,744.4−18,963.611,892.3−22.5279.6−787.7271.1
District of Columbia538.71,597.0−2,981.04,286.1329.0200.8−258.8990.0
Florida−435.22,248.4−11,745.412,904.50.00.00.00.0
Georgia−541.12,001.8−12,930.05,308.034.4330.3−3,384.0522.3
Hawaii62.25,470.3−17,830.610,489.5−381.21,122.1−5,724.1359.5
Idaho−602.61,378.9−5,624.5431.3−78.0469.6−1,942.5386.1
Illinois−316.52,456.1−9,495.79,300.8−30.0124.6−1,200.0182.8
Indiana−238.11,890.6−6,890.06,167.66.2106.9−428.7453.2
Iowa−257.12,051.4−7,682.53,956.165.0208.1−1,343.9490.9
Kansas−290.6885.1−3,084.02,668.664.9385.1−816.6954.6
Kentucky−315.01,766.0−11,486.54,432.6−26.2345.1−3,075.8418.8
Louisiana−669.52,426.8−8,925.06,283.8−48.0187.0−845.7237.0
Maine−465.31,317.9−5,824.92,908.1−16.7496.2−1,531.0994.5
Maryland−19.52,870.5−9,493.812,152.553.9311.1−1,714.81,304.2
Massachusetts−138.72,617.7−12,687.215,020.5−11.3242.6−1,584.7665.0
Michigan−163.12,190.4−9,625.08,984.7−33.5266.5−1,735.71,152.0
Minnesota54.71,879.8−5,358.69,908.5324.2509.2−924.72,355.0
Mississippi−734.92,398.4−9,016.05,775.3−31.7188.6−519.0215.0
Missouri−376.21,935.9−9,625.010,932.626.5332.0−1,245.01,859.0
Montana−621.81,798.7−11,142.9844.6−156.5624.9−3,750.7174.3
Nebraska−12.01,894.4−3,522.06,693.799.6393.8−763.01,145.7
Nevada−181.31,768.6−8,403.83,832.10.00.00.00.0
New Hampshire320.11,991.3−9,817.08,595.1−1.010.8−120.00.0
New Jersey−228.82,577.0−13,168.88,236.3−53.4614.8−2,339.01,492.5
New Mexico−592.91,986.2−8,971.14,850.9133.4320.4−799.0703.2
New York−268.53,202.5−17,864.614,178.5372.81,515.1−4,589.512,579.0
North Carolina−564.22,062.9−6,090.79,483.9−30.1387.6−2,197.5749.5
North Dakota−632.01,006.2−3,147.50.0−54.463.0−164.70.0
Ohio−618.12,374.9−16,842.84,331.3281.8446.2−1,708.31,676.0
Oklahoma−69.21,072.1−3,386.31,931.398.1299.9−749.0902.5
Oregon−687.72,127.2−10,065.04,106.481.0630.4−4,313.61,458.4
Pennsylvania−236.41,934.5−8,275.58,836.3−49.4193.8−1,074.5184.2
Rhode Island−820.81,972.4−9,625.02,424.620.9512.4−1,266.01,145.8
South Carolina−595.21,743.0−6,132.74,431.360.9355.2−654.5713.6
South Dakota−471.5801.4−1,378.8505.90.00.00.00.0
Tennessee−766.32,019.7−10,263.83,135.2−7.120.7−75.00.0
Texas−232.12,302.0−14,438.111,056.00.00.00.00.0
Utah−617.72,109.4−8,420.25,221.9144.1310.2−669.2892.5
Vermont−432.22,182.7−9,408.79,555.1110.6530.2−2,136.51,408.1
Virginia−509.12,099.5−6,919.19,915.618.5215.2−1,249.5477.6
Washington−301.11,967.9−9,499.44,269.30.00.00.00.0
West Virginia−255.01,638.1−4,291.54,431.3322.9541.8−1,111.51,125.0
Wisconsin225.71,611.2−4,294.87,532.033.1403.2−1,235.91,709.7
Wyoming−1,331.9112.4−1,528.8−1,187.00.00.00.00.0

If we assume that couples who would face a federal tax liability increase of over $5,000 would not marry if allowed, then both federal and state income tax revenues would experience an aggregated reduction. The reduction in federal income tax revenue would be approximately $303.5 million (a drop of an additional $116.4 million), while the reduction in aggregated state tax revenues would be around $15.7 million (an additional $5.4 million), summarized in Table 11, columns 2 and 3. Column 1 in Table 11 details the marriage tax of the 95th percentile for couples in every state. For the majority of states this falls well below the $5,000 mark, suggesting that the majority of couples would still marry if indeed an increase of $5,000 would entice these couples not to marry. If a change in tax liability less than $5,000 would deter marriage, the net negative influence on federal income taxes would increase.

Table 11. Income tax revenue effects of same-sex marriage assuming that couples with marriage tax greater than $5,000 do not marry
 Federal marriage taxFederal revenueState revenue
Statefor 95th percentileincrease/decreaseincrease/decrease
Alabama1,138−1,925,427195,223
Alaska3,830−118,2070
Arizona1,992−4,256,205−471,370
Arkansas1,931−1,327,106−249,961
California4,194−68,400,000−30,800,000
Colorado1,257−5,375,852587,851
Connecticut3,119−7,451,40363,310
Delaware4,286963,749−40,324
District of Columbia11,296−852,308799,331
Florida2,044−23,400,0000
Georgia1,502−9,192,033548,000
Hawaii10,489−4,014,969−1,074,058
Idaho256−1,098,504−142,175
Illinois3,258−9,864,439−607,349
Indiana2,038−4,031,91465,949
Iowa3,744−1,225,228309,728
Kansas773−1,180,680263,577
Kentucky1,042−2,412,298−200,854
Louisiana1,524−4,594,065−276,207
Maine675−2,072,495−74,390
Maryland4,678−4,055,935335,664
Massachusetts2,639−8,095,727−228,518
Michigan4,131−4,357,306−481,736
Minnesota2,265−897,4253,828,098
Mississippi600−2,140,350−86,598
Missouri895−5,144,843300,057
Montana845−904,150−227,573
Nebraska6,694−1,130,66085,053
Nevada2,155−1,113,4190
New Hampshire2,359300,888−3,120
New Jersey3,728−5,712,219−946,291
New Mexico1,501−3,315,889745,883
New York5,601−31,300,0005,355,809
North Carolina972−12,500,000−679,842
North Dakota0−660,441−56,802
Ohio989−10,800,0004,933,772
Oklahoma1,931−359,622509,476
Oregon2,494−6,168,349726,684
Pennsylvania1,733−6,559,777−1,109,222
Rhode Island757−2,912,99374,309
South Carolina1,684−3,211,670328,461
South Dakota506−301,2920
Tennessee1,487−6,858,221−63,738
Texas2,276−16,400,0000
Utah1,190−2,880,848589,323
Vermont1,829−1,388,267170,350
Virginia1,158−7,987,146208,759
Washington2,305−4,971,7090
West Virginia3,086−449,767569,650
Wisconsin2,368721,685440,397
Wyoming−1,187−169,1460
Total −303,508,672−15,785,414
Table 12. Estimated number of same-sex couple households from different data sources
 Census 2010ACS 2010 weightedACS 2010 weighted without
Statecorrected totalstotalsaallocated statusb
  1. a

    Estimated using the household weights provided by the ACS 2010. These numbers are comparable to those estimated for the ACS 2010 by the Census Bureau, and fall within its estimated margin of error.

  2. b

    These are obtained by dropping all those couples for whom the census has allocated their marital status, and using household weights.

Alabama6,5825,4532,906
Alaska1,2281,185705
Arizona15,81715,21011,800
Arkansas4,2263,9772,226
California98,15389,11556,121
Colorado12,42412,4019,159
Connecticut7,8528,8534,966
Delaware2,6461,7891,103
District of Columbia4,8222,9222,707
Florida48,49640,28129,428
Georgia21,31816,20211,227
Hawaii3,2392,7941,417
Idaho2,0421,8231,090
Illinois23,04920,26513,804
Indiana11,07410,6048,543
Iowa4,0934,7662,816
Kansas4,0094,0633,352
Kentucky7,1957,6575,276
Louisiana8,0766,1494,468
Maine3,9584,4543,334
Maryland12,53810,5657,302
Massachusetts20,25621,12310,524
Michigan14,59814,36710,311
Minnesota10,20712,0259,045
Mississippi3,4842,4411,442
Missouri10,55710,5968,549
Montana1,3481,454529
Nebraska2,3562,7401,776
Nevada7,1406,1404,417
New Hampshire3,2603,1951,464
New Jersey16,87515,65510,267
New Mexico5,8255,5934,528
New York48,93242,32629,421
North Carolina18,30917,08311,808
North Dakota5591,045564
Ohio19,68417,50813,806
Oklahoma6,1345,1943,887
Oregon11,7738,9706,492
Pennsylvania22,33622,46614,520
Rhode Island2,7853,5492,718
South Carolina7,2145,3963,794
South Dakota714639185
Tennessee10,8988,9506,921
Texas46,40142,84332,469
Utah3,9094,1312,893
Vermont2,1432,1511,176
Virginia14,24311,8977,462
Washington19,00316,51412,817
West Virginia2,8481,7641,160
Wisconsin9,1798,5056,343
Wyoming65712799
United States646,464586,915405,137
Table 13. Estimated income tax revenue effects using each state's average marriage tax/subsidy and number of same-sex couples from different sources
 Federal revenueState revenue
   ACS 2010  ACS 2010
   weighted  weighted
 Census 2010ACS 2010withoutCensus 2010ACS 2010without
 correctedhouseholdallocatedcorrectedhouseholdallocated
Statetotalsweightsstatusatotalsweightsstatusa
  1. a

    These are obtained by dropping all those couples for whom the Census has allocated their marital status, and using household weights.

Alabama−2,324,071−1,925,428−1,026,094235,643195,223104,038
Alaska−122,497−118,206−70,326000
Arizona−2,524,275−2,427,408−1,883,192−505,756−486,346−377,310
Arkansas175,089164,77492,226−265,612−249,962−139,908
California−35,288,713−32,000,000−20,177,048−31,688,706−28,800,000−18,118,670
Colorado−4,136,944−4,129,296−3,049,764604,277603,156445,474
Connecticut8,234,981−4,305,5205,208,217−73,761−83,164−46,650
Delaware−1,286,842963,750−536,428−59,641−40,324−24,862
District of Columbia2,597,6513,064,5281,458,2831,586,312961,260890,532
Florida−21,107,549−17,500,000−12,808,334000
Georgia−11,535,820−8,767,408−6,075,272733,009557,100386,035
Hawaii201,341173,68088,083−1,234,829−1,065,178−540,214
Idaho−1,230,469−1,098,504−656,813−159,255−142,175−85,009
Illinois−7,295,320−6,414,144−4,369,152−690,786−607,348−413,711
Indiana−2,636,861−2,524,944−2,034,19868,87265,94853,131
Iowa−1,052,215−1,225,226−723,928265,992309,728183,003
Kansas−1,164,988−1,180,680−974,068260,074263,577217,453
Kentucky−2,266,747−2,412,296−1,662,176−188,735−200,854−138,397
Louisiana−5,406,546−4,116,500−2,991,140−387,866−295,318−214,585
Maine−1,841,701−2,072,496−1,551,347−66,106−74,390−55,684
Maryland−244,010−205,600−142,109676,406569,964393,932
Massachusetts−2,808,549−2,928,768−1,459,181−228,674−238,464−118,807
Michigan−2,380,786−2,343,104−1,681,620−489,481−481,736−345,735
Minnesota558,701658,224495,0973,309,0053,898,3802,932,297
Mississippi−2,560,311−1,793,834−1,059,693−110,326−77,298−45,663
Missouri−3,971,317−3,985,984−3,215,951279,647280,680226,456
Montana−838,236−904,150−328,952−210,983−227,573−82,797
Nebraska−28,280−32,888−21,318234,691272,943176,915
Nevada−1,294,758−1,113,416−800,973000
New Hampshire1,043,6871,022,880468,699−3,183−3,120−1,430
New Jersey−3,861,395−3,582,240−2,349,330−900,331−835,240−547,775
New Mexico−3,453,433−3,315,892−2,684,489776,823745,883603,855
New York−13,137,293−11,400,000−7,898,96818,243,42015,800,00010,969,093
North Carolina−10,330,214−9,638,480−6,662,252−550,559−513,688−355,072
North Dakota−353,289−660,441−356,449−30,385−56,802−30,657
Ohio−12,167,306−10,800,000−8,533,9285,546,9714,933,7723,890,545
Oklahoma−424,705−359,624−269,128601,680509,476381,273
Oregon−8,095,872−6,168,344−4,464,317953,763726,684525,935
Pennsylvania−5,280,787−5,311,520−3,432,890−1,102,804−1,109,220−716,901
Rhode Island−2,285,907−2,912,992−2,230,91458,31274,30956,910
South Carolina−4,293,734−3,211,672−2,258,168439,125328,461230,945
South Dakota−336,655−301,292−87,228000
Tennessee−8,350,938−6,858,224−5,303,436−77,611−63,738−49,288
Texas−10,771,180−9,945,216−7,537,110000
Utah−2,414,729−2,551,866−1,787,110563,353595,347416,930
Vermont−926,164−929,622−508,245237,053237,939130,086
Virginia−7,250,868−6,056,560−3,798,777263,933220,460138,276
Washington−5,721,047−4,971,712−3,858,689000
West Virginia−726,155−449,767−295,765919,707569,650374,600
Wisconsin2,071,3711,919,2721,431,387303,422281,142209,675
Wyoming−875,032−169,146−131,854000
United States−201,521,685−187,153,302−124,506,127−1,863,901−2,641,1521,488,264

However, given the large size of the increased tax liability faces by these couples in particular, imposing this restriction does change the number of states that contribute a net positive change to the federal income tax. For the original estimates, Arkansas, the District of Columbia, Delaware, Hawaii, Minnesota, New Jersey, and Wisconsin had couples paying a net positive average increase in federal income tax. Dropping couples whose federal income tax liability increased by more than $5,000, eliminates Arkansas, the District of Columbia, Hawaii, and Minnesota from this statistic.

The estimated change in state income tax revenue if these couples are assumed to not marry is detailed in column 3 in Table 11. All 23 states that were previously estimated to experience a net increase in revenues are still estimated to experience an increase.20 New York is still the state with the largest estimated increase ($5.3 million), but this is roughly one-third of the original estimated increase ($15.8 million). California, the state that is estimated to lose the most state income tax revenue, would experience an additional drop of approximately $2 million, adjusting the aggregated net decline in state income tax revenues to approximately $30.8 million.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

The income taxes paid by same-sex couples are one of many potential effects of legalizing same-sex marriage. We provide updated estimates of potential federal government revenue changes using current and individual-level data. We also consider the potential impacts on state government income tax collections of legalization.

Overall, we find quite modest impacts on state budgets, on federal collections, and on per household liabilities. In total, states are estimated to gain on an annual basis between $6.7 million and $23.7 million in additional state income tax under the most plausible set of assumptions, but may generate as much as $48 million or as little as $3 million. New York is projected to experience the largest increase in additional tax revenues ($12 million), while California will experience the least (−$21 million). Same-sex couples in Ohio will pay the largest state per household marriage penalty ($380), while couples in California will receive the largest subsidy (−$272). In all, we estimate that 31 states will gain additional income tax revenues, 12 will lose, and seven will experience no change. At the federal level, our estimates range from an annual gain of $5.7 million to an annual loss of −$315.8 million. For both the state and federal tax liability changes, the distribution of the effects varies significantly across the states. Again, however, the effects are small.

Our estimates are of course dependent upon the many assumptions that we necessarily made: on the number of gay and lesbian individuals, the number in relationships, the number in relationships who would marry, their labor supply responses, and their income and use of tax preferences. We have considered several alternative scenarios, and generally found our base estimates to be only marginally affected. Behavioral changes, if any, will also influence our estimations. For example, if as a result of marriage same-sex couples have more children, then this will increase the available deductions, decrease federal income tax revenue, and also reduce any projected gain to many states. Similarly, an increased probability of homeownership by married same-sex households will increase deductions and reduce taxable income further. Although marriage taxes/subsidies have been shown to have a small impact on the probability of marriage (Alm & Whittington, 1997), it is possible that those couples who would see an increase in their tax liability will choose to remain unmarried.

Even though these revenue effects in the aggregate are typically modest, individual couples may still experience large changes in their income tax liabilities simply from a change in legal marital status. Any such changes in taxes do not need to occur, and are entirely dependent on how the tax law defines the unit of taxation (Alm, Dickert-Conlin, & Whittington, 1999). An income tax system that defines the unit as the family will exhibit the types of changes in tax liabilities that we have calculated here. In contrast, in a tax system that makes the individual the unit of taxation, a couple's tax liability will not change simply because of a change in legal marital status. Indeed, the dominant current practice in income tax systems around the world is to designate the individual rather than the family as the unit of taxation (Alm & Melnik, 2005). Given the enormous, and increasing, range of family types in the United States—traditional single-earner households with a stay-at-home spouse, two-earner families, nonmarital cohabitation among opposite and same-sex couples, extended families, unrelated individuals living together—it may well be time to recognize that the tax laws of such a diverse society should treat all families the same. The choice to make the individual the unit of taxation would ensure that tax liabilities are independent of legal marital status. The recent legal cases regarding same-sex marriages make it even more pressing that this choice be addressed.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies

We are grateful to the Editor, Maureen Pirog, to the Co-Editor, Brad Heim, and to three anonymous referees for many helpful comments. We are also grateful to Wesley Howden, Rachel Donga, and Tamara Walker for their assistance in assembling and analyzing the data set, as part of the Tulane Research Experience for Undergraduates in Applied Microeconomics and Program Evaluation (Treu-Ampe), funded by National Science Foundation grant award SMA-1004569 and supervised by Professors Keith Finlay and Jonathan Pritchett. Daniel Feenberg provided invaluable help with the NBER TAXSIM calculations.

  1. 1

    These states include California, Connecticut, Delaware, Iowa, Maine, Maryland, Massachusetts, Minnesota, New Hampshire, New York, Rhode Island, Vermont, and Washington. Additionally, Colorado, Hawaii, Illinois, and New Jersey recognize same-sex civil unions.

  2. 2

    Another economic issue is the potential impact on legalization on employer-provided benefits. For example, Buchmueller and Carpenter (2012) evaluated the effect of 2005 California legislation that required private employers to provide the same health benefits to employees in committed same-sex relationships as those provided to employees in different-sex marriages. They found that the reform had no effects on differences in insurance outcomes between gay and straight men, but that the reform may have reduced insurance disparities between lesbians and heterosexual women. There are of course many other issues.

  3. 3

    There are also several studies that focus on the state-level effects. For example, Badgett (2010) used American Community Survey individual-level data on same-sex couples in Massachusetts to estimate 2009 federal income tax liabilities for same-sex couples in Massachusetts if they filed jointly as a married couple and if they filed as two separate individuals; she found that 66 percent of the same-sex couples in Massachusetts would pay on average $2,325 less in federal taxes if they could file as a married couple, 11 percent of same-sex couples in Massachusetts would see no change in their federal income taxes if filing as married, and the remaining 23 percent would see an average increase in taxes of $502. Other studies examine the economic impact of legalization at the state level, including the impacts on state income tax collections and on the overall state budget. For example, see Badgett (1998); Badgett, Sears, and Kukura (2005a, 2005b); Badgett et al. (2007); Badgett et al. (2008); Sears and Badgett (2008); Sears, Ramos, and Badgett (2009); and Herman, Konnoth, and Badgett (2011), among others. Many of these studies have been done by researchers at the Williams Institute of the UCLA School of Law (http://williamsinstitute.law.ucla.edu/).

  4. 4

    For example, see Kastanis, Badgett, and Herman (2012) for analysis of legalization in the State of Washington. These calculations include wedding and tourism expenditures generated by legalization. Again, see the many studies done by researchers at the Williams Institute of the UCLA School of Law.

  5. 5

    These tax cuts did not eliminate the entire marriage penalty once one includes the effect of an increase in family income resulting from marriage on qualification for government transfers.

  6. 6

    Historically, same-sex couples have been assumed to be less likely to have children and, together with the implications resulting from gender-based discrimination, same-sex couples have been assumed to be part of a two-earner couple (Alm, Badgett, & Whittington, 2000; Becker, 1991). In contrast, Becker (1991) predicted that heterosexual couples would choose to specialize in either home production or labor market production, implying more single-earner heterosexual couples.

  7. 7

    The ACS data used to determine family characteristics, income, and potential tax deductions were obtained from the Minnesota Population Center (Ruggles et al., 2010).

  8. 8

    We provide a comparison of the estimated number of same-sex couples and the corresponding effects on the estimated income tax revenue changes for each specification (e.g., the original ACS numbers, the ACS numbers with reallocated observations dropped, and the census-estimated numbers) in Tables 12 and 13.

  9. 9

    Interest paid on mortgage was calculated by assuming that home-owning households are 15 years into a 30-year mortgage with a fixed interest rate of 6.5 percent. This interest rate represents a 15-year average mortgage interest rate obtained from Freddie Mac. This includes home equity loans or second mortgages and takes the mean and median interest paid for each state representative household.

  10. 10

    The presence and magnitude of the marriage tax/subsidy depends heavily on the relative wages of the coupled individuals. To that end, couples are classified as a one-earner household if only one partner works for wages and as a two-earner household if both partners earn wages; two-earner households are further classified as equal earner if each partner earns the same gross wage and as unequal earner if the gross wages differ.

  11. 11

    We are grateful to Daniel Feenberg for his help with the NBER TAXSIM calculations.

  12. 12

    We suspect that a nontrivial portion of these children is the product of previous relationships rather than jointly adopted, so they cannot legally be claimed by the same-sex partner even if it would lower their joint tax liability. Adopted children make up a very small portion of the overall sample, and it is not possible to tell from our data if any adopted children have been adopted by both partners.

  13. 13

    Note that the tax-minimizing estimation does not imply that we assume that all couples itemize; rather, we assume that couples will choose to itemize if it reduces their overall tax liability. In some cases, individuals in a household may itemize when unmarried, but may then take the standard deduction when married (or vice versa), if this minimizes their overall liabilities. For 22 of the states that have income taxes, the standard deduction estimation is identical to the tax-minimizing estimation because the couples in the sample do not have sufficiently high qualifying deductions to warrant itemizing.

  14. 14

    Our negative result is comparable to the results obtained by Stevenson (2012) when he assumes an instantaneous penalty; his estimation based on the assumption of an endogenous penalty is larger, but still modest.

  15. 15

    Although it may seem counterintuitive that the marriage subsidy decreases when we minimize taxes, it occurs because individuals stand to gain a larger amount from itemizing as a single filer relative to filing as married. Remember that we are measuring the difference in tax liabilities rather than the overall level of tax liability.

  16. 16

    Our estimates suggest larger effects in Wyoming, North Dakota, and Montana, but these states do not seem to present an appropriate comparison given their relatively small sample sizes. See our later discussion in sensitivity analysis.

  17. 17

    Unweighted sample sizes for all states are reported in Table 1.

  18. 18

    As previously mentioned, there are likely legal deterrents that prohibit unmarried same-sex couples from choosing which partner will claim any children present as dependents.

  19. 19

    While the CBO (2004) assumed a 100 percent marriage rate of same-sex couples, Stevenson (2012) considered several potential marriage rates, including those suggested by Badgett (2010) and those that depend upon the demographics of the same-sex couples.

  20. 20

    Additionally, Connecticut, which was not previously estimated to experience an increase, now is estimated to experience an increase. This is because the restriction of eliminating couples whose federal tax liability increased by more than $5,000 also eliminated some couples who were experiencing an estimated reduction in state tax liability.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies
  • Alm, J., & Melnik, M. I. (2005). Taxing the “family” in the individual income tax. Public Finance and Management, 5, 67109.
  • Alm, J., & Whittington, L. A. (1997). Income taxes and the timing of marital decisions. Journal of Public Economics, 64, 219240.
  • Alm, J., Dickert-Conlin, S., & Whittington, L. A. (1999). The marriage tax. The Journal of Economic Perspectives, 13, 193204.
  • Alm, J., Badgett, M. V. L., & Whittington, L. A. (2000). Wedding bell blues: The income tax consequences of legalizing same-sex marriage. National Tax Journal, 66, 297316.
  • Amato, P. R. (2007). Strengthening marriage is an appropriate social policy goal. Journal of Policy Analysis and Management, 26, 952955.
  • Badgett, M. V. L. (1998). The fiscal impact on the state of Vermont of allowing same-sex couples to marry. Technical Report. Amherst, MA: Institute for Gay and Lesbian Strategic Studies.
  • Badgett, M. V. L. (2010). The economic value of marriage for same-sex couples. Drake Law Review, 58, 10811116.
  • Badgett, M. V. L., Baumle, A., Dravich, S., Romero, A., & Sears, B. (2007). The impact on Maryland's budget of allowing same-sex couples to marry. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Badgett, M. V. L., Baumle, A., Romero, A., & Sears, B. (2008). The impact on Iowa's budget of allowing same-sex couples to marry. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Badgett, M. V. L., Sears, B., & Kukura, E. (2005a). Counting on couples: Fiscal savings from allowing same-sex couples to marry in Connecticut. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Badgett, M. V. L., Sears, B., & Kukura, E. (2005b). The impact on New Hampshire's budget of allowing same-sex couples to marry. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Becker, G. S. (1991). Treatise on the family. Cambridge, MA: Harvard University Press.
  • Black, D., Sanders, S., & Taylor, L. (2000). Demographics of the gay and lesbian population in the United States: Evidence from available data sources. Demography, 37, 139154.
  • Black, D., Sanders, S., & Taylor, L. (2007). The economics of gay and lesbian families. The Journal of Economic Perspectives, 21, 5370.
  • Buchmueller, T. C., & Carpenter, C. S. (2012). The effect of requiring private employers to extend health benefit eligibility to same-sex partners of employees: Evidence from California. Journal of Policy Analysis and Management, 31, 388403.
  • Cohn, D., Passel, J. S., Wang, W., & Livingston, G. (2011). Barely half of U.S. adults are married—A record low. Pew Research, Social and Demographic Trends. Retrieved November 21, 2013, from http://www.pewsocialtrends.org/2011/12/14/barely- half-of-u-s-adults-are-married-a-record-low/.
  • Feenberg, D., & Coutts, E. (1993). An introduction to the TAXSIM model. Journal of Policy Analysis and Management, 12, 189194.
  • Furstenberg, F. F. (2007). Should government promote marriage? Journal of Policy Analysis and Management, 26, 956960.
  • Gale, W. G., & Potter, S. R. (2002). An economic evaluation of the Economic Growth and Tax Reconciliation Act of 2001. National Tax Journal, 55, 133186.
  • Gates, G. J. (2010). Same-sex couples in U.S. Census Bureau data: Who gets counted and why. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Gates, G. J., & Steinberger, M. (2010). Same-sex unmarried couples in the American Community Survey: The role of misreporting, miscoding, and misallocation. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Glaeser, E. (2002). The benefits of the home mortgage interest deduction. NBER Working Paper 9284. Cambridge, MA: National Bureau of Economic Research.
  • Graham, J., & Barr, J. (2008). Assessing the geographic distribution of same-sex and opposite-sex couples across the United States: Implications for claims of causality between traditional marriage and same-sex unions. Review of Economics of the Household, 6, 347367.
  • Herman, J., Konnoth, C., & Badgett, M. V. L. (2011). The impact on Rhode Island's budget of allowing same-sex couples to marry. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Holtz-Eakin, D. (2004). The potential budgetary impact of recognizing same-sex marriages. Report for Subcommittee on the Constitution. Congressional Budget Office. Washington, DC.
  • Kastanis, A., Badgett, M. V. L., & Herman, J. (2012). The economic impact of extending marriage to same-sex couples in Washington State. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Ramos, C., & Badgett, M. V. L. (2009). The impact on Maine's budget of allowing same-sex couples to marry. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Ruggles, S., Alexander, J. T., Genadek, K., Goeken, R., Schroder, M., & Sobek, M. (2010). Integrated public use microdata series: Version 5.0 [Machine-readable database]. Minneapolis, MN: University of Minnesota.
  • Sears, B., & Badgett, M. V. L. (2008). The impact of extending marriage to same-sex couples on the California budget. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Sears, B., Ramos, C., & Badgett, M. V. L. (2009). The impact of extending marriage to same-sex couples on the New Jersey budget. The Williams Institute, Research Report. Los Angeles, CA: UCLA School of Law.
  • Stevenson, A. (2012). The labor supply and tax revenue consequences of federal same-sex marriage legalization. National Tax Journal, 65, 783806.
  • United States Bureau of the Census (2011). America's families and living arrangements: 2011. Washington, DC: Bureau of the Census.

Biographies

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A BRIEF HISTORY OF MARRIAGE TAXES AND SUBSIDIES IN THE UNITED STATES
  5. DATA AND METHODOLOGY
  6. ESTIMATIONS
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Biographies
  • JAMES ALM is Professor and Chair of the Department of Economics, Tulane University, 208 Tilton Hall, New Orleans, LA 70118.

  • J. SEBASTIAN LEGUIZAMON is Professor in the Department of Economics, Vanderbilt University, VU Station B Number 351819, 2301 Vanderbilt Place, Nashville, TN 37235–1819.

  • SUSANE LEGUIZAMON is Professor in the Department of Economics, Western Kentucky University, 1906 College Heights Boulevard, Bowling Green, KY 42101.