FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL

FOR BETTER OR FOR WORSE:

MARRIAGE AND THE

FEDERAL INCOME TAX

JUNE 1997

The Congress of the United States

Congressional Budget Office

NOTES

Because numbers are often rounded, their sums may not equal totals shown in tables,

figures, and text.

For simplicity of exposition, this paper sometimes uses the term "families" to refer to

both families and individuals.

Values for earnings are sometimes negative. People can have negative earnings if

they lose money in self-employment and the losses exceed incomes from wages or

salaries.

Preface

his study examines the ways in which federal tax law affects the income taxes that

married couples pay and how demographic and labor market changes over the last two

decades have altered those effects. It also discusses a variety of possible changes in

the federal tax code that would reduce the higher tax liabilities that married couples often incur

because they cannot file individual tax returns. The Congressional Budget Office (CBO)

prepared the study at the request of Representative Barbara Kennelly of the Committee on

Ways and Means.

Roberton Williams and David Weiner of CBO's Tax Analysis Division wrote the study

under the direction of Rosemary Marcuss and Frank Sammartino. A number of people inside

and outside CBO reviewed drafts and provided valuable criticism and suggestions. They

include Mark Booth, Albert Davis, Theresa Devine, Daniel Feenberg, Diana Furchtgott-Roth,

Amy Rehder Harris, Richard Kasten, Pearl Richardson, and Eugene Steuerle.

Sherwood Kohn edited the manuscript and Marlies Dunson provided editorial assistance.

Simone Thomas produced drafts of the study. Kathryn Quattrone and Jill Sands prepared the

report for publication.

June E. O'Neill

Director

June 1997

Contents

SUMMARY xiii

ONE MARRIAGE PENALTIES AND BONUSES 1

Defining Marriage Penalties and Bonuses

A History of the Treatment of Married Couples

in the Federal Income Tax

The Significance of Marriage Penalties and Bonuses

TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 15

Factors Giving Rise to Marriage Penalties

and Bonuses

Which Couples Pay the Marriage Tax?

THREE MAGNITUDES OF MARRIAGE PENALTIES

AND BONUSES 27

Measuring Marriage Penalties and Bonuses

Total Marriage Penalties and Bonuses

The Distribution of Marriage Penalties

and Bonuses

Effective Tax Rates of Married Couples

FOUR LABOR MARKET AND DEMOGRAPHIC CHANGES

AFFECTING MARRIAGE PENALTIES AND BONUSES 37

Working-Age Married Couples

Changes in Earnings Among Income Categories

Changes in Earnings Among Age Categories

Changes in Earnings by Number of Children

Overview

FIVE REDUCING MARRIAGE PENALTIES 47

Widen Tax Brackets and Raise the Standard

Deduction for Joint Filers

Exempt from Taxes Some Income of

Lower-Earning Spouse

vi FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Modify Earned Income Tax Credit to Reflect

Number of Adult Earners

Restore the Requirement That Spouses File

Individual Returns

Allow Couples Choice of Filing Status

Fundamental Tax Reform

APPENDIXES

A Tax Treatment of Married Couples in

Other Countries

B Treatment of Married Couples Under State

Income Taxes

C Characteristics of Married and Unmarried

Couples, 1995

D Additional Supporting Tables

E Sources of Data

F Estimated Marriage Penalties and Bonuses

Under a Divorce Model

G The Income Distribution of Families

H Comparison of Demographic Changes for Working-Age

Couples and All Married Couples

CONTENTS vii

TABLES

S-1. Distribution of Joint Tax Returns and Marriage Penalty

or Bonus as a Percentage of Adjusted Gross Income,

by Adjusted Gross Income of Couple and Penalty or

Bonus Status, Projected 1996 xiv

S-2. Joint Tax Returns by Penalty or Bonus Status and

Division of Earnings Between Spouses, Projected 1996 xv

S-3. Effective Individual Income Tax Rates of Married Couples

Filing Individual or Joint Returns, Projected 1996 xvi

S-4. Alternative Approaches to Reducing Marriage Penalties xviii

1. Effects on Marginal Tax Rates of Married Couples Filing

Jointly Rather than Individually, Projected 1996 12

2. Marriage Penalties and Bonuses for Married and Unmarried

Couples, Simulated 1995 13

3. Factors Determining Whether Couples Face Marriage

Penalties or Bonuses, 1996 16

4. Total Value of and Tax Returns with Marriage Penalties

and Bonuses Under Alternative Measures, Projected 1996 30

5. Couples Receiving Marriage Penalties and Bonuses by

Adjusted Gross Income, Projected 1996 31

6. Marriage Penalties and Bonuses by Adjusted Gross Income

of Couple, Projected 1996 32

7. Joint Tax Returns by Penalty or Bonus Status and Division

of Earnings Between Spouses, Projected 1996 33

8. Marriage Penalties and Bonuses by Penalty or Bonus Status

and Division of Earnings Between Spouses, Projected 1996 34

9. Effective Individual Income Tax Rates of Married Couples

Filing Individual or Joint Returns, by Adjusted Gross

Income and Division of Earnings Between Spouses,

Projected 1996 35

10. Distribution of Working-Age Married Couples, by Division

of Earnings Between Spouses in 1969, 1979, 1989, and 1995 39

11. Percentage of Working-Age Married Couples with Two

Earners, by Total Income in 1969, 1979, 1989, and 1995 40

viii FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

12. Percentage Distribution of Working-Age Married Couples, by

Division of Earnings Between Spouses and Total Income in

1969, 1979, 1989, and 1995 41

13. Percentage of Working-Age Married Couples with Two

Earners, by Age of Older Spouse in 1969, 1979, 1989, and 1995 42

14. Percentage Distribution of Working-Age Married Couples,

by Division of Earnings Between Spouses and Age of Older

Spouse in 1969, 1979, 1989, and 1995 43

15. Percentage of Working-Age Married Couples with Two

Earners, by Number of Children in 1969, 1979, 1989, and 1995 44

16. Percentage Distribution of Working-Age Married Couples,

by Division of Earnings Between Spouses and Number of

Children in 1969, 1979, 1989, and 1995 45

17. Percentage Distribution of Families with Children, by Type

and Number of Earners in Married Couples in 1969, 1979,

1989, and 1995 46

18. Simulated Effects of Demographic Change on Marriage

Penalties and Bonuses, 1969 and 1996 46

C-1. Distribution of Married and Unmarried Couples

by Total Income, 1995 65

C-2. Distribution of Married and Unmarried Couples

by Age of Older Spouse, 1995 66

C-3. Distribution of Married and Unmarried Couples

by Number of Children, 1995 67

C-4. Simulated Marriage Penalties and Bonuses for

Married and Unmarried Couples, 1995 68

D-1. Factors Determining Marriage Penalties and Bonuses

by Adjusted Gross Income, Simulated 1996 70

D-2. Married Couples by Earnings of Each Spouse, Simulated 1996 71

D-3. Married Couples with Marriage Penalties or Bonuses

and Average Penalty or Bonus by Earnings of Each

Spouse, Simulated 1996 72

D-4. Distribution of Marriage Penalties and Bonuses by

Number of Earners, Simulated 1996 74

CONTENTS ix

F-1. Marriage Penalties and Bonuses for Married Couples Under

Alternative Assumptions about Distribution of Children

Between Spouses, Simulated 1996 78

G-1. Percentage Distribution of Families by Income

Category and Family Type, 1995 80

G-2. Distribution of Working-Age Married Couples with

Earnings by Income Category in 1969, 1979, 1989, and 1995 81

H-1. Selection of Married Couples for Analysis Based

on Age and Earnings, 1995 84

H-2. Percentage of Couples with Two Earners: All Married

Couples and Working-Age Married Couples with Positive

Earnings in 1969, 1979, 1989, and 1995 85

H-3. Distribution of Working-Age and All Married Couples

by Division of Earnings Between Spouses in 1969,

1979, 1989, and 1995 86

H-4. Percentage Distribution of Working-Age and All Married

Couples with Two Earners, by Total Income in 1969, 1979,

1989, and 1995 87

H-5. Percentage Distribution of Working-Age and All Married

Couples by Division of Earnings Between Spouses and

Total Earnings in 1969, 1979, 1989, and 1995 88

H-6. Percentage Distribution of Working-Age and All Married

Couples with Two Earners by Age of Older Spouse in 1969,

1979, 1989, and 1995 90

H-7. Percentage Distribution of Working-Age and All Married

Couples by Division of Earnings Between Spouses and Age

of Older Spouse in 1969, 1979, 1989, and 1995 91

H-8. Percentage Distribution of Working-Age and All Married

Couples with Two Earners by Number of Children in 1969,

1979, 1989, and 1995 93

H-9. Percentage Distribution of Working-Age and All Married

Couples by Division of Earnings Between Spouses and

Number of Children in 1969, 1979, 1989, and 1995 94

x FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

FIGURES

1. The Conflict Among Goals of Income Taxation 3

2. Earned Income Tax Credit by Number of Children

and Earnings, 1996 Tax Law 20

3. Marriage Penalties and Bonuses Resulting from the

Earned Income Tax Credit, 1996 Tax Law 22

4. Marriage Penalties and Bonuses by Number of

Children and Earnings Split, 1996 Tax Law 24

5. Work Patterns of Married Couples, 1969-1995 37

6. Distribution of Families and Unrelated Individuals

by Type, 1969-1995 38

7. Projected 1996 Marriage Penalty Relief Under Alternative

Proposals, Basic Measure of Marriage Penalty 49

8. Projected 1996 Target Efficiency of Alternative Proposals

to Relieve Marriage Penalties, Basic Measure of Marriage

Penalty 50

BOXES

1. A Marriage Penalty 4

2. A Marriage Bonus 5

3. 1996 Marriage Bonuses Resulting from Tax Brackets

for One-Earner Couples 18

4. 1996 Marriage Penalties Resulting from Tax Brackets

for Two-Earner Couples in Which Spouses Have

Equal Earnings 19

5. A Marriage Penalty for a Low-Income Couple

with Children 21

6. Alternative Measures of Marriage Penalties

and Bonuses 29

7. Setting Tax Brackets and Standard Deductions for

Joint Filers to Twice Those for Single Filers 48

8. The Two-Earner Deduction 52

CONTENTS xi

A-1. Tax Treatment of Earned Income in OECD Countries, 1993 60

B-1. Treatment of Married Couples Under State Income

Taxes, 1993 62

Summary

arriage affects the income taxes people pay,

sometimes causing married couples to pay

more than they would if they were single and

sometimes causing them to pay less. That is because

the household the married couple or the single indi-

vidual is the basic tax-paying unit in the U.S. system.

Differences in the tax liabilities of households reflect

attempts by the Congress to maintain balance in tax

treatment among different kinds of families. On the

one hand, the tax code seeks to levy the same tax on

couples with the same income, regardless of who earns

the income. On the other hand, the code tries to mini-

mize the effect of marriage on a couple's tax liability.

A tax structure with progressive rates, however, cannot

attain both goals. The incompatibility of those three

goals progressive rates, equal treatment of married

couples, and marriage neutrality results in a continu-

ing tension within the tax code. The third goal, mar-

riage neutrality, has proven to be the most elusive in

our system.

The balance among the goals has shifted over time,

partly in response to complaints of unfair treatment and

partly in reaction to changing demographic patterns.

Growing numbers of single taxpayers, the greater like-

lihood that husbands and wives both work, increasing

parity of earnings for husbands and wives who do

work, and changes in tax rate brackets have all contrib-

uted to the pressures on the balance.

Differences in income tax liabilities caused by mar-

ital status are embodied in a number of tax code provi-

sions, beginning with separate rate schedules for mar-

ried couples and single individuals. The most impor-

tant other differences are those in the standard deduc-

tion and the earned income tax credit (EITC). Those

factors cause most two-earner couples in which hus-

band and wife have roughly equal incomes to pay more

tax than they would if they could file individual tax re-

turns, incurring what has become known as a "marriage

penalty." At the same time, couples with just one

earner or in which husband and wife have quite differ-

ent incomes generally pay lower taxes as joint tax filers

than they would if they could file as single taxpayers.

Those couples receive what analysts have termed "mar-

riage bonuses."

Marriage penalties also exist outside the federal

income tax in the form of transfer program rules that

limit married couples' benefits to amounts below what

they would receive as single individuals. For example,

a low-income single mother might qualify for welfare

assistance, food stamps, and Medicaid, benefits that she

would lose if she obtained a good-paying job or married

a man who had a moderate income. That loss of bene-

fits can offset much of the income gains from work or

marriage, thus providing disincentives for each. Al-

though such marriage penalties outside the tax code can

be substantial and often are larger than those created

by the tax code this study addresses only the different

income taxes paid by joint and individual filers.

If marriage penalties and bonuses were small, they

would be of little consequence. Taxpayers would not

perceive as unfair the small differences in taxes created

by marriage. Perhaps more important, those differ-

ences would not lead people to change their behavior

toward work or marriage itself. But the penalties and

bonuses can be large. Empirical evidence indicates that

they may affect work patterns, particularly for a cou-

xiv FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

ple's second earner. Joint tax filing often imposes a

substantially higher rate on a couple's second earner

than does individual filing. The higher rates reduce the

after-tax wage and may cause second earners to work

fewer hours or not to work at all, thus reducing eco-

nomic efficiency. Furthermore, large differences in tax

liabilities between otherwise similar couples lead to

perceptions of unfairness and calls for changes in the

tax code that would reduce those differences.

The size and imposition of marriage penalties and

bonuses depend on which components of the tax code

are included in their measurement. In general, the more

components of the code that are included in the yard-

stick, the more couples incur marriage penalties and the

fewer couples receive bonuses. The components in-

cluded depend on which taxpaying circumstances are

assumed. The measure used in this study includes the

code components that account for the bulk of the pen-

alty and bonus amounts. The estimates of those penal-

ties and bonuses, however, assume no induced changes

in labor supply or marriage and thus underestimate the

economic costs involved.

To be more specific, marriage penalties and bo-

nuses occur because the tax code allows different stan-

dard deductions in determining the taxable incomes of

couples and single taxpayers, has different tax rates for

given levels of taxable income for the two groups, and

provides the EITC without regard to marital status.

Couples in which spouses have similar incomes gener-

ally can claim smaller deductions, face higher marginal

tax rates, and lose some or all of the EITC they would

have if they filed individual tax returns and thus in-

curred marriage penalties. Couples in which one

spouse earns much less than the other or does not work

face the opposite situation. They may qualify for larger

deductions, incur lower tax rates, and receive more

EITC than under individual filing and thus receive mar-

riage bonuses.

Distribution of

Marriage Penalties

And Bonuses

An estimated 42 percent of couples incurred marriage

penalties in 1996, 51 percent received bonuses, and 6

percent paid taxes unaffected by their marital status

(see Summary Table 1). That distribution varies mark-

edly across the income distribution. Only 12 percent of

couples with incomes below $20,000 had penalties and

63 percent received bonuses. Couples with incomes

between $20,000 and $50,000 were somewhat more

likely to receive bonuses than to incur penalties,

whereas couples with incomes above $50,000 were

somewhat more likely to incur penalties than to receive

bonuses.

Measured as a percentage of income, marriage pen-

alties and bonuses are largest for low-income families

and least for high-income families (see Summary

Table 1). Couples with incomes below $20,000 who

bear penalties are estimated to pay nearly 8 percent

Summary Table 1.

Distribution of Joint Tax Returns and Marriage

Penalty or Bonus as a Percentage of Adjusted

Gross Income, by Adjusted Gross Income of

Couple and Penalty or Bonus Status,

Projected 1996

Less Than $20,000- More Than All

$20,000 $50,000 $50,000 Incomes

Returns in Income Category (Percent)

With Penalties12445442

Unaffected 25136

With Bonuses 63

55 44 51

All Returns 100 100 100 100

Penalty or Bonus as a Percentage

of Adjusted Gross Income

Penalties 7.6 3.2 1.6 2.0

Bonuses 5.0 2.6 2.0 2.3

Total Penalties and Bonuses (Billions of dollars)

Penalties 0.9 9.6 18.3 28.8

Bonuses 3.9

8.7 20.3 32.9

Net Effect on

Tax Liability -3.0 0.9 -2.0 -4.1

SOURCE: Congressional Budget Office simulations based on data

from the 1993 Statistics of Income.

a. Positive values indicate net penalties; negative values indicate net

bonuses.

SUMMARY xv

more of their income in taxes measured at 1996 income

levels than they would pay if they could file individual

tax returns. For their counterparts who received them,

however, bonuses averaged 5 percent of income. Be-

cause bonus recipients in that income category far out-

number those incurring penalties, low-income married

couples received a net bonus of $3 billion. Both penal-

ties and bonuses for middle-income couples averaged

roughly 3 percent of the incomes of affected families.

Slightly higher average penalties offset a slightly lower

proportion of couples with penalties, yielding a net pen-

alty of nearly $1 billion. Couples with incomes over

$50,000 received a total net bonus of $2 billion. Cou-

ples with bonuses saved 2 percent of their incomes by

filing jointly, more than offsetting the added taxes

equal to 1.6 percent of income paid by penalized

couples.

Couples with a more equal division of incomes be-

tween husband and wife are more likely to incur mar-

riage penalties and less likely to receive bonuses than

couples in which one spouse has significantly less in-

come than the other or does not work at all. One-earner

couples received seven-eighths of the total value of bo-

nuses in 1996 but represented less than one-half of all

couples (see Summary Table 2). On net, their income

tax liabilities were nearly $29 billion lower than if they

were required to file individual tax returns. By

contrast, the one-quarter of all couples in which hus-

band and wife both earn at least one-third of the cou-

Summary Table 2.

Joint Tax Returns by Penalty or Bonus Status and Division of Earnings Between Spouses, Projected 1996

Both Spouses

Each Spouse Had Earnings

Earned and One Spouse

at Least Contributed Only One

One-Third of More Than Spouse

Couple's Two-Thirds of Had All

Earnings Couple's Earnings Earnings Couples

Joint Returns in Penalty or Bonus Category (Percent)

With Penalties 52 48 0 100

Unaffected 21 5 75 100

With Bonuses 2 21 77 100

All Joint Tax Returns 25 31 44 100

Total Penalty or Bonus (Billions of dollars)

Penalties 15.2 13.6 0 28.8

Bonuses 0.1

4.4 28.5 32.9

Net Effect on Tax Liability 15.1 9.2 -28.5 -4.1

SOURCE: Congressional Budget Office projections based on data from the 1993 Statistics of Income.

a. Couples with no earners and those with one earner incur similar marriage penalties and bonuses and are thus combined in the analysis. For

simplicity of exposition, the text refers to both types of couples as having one earner.

b. Positive values indicate net penalties; negative values indicate net bonuses.

xvi FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

ple's income filed one-half of all joint returns incurring

marriage penalties but only 2 percent of those receiving

bonuses. As a result, those couples paid $15 billion

more in taxes than they would have if they could have

filed as individuals. Those estimates are based on ob-

served earnings under current tax laws, however, and

do not take account of the costs of couples altering their

work and marriage patterns in response to the tax code.

Although the prevalence of marriage penalties and

bonuses indicates that the tax code fails to provide mar-

riage neutrality, it is more successful in achieving equal

treatment of married couples with similar incomes. If

couples were required to file individual tax returns,

those with one earner would face substantially higher

tax rates than those with two earners who have roughly

equal incomes (see Summary Table 3). Because the tax

code generally requires couples to file jointly, those

with different divisions of earnings between spouses

incur more nearly equal tax rates. Marriage penalties

and bonuses arise from this equalization of tax rates for

couples with different divisions of earnings.

Summary Table 3.

Effective Individual Income Tax Rates of

Married Couples Filing Individual or

Joint Returns, Projected 1996 (In percent)

Both Spouses

Each Spouse Had Earnings

Earned and One Spouse

at Least Contributed Only One

One-Third of More Than Spouse

Tax Filing Couple's Two-Thirds of Had

Status Earnings Couple's Earnings Earnings

Joint 13.2 15.2 15.0

Individual 11.2 14.3 17.6

SOURCE: Congressional Budget Office projections based on data

from the 1993 Statistics of Income.

NOTE: Effective tax rates equal federal individual income tax liability

as a percentage of adjusted gross income.

a. Couples with no earners and those with one earner incur similar

marriage penalties and bonuses and are thus combined in the

analysis. For simplicity of exposition, the text refers to both types

of couples as having one earner.

Demographic Changes

Affecting Marriage

Penalties and Bonuses

In the last two decades, a sharp rise in the share of mar-

ried couples with two earners has taken place, accom-

panied by increasing equality of the earnings of hus-

bands and wives, particularly among couples with

higher incomes. Those changes, occurring for couples

with and without children, have contributed to a rise in

the share of couples incurring marriage penalties.

Between 1969 and 1995, the proportion of working-

age married couples with two workers grew from 48

percent to 72 percent. By itself, that shift tended to

increase marriage penalties and reduce bonuses. Off-

setting that trend, however, was a decline in married

couples as a share of all families and individuals: cou-

ples made up two-thirds of all such tax units in 1969,

but only one-half by 1995. Thus, a smaller fraction of

tax returns are now joint (and hence may be subject to

marriage penalties and bonuses) than was the case a

generation ago.

The shift toward two-earner couples has been ac-

companied by increasing equality of the incomes of

husbands and wives. Between 1969 and 1995, the frac-

tion of working-age couples in which both husband and

wife earned at least one-third of the couple's income

doubled from 17 percent to 34 percent. The greater

equality of spouses' earnings increased both the share

of couples incurring marriage penalties and the size of

those penalties.

Those changes in earnings patterns for married cou-

ples occurred for all subgroups defined by earnings,

age, and presence and number of children. However,

the changes were largest among couples with the high-

est incomes, those with both spouses under age 55, and

those with children.

If the demographic characteristics of married cou-

ples had not changed between 1969 and 1996, more

couples would have received bonuses under 1996 tax

law and fewer would have incurred penalties. Simula-

tions indicate that without the demographic change of

the last quarter century, two-thirds of couples would get

bonuses and less than one-third would pay penalties

SUMMARY xvii

under joint filing, compared with the nearly even split

between the two categories in 1996.

Reducing Marriage Penalties

Recent growth in the number of married couples who

incur marriage penalties and the increasing size of those

penalties have focused attention on the effects of mar-

riage on taxes and on alternative ways to alleviate them.

One approach involves relatively minor alterations of

the current tax code such as changes in tax brackets and

standard deductions, restoration of the two-earner de-

duction, or credits to offset penalties. More significant

changes would allow couples to file either joint or sin-

gle returns or require that all taxpayers file individual

returns. Finally, comprehensive tax reform, replacing

the current income tax with either a simpler tax on in-

comes or some form of consumption tax, would entail

some pattern of penalties and bonuses unless individual

filing was incorporated into the new system.

Changes in the tax code that reduce or eliminate

marriage penalties face an inevitable trade-off: cutting

penalties means lower taxes for some couples and ei-

ther reduced federal tax revenues or higher taxes on

other taxpayers. Revenue-neutral options require a re-

distribution of taxes from those now incurring penalties

to other taxpayers, either couples now receiving bo-

nuses or single people.

Six basic approaches address the problem of mar-

riage penalties directly (see Summary Table 4). Com-

bining options would generate interactions in ways not

analyzed in this study.

Widen Tax Brackets and Raise the Standard De-

duction for Joint Filers.

Increasing tax bracket

widths and standard deductions for joint filers to

twice those for single people would reduce total

marriage penalties by nearly one-half at an esti-

mated revenue cost of about $25 billion measured at

1996 income levels. Nearly 90 percent of tax reduc-

tions would accrue to couples with incomes above

$50,000. Half would go to those receiving bonuses

under current law.

Exempt from Taxes Some Income of the Lower-

Earning Spouse.

Exempting some earnings of

lower-earning spouses would ease the high marginal

tax rates they may face on even low amounts of

earnings. For example, restoring the 10 percent de-

duction for up to $30,000 of earnings would reduce

total marriage penalties by one-third at an estimated

cost of about $9 billion in lost revenues measured at

1996 income levels. Principal beneficiaries would

be couples with incomes between $50,000 and

$100,000, for whom penalties would fall by more

than one-half. Couples with lower incomes would

see their penalties decline by about one-tenth.

Modify the EITC to Reflect Number of Adult Earn-

ers.

A significant share of the marriage penalties

incurred by low-income couples is caused by the

loss of EITC when individual incomes are com-

bined. Providing two-earner couples with EITC pa-

rameters double those for single-earner tax units

would reduce marriage penalties for low-income

couples by about one-fourth, producing an estimated

revenue loss of about $4 billion measured at 1996

income levels. Virtually all of the benefits would go

to couples with incomes below $50,000, and less

than 1 percent would go to those now receiving bo-

nuses because of the EITC. Nearly 3.7 million cou-

ples would become newly eligible for the EITC.

Require Spouses to File Individual Tax Returns.

Requiring all taxpayers to file single returns would

eliminate all marriage penalties and bonuses, but

would redistribute roughly $30 billion in tax liabili-

ties from couples now receiving bonuses to those

incurring penalties. Marriage would no longer affect

a couple's tax bill. Although revenue gains from

lost bonuses would outweigh revenue lost in elimi-

nated penalties, yielding an estimated $4 billion in

net additional revenue in 1996, changes in the tax

liabilities of couples would be enormous. The 25

million couples now receiving bonuses would pay an

average of $1,300 in added taxes, and 21 million

couples now incurring penalties would find their

taxes down an average of nearly $1,400.

Allow Couples to Choose Filing Status.

Allowing

couples to choose to file either jointly or as individu-

als would eliminate all marriage penalties at an esti-

xviii FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

mated cost of about $29 billion in lost revenues

measured in 1996 income levels. All existing mar-

riage bonuses would continue.

A less extreme version of that approach would allow

couples a tax credit based on a simplified calcula-

tion of the marriage penalty they now pay. If that

calculation was limited to earnings, standard deduc-

tions, and a single personal exemption for each

spouse, the resulting credits would cut total mar-

riage penalties by about one-third, producing an es-

timated revenue loss of about $10 billion measured

at 1996 income levels. Couples with incomes above

$50,000 would have their penalties cut by one-half,

whereas low-income couples would be relieved of

just one-tenth of their current penalties, largely be-

cause the credit would ignore the EITC.

Fundamental Radical Tax Reform.

Fundamental

changes in the federal tax system such as those dis-

Summary Table 4.

Alternative Approaches to Reducing Marriage Penalties

Share Share to Tax Units (Percent)

of With With

Revenue Loss Penalty Incomes Incomes Currently

(Billions of Removed Below Above Receiving

Approach dollars) (Percent) $20,000 $50,000 Bonuses

cc d

Set Standard Deduction and Width of

Tax Brackets for Joint Filers Equal to

Double Those for Single Filers 25 44 6 87 51

Restore Two-Earner Deduction 9 32 1 82 10

Double Earned Income Tax Credit

Parameters for Two-Earner Couples 4 12 41 1 1

Require Spouses to

File Individual Returns -4 100 e e e

Allow Couples to

Choose Filing Status 29 100 3 64 0

Replace Income Tax

with Consumption Tax f f f f f

SOURCE: Congressional Budget Office simulations.

NOTE: See text for complete description of approaches.

a. Negative value indicates revenue gain because tax increases for some taxpayers exceed tax reductions for others.

b. Total tax reduction for couples currently incurring tax penalty as a percentage of current total tax penalty.

c. Percentage of revenue loss going to couples with income in given range. Share includes tax reductions for couples getting either penalties or

bonuses.

d. Fraction of total revenue loss going to couples currently receiving marriage bonuses.

e. All couples incurring penalties get tax reductions equal to penalties. All couples receiving bonuses incur tax increases equal to bonuses. Shares

going to each income group and to those receiving bonuses are misleading and therefore not shown.

f. Revenue loss depends on formulation of consumption tax. Distribution among income groups and to those now receiving bonuses cannot be

determined.

SUMMARY xix

cussed during the 104th Congress could markedly

reduce marriage penalties. Moving to a flat

(single-rate) tax, for example, would eliminate all

penalties (and bonuses) resulting from tax brack-

ets. If reform also eliminated other tax features

such as standard deductions and the EITC, penal-

ties and bonuses arising from those features would

also disappear. Alternatively, switching from a tax

on income to a tax on consumption would wipe out

all marriage penalties and bonuses if exemptions

from the tax were independent of marital status and

marriage did not affect a couple's combined con-

sumption.

Chapter One

Marriage Penalties and Bonuses

ederal income tax laws generally require that a

married couple file a joint tax return based on

the combined income of husband and wife. As

a result, husbands and wives with similar incomes usu-

ally incur a larger combined tax liability than they

would if they could file individually. At the same time,

spouses who have markedly different incomes but file

as a couple generally face smaller tax bills than they

would if they were single. Those two possibilities, of-

ten referred to as marriage "penalties" and "bonuses,"

result from conflicting goals of a tax system that at-

tempts to balance fairness between married and unmar-

ried couples, among married couples, and among tax-

payers with differing incomes.

Under 1996 tax law, a married couple could face a

federal tax bill that was more than $20,000 higher than

the amount they would pay if they were not married and

could file individual tax returns, whereas other couples

may find that filing a joint tax return reduces their tax

bill by more than $4,000. Although there are various

ways of defining marriage penalties and bonuses, one

broad measure indicates that more than 21 million mar-

ried couples paid an average of nearly $1,400 in addi-

tional taxes in 1996 because they must file jointly,

whereas another 25 million found that the benefits of

filing jointly decrease their tax bills by an average of

about $1,300. Marriage penalties totaled about $29

billion in 1996, and bonuses added up to roughly $33

billion.

Marriage penalties are a relatively recent aspect of

federal income taxes, the result of changes made in the

tax code in 1969 to redress what some individuals

viewed as unfairly high taxes paid by single people.

Although the impact and size of marriage penalties and

bonuses have changed over the years with changes in

the tax code, recent tax increases in the form of addi-

tional tax brackets, and shifts in the work patterns of

married couples have combined to expand the propor-

tion of couples subject to penalties and boost the mag-

nitude of those payments. Those movements have

raised interest in policies that would reduce the impact

of marriage penalties in the tax code.

Marriage penalties and bonuses are not deliberately

intended to punish or reward marriage. Rather they are

the result of a delicate balance among disparate goals of

the federal income tax system. Largely for historical

reasons, the United States imposes income taxes not on

individuals as do most other industrialized nations, but

on couples, regardless of the division of incomes be-

tween spouses. Under a progressive structure of tax

rates, treating families equally generally means that a

couple's taxes will differ depending on their marital

status. Over time, the ebb and flow of political pres-

sures have shifted the equilibrium between marriage

penalties and bonuses as the Congress has attempted to

balance the competing demands of married and unmar-

ried taxpayers. Shifting demographics, particularly the

1. Federal income taxes are only one source of marriage penalties. Oth-

ers include state and local income taxes and federal transfer programs

such as Social Security, in which benefits depend in part on both mari-

tal status and resources. This study does not consider any effects of

marriage beyond those affecting a couple's federal individual income

tax liability. Two recent studies that examine the broader question

of marriage penalties and bonuses are Linda Giannarelli and Eugene

Steuerle,

The Twice-Poverty Trap: Tax Rates Faced by AFDC Re-

cipients

(Washington, D.C.: Urban Institute, April 1995), and Stacy

Dickert-Conlin and Scott Houser,

Taxes and Transfers: A New Look

at the Marriage Penalty

(unpublished working paper, Lexington:

University of Kentucky, July 1996).

2 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

growing proportion of couples with two earners, have

exerted further pressures on that fragile equilibrium.

Whether a couple incurs a penalty or receives a bo-

nus in their taxes depends on the level of their income,

the division of earnings between husband and wife, and

how many children they have. Over the last quarter

century, as more wives have entered the labor force and

their wages have risen in relation to men's, earnings of

spouses have grown more equal. The change increases

the likelihood that a couple will incur a tax penalty, as

compared with two single people whose earnings are

the same. Changes in demographics and earnings alone

appear to have increased by one-half the proportion of

couples incurring marriage penalties.

In 1995, the House of Representatives passed legis-

lation that would slightly lower the tax penalty imposed

on married couples, which would have cost $2 billion

annually in reduced government revenues. That same

year the Senate included in its budget reconciliation bill

a provision that would, over a 10-year period, raise the

standard deduction for married couples filing jointly to

twice that of single filers. When fully phased in, that

change would reduce the marriage penalty for many

couples and increase the marriage bonus for others at

an annual cost of roughly $5 billion in lower revenues.

Other options that would further reduce or eliminate

marriage penalties could have significantly higher reve-

nue costs.

Definin

Marria

e Penalties

and Bonuses

Marriage penalties and bonuses in the federal individual

income tax are defined as the increase or decrease in

taxes that some married couples incur because they pay

taxes as couples rather than as individuals. Penalties

and bonuses result from the tax code's pursuit of three

conflicting goals: equal treatment of married couples,

marriage neutrality, and progressive taxation.

Equal Treatment of Married Couples

According to the principle of equal treatment, married

couples who have equal incomes should pay the same

income taxes, regardless of the way earnings are di-

vided between spouses. Thus, a couple in which each

spouse earns half the income should pay the same taxes

as a couple with the same income but in which only one

spouse works. That concept runs afoul of equity princi-

ples when married couples with equal incomes have

different abilities to pay. A prime example involves

one- and two-earner families with the same total in-

come. Because the nonearning spouse has time to per-

form household tasks, the one-earner couple is better

off than the couple with two workers, which often has

to pay for such services as child care and household

chores. It thus has a greater ability to pay and, by some

standards of equity, should pay higher taxes. But deter-

mining the difference in ability to pay is difficult, if not

impossible, and the government has shied away from

imposing different taxes on that basis.

Marriage Neutrality

Under the principle of marriage neutrality, a couple's

income tax bill should not depend on their marital sta-

tus. Without neutrality, the tax code would provide an

incentive, even though it is small, for couples either to

marry

if marriage would lower their tax bill or to

remain single if marriage increases their taxes. The

concept of marriage neutrality may violate principles of

equity, however, if marriage (or divorce) either in-

creases or decreases couples' combined ability to pay.

If, for example, marriage enables couples to reduce

their housing costs because they live together, marriage

would increase their ability to pay, and in fact some

standards of equity would dictate that they pay higher

taxes than they did before they married.

Progressive Taxation

The principle of progressive taxation holds that taxpay-

ers who have higher incomes should pay a larger per-

centage of their incomes in taxes than do those with

lower incomes. The federal income tax, for example,

imposes no tax on families and individuals who have

the lowest incomes, and taxes income above various

thresholds according to a series of graduated rates rang-

ing from 15 percent to 39.6 percent. Graduated tax

rates, however, are not necessary to achieve progres-

sivity. Combining a single rate and a low-income de-

duction or credit would also be progressive.

Couple A: married

with both spouses

earning $30,000

Same Tax

(Marriage Neutrality)

Same Tax

(Marriage Neutrality)

Same Tax

(Equal Treatment

of Married Couples)

Different Taxes

(Progressivity)

Couple A: not married,

each person

earning $30,000

Couple B: married

with one worker

earning $60,000

Couple B: not married,

one worker

earning $60,000

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 3

The Unattainable Ideal

A simple example demonstrates the impossibility of

simultaneously achieving all three of those goals (see

Figure 1). Consider two couples, each earning $60,000

and having no unearned income. Both members of cou-

ple A work, each earning $30,000, whereas only one

member of couple B works, earning the couple's entire

$60,000. Satisfying the goal of taxing married couples

with the same incomes equally requires that married

couple A and married couple B pay the same taxes. To

meet the goal of marriage neutrality, the tax bill of each

couple must not depend on their marital status; that is,

the tax on two single people, each earning $30,000,

must equal the tax on a married couple in which each

spouse earns $30,000, and the tax on two single people,

one with a $60,000 income and the other with no in-

come, must equal the tax on a one-earner married cou-

ple with the same $60,000 income.

Combining the two goals implies that two single

people, each with a $30,000 income, pay the same total

tax as two single people, one earning $60,000 and the

other with no income, or, equivalently, that a single per-

son with a $60,000 income pay exactly double the tax

of a single person earning $30,000. But that outcome

directly violates the goal of progressive taxation: under

that principle, a person who has twice the income of

another pays more than twice the taxes.

Relaxing one or more of the three goals solves the

contradiction among them. Forgoing progressivity, for

example, would allow a tax system to satisfy both the

goal of marriage neutrality and that of equal treatment

Figure 1.

The Conflict Among Goals of Income Taxation

SOURCE: Congressional Budget Office, derived from Treasury Department, Office of Tax Analysis,

Federal Income Tax Treatment of Married and

Single Taxpayers

(September 13, 1977), p. 2.

4 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box 1.

A Marriage Penalty

A couple with $75,000 in total earnings, split evenly

between the husband and the wife, would have incurred

a marriage penalty of nearly $1,400 under 1996 tax law.

The penalty results from two factors. First, the com-

bined standard deduction for two individual tax filers

would have been $8,000

$1,300 more than the stan-

dard deduction available on a joint return. At the cou-

ple's marginal tax rate of 28 percent, the lower deduc-

tion would have increased the couple's tax liability by

$364 (28 percent of $1,300). Second, because tax

brackets for joint returns were less than twice as wide as

those for individual returns, $7,900 that is taxed at 15

percent on individual returns would have incurred a 28

percent rate on a joint return. That higher tax rate

would have raised the couple's tax liability by an addi-

tional $1,027 (28 percent minus 15 percent equals 13

percent of $7,900). In combination, the two factors

would have increased the couple's tax liability by 1.9

percent of their adjusted gross income.

Husband

Wife Couple

Adjusted Gross Income $37,500 $37,500 $75,000

Less personal exemptions 2,550 2,550 5,100

Less standard deduction 4,000 4,000 6,700

Equals taxable income 30,950 30,950 63,200

Taxable at 15 percent 24,000 24,000 40,100

Taxable at 28 percent 6,950 6,950 23,100

Tax Liability 5,546 5,546 12,483

Marriage Penalty $1,391

As a Percentage of Adjusted Gross Income 1.9

SOURCE: Congressional Budget Office.

of married couples. Alternatively, married couples with

equal incomes but unequal division of income between

spouses could pay different taxes based on the principle

that the division of income within couples measures

different abilities to pay. Or, as in the United States,

the taxes of a married couple could depend only on their

total income, thus satisfying the goal of equal treatment

of married couples but sacrificing the goal of marriage

neutrality because some married couples would pay

higher or lower taxes than they would if they were sin-

gle. In practice, the last solution to the contradiction

among the three goals can impose a marriage penalty

(that is, higher taxes than if they were not married) on

some couples while granting a marriage bonus (lower

taxes) to others. Two examples based on the federal

individual income tax demonstrate these possibilities.

Example: A Marriage Penalty

. A married couple

with $75,000 in earnings, split evenly between husband

and wife, paid $1,391 more in federal individual in-

come taxes in 1996 than they would have if they had

not been married (see Box 1). Two factors contribute

to that marriage penalty. First, when filing a joint tax

return, the couple can claim a standard deduction that is

$1,300 less than the two single standard deductions

$4,000 each they would receive if they were not mar-

ried. At a marginal tax rate of 28 percent, the smaller

deduction costs them $364 (28 percent of $1,300) in

additional taxes. Second, because tax brackets for mar-

ried couples are not twice as wide as those for single

tax filers, $7,900 that is taxed at 15 percent on the two

single returns faces a 28 percent rate on the joint return,

yielding an additional tax of $1,027 (28 percent minus

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 5

Box 2.

A Marriage Bonus

A couple with $75,000 in total earnings, all earned by

the wife, would have received a marriage bonus of

nearly $4,000 under 1996 tax law. The bonus results

from three factors. First, filing jointly, the couple would

have claimed $5,100 in personal exemptions, twice

what they could have claimed on two single returns. At

a 31 percent tax rate, the larger exemption would have

reduced the couple's tax liability by $791 (31 percent of

$2,550). Second, the standard deduction of $6,700 on a

joint return would have been $2,700 more than the

$4,000 standard deduction the wife could have claimed

on an individual return. (The husband, filing individu-

ally with no income, could not take the deduction.) At

the couple's marginal tax rate of 31 percent, the larger

deduction would have reduced the couple's tax liability

by $837 (31 percent of $2,700). Finally, because tax

brackets for joint returns were wider than those for indi-

vidual returns, $16,100 that is taxed at 28 percent on

individual returns would have been taxed at only 15 per-

cent on a joint return and $5,050 taxed at 31 percent

rather than at 28 percent. Those lower tax rates would

have reduced the couple's tax liability by an additional

$2,245 (28 percent minus 15 percent equals 13 percent

of $16,100 plus 31 percent minus 28 percent equals 3

percent of $5,050). In combination, the three factors

would have lowered the couple's tax liability by 5.2 per-

cent of their adjusted gross income.

Husband

Wife Couple

Adjusted Gross Income $0 $75,000 $75,000

Less personal exemptions 2,550 2,550 5,100

Less standard deduction 4,000 4,000 6,700

Equals taxable income 0 68,450 63,200

Taxable at 15 percent 0 24,000 40,100

Taxable at 28 percent 0 34,150 23,100

Taxable at 31 percent 0 10,300 0

Tax Liability 0 16,355 12,483

Marriage Bonus $3,872

As a Percentage of Adjusted Gross Income 5.2

SOURCE: Congressional Budget Office.

15 percent equals 13 percent of $7,900). In combina-

tion, the two factors generate a marriage penalty of

$1,391 in higher taxes. Because the couple receives the

full value of personal exemptions whether or not they

are married, exemptions cannot contribute to their mar-

riage penalty.

Example: A Marriage Bonus

. The couple described

in the preceding example would have received a mar-

riage bonus of more than $3,500 in 1996 if all of the

couple's income were earned by one spouse

the wife in

this instance (see Box 2). Three factors generate the

bonus. First, because the husband has no earnings, he

would receive no benefit from the personal exemption

when filing as a single person. Filing jointly, the couple

thus receives an additional $2,550 exemption, worth

$791 in lower taxes (31 percent of $2,550). Second,

the married couple can claim a standard deduction that

is $2,700 greater than that which they can claim as sin-

gle filers

again because the husband, without earnings,

receives no value from the deduction as a single filer.

The additional standard deduction lowers the couple's

6 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

taxes by $837 (31 percent of $2,700). Finally, because

tax brackets for joint filers are wider than those for sin-

gle people, $16,100 more of the couple's income is

taxed at 15 percent rather than 28 percent, yielding tax

savings of $2,093 (28 percent minus 15 percent equals

13 percent of $16,100), and $5,050 is taxed at 28 per-

cent rather than 31 percent, saving another $152 (31

percent minus 28 percent equals 3 percent of $5,050).

The three factors combine to provide the couple with a

marriage bonus of $3,872 in lower taxes.

A Histor

of the Treatment

of Married Cou

les in the

Federal Income Tax

Since its inception in 1913, the federal individual in-

come tax has been periodically adjusted in its treatment

of married couples. Until 1948, the government for-

mally imposed income taxes on individuals, but couples

were allowed to file joint returns as early as 1918, even

though joint returns used the same schedules as individ-

ual returns. Individual taxation meant that marital sta-

tus had no effect on taxes and, as a result, married cou-

ples with identical total incomes could pay quite differ-

ent taxes, depending on the division of income between

husband and wife.

Community Property Laws

Community property laws in some states complicated

the situation by requiring that husbands and wives

share all of their income equally. The laws therefore

provided a basis for each spouse to pay federal income

taxes on one-half of the couple's income. That claim

equalized taxes for couples with the same income, but

created a situation in which some couples

those in

which husband and wife had different earnings re-

ceived a marriage bonus, paying lower taxes after mar-

riage than before. In 1930, the Supreme Court vali-

dated the splitting of income in community property

states [

Poe v. Seaborn

, 282 U.S. 101 (1930)], but de-

nied that opportunity to couples in other states, even if

they agreed by contract to share their incomes equally

[

Lucas v. Earl

, 281 U.S. 111 (1930)].

The distinction drawn between couples in states

that have community property laws and those that have

common property laws meant that couples with identi-

cal incomes could pay higher federal taxes in the latter

states than in the former. Because only a small fraction

of the population paid any income taxes before World

War II fewer than 4 million people owed any tax in

1939 the Congress felt little pressure to change this

situation during the 1930s. In fact, before 1948, the

Congress's only effort to change the filing status of cou-

ples came in 1941 in a proposal to require couples to

file joint returns at individual rates. The proposal failed

when opponents objected that the change would raise

the taxes paid by virtually all two-earner couples.

As income taxes increasingly became a means to

finance the war effort in the 1940s between 1939 and

1945, taxable returns increased more than tenfold to

nearly 43 million state governments more widely real-

ized that their citizens would pay lower federal taxes if

the states adopted community property laws. After fail-

ing to satisfy a skeptical Supreme Court with optional

community property laws [

Commissioner v. Harmon

323 U.S. 44 (1944)], Oklahoma and Oregon enacted

mandatory community property rules that met the

court's objections, even though couples were allowed to

opt out of their coverage. Other states quickly followed

suit. Concerns about problems that community prop-

erty laws could create in nontax areas brought pressure

on the Congress to act on the issue.

Joint Returns

The Congress responded in 1948 by allowing all mar-

ried couples to file joint returns and pay taxes on their

total income using schedules that had brackets twice the

width of those for single individuals. That approach

effectively assigned each spouse one-half of the cou-

2. This history of the treatment of couples in the federal individual in-

come tax draws from Boris I. Bittker, "Federal Income Taxation and

the Family,"

Stanford Law Review

, vol. 27 (July 1975), pp. 1389-

1463; June O'Neill, "Family Issues in Taxation," Rudy Penner, ed.,

Taxing the Family

(Washington, D.C.: American Enterprise Institute,

1983), pp. 1-22; Edward J. McCaffery, "Taxation and the Family: A

Fresh Look at Behavioral Gender Biases in the Code,"

UCLA Law

Review

, vol. 40 (April 1993), pp. 983-1060; Harvey S. Rosen, "The

Marriage Tax Is Down But Not Out,"

National Tax Journal

, vol. XL,

no. 4 (December 1987), pp. 567-575; and Gregg A. Esenwein,

Mar-

riage Tax Penalties and Bonuses Under the Federal Income Tax

CRS Report for Congress 93-475E (Congressional Research Service,

May 7, 1993).

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 7

ple's income and meant that a couple paid taxes equal

to twice what a single person would pay if he earned

one-half of the couple's total income. As a result, a sin-

gle taxpayer would pay higher taxes than a married

couple with the same total income. In effect, the new

law provided a marriage bonus for most couples, and

eliminated neutrality of tax laws with respect to mar-

riage. Under the new law, marriage would leave tax

liabilities unchanged only for couples who had no tax

liability or in instances in which husband and wife had

identical earnings.

In 1951, the Congress extended some of the advan-

tage that married taxpayers received from income-split-

ting to unmarried taxpayers supporting dependents by

establishing new schedules for "heads of household."

Taxpayers in this new category incurred tax liability

that was less than that of single taxpayers but greater

than that of married couples with the same total in-

come. The head-of-household filing status reduced

marriage bonuses for couples with children by lowering

the tax liabilities of single parents.

Although creating the joint filing status eliminated

the unequal treatment of married couples in different

states, it created sizable differences in the taxes paid by

single and married people. Predictably, unmarried peo-

ple viewed those differences as a tax on singles rather

than as a bonus for couples, an effect magnified by the

high marginal tax rates paid by upper-income taxpay-

ers. Under 1970 rates, for example, a single person

earning $20,000 paid $5,328 in income taxes, 42 per-

cent more than the $3,750 paid by a married couple

with the same total income. (About 15 percent of the

difference resulted from the two personal exemptions

allowed the married couple, whereas the single taxpayer

was allowed only one. The rest of the difference de-

rived from higher tax rates for single taxpayers.)

The Tax Reform Act of 1969 partially remedied the

disparity by limiting the tax liability of single people to

no more than 20 percent above that of married couples

with the same taxable income, effective for tax years

after 1970. For example, the change reduced the dis-

parity in taxes between a single taxpayer and a married

couple, each with $20,000 in total income, from 42 per-

cent to 24 percent. But the limitation created marriage

penalties for couples in which husband and wife had

similar incomes and left in place marriage bonuses

when the earnings split between spouses was less equal.

The Earned Income Tax Credit

The Tax Reform Act of 1975 created the Earned In-

come Tax Credit (EITC), a refundable credit available

to low-income taxpayers with earnings. Because indi-

vidual and joint filers have the same restrictions on

their incomes for receiving the credit, the EITC added a

potential penalty on married couples. If a low-income

person who qualified for the EITC got married, the

combined income of the couple could make them ineli-

gible for the credit. At the same time, the requirement

that credit claimants have a dependent child meant that

couples could receive a tax bonus if a low-earner mar-

ried a single parent with no earnings. The magnitude of

both marriage penalties and bonuses resulting from the

EITC was increased by subsequent tax acts that created

different credits for taxpayers with one and more than

one child, raised the maximum value of the credit, and

extended eligibility to childless families.

Recent Changes in Tax Law

Complaints from two-earner couples about the "mar-

riage tax" they had to pay led the Congress to reduce

the taxes for those couples as part of the Economic Re-

covery Tax Act of 1981. That act established a two-

earner deduction equal to 10 percent of the earnings of

the lower-earning spouse, up to a maximum deduction

of $3,000. The deduction reduced the marriage penalty

for all couples facing the penalty, and eliminated the

penalty entirely for some, but increased the size of the

marriage bonus for others. Even with the new deduc-

tion, however, some couples paid more than $4,000 in

added taxes because they filed joint returns.

The Tax Reform Act of 1986 repealed the two-

earner deduction as part of a broad reform that in-

creased the standard deduction for married couples and

collapsed the tax schedule from 15 brackets with a

maximum rate of 50 percent to just two brackets with a

maximum rate of 28 percent. (Married couples with

taxable income between $71,900 and $149,250 faced

tax rates of 33 percent as the act phased out the bene-

fits of the 15 percent bracket over that income range.

The 33 percent rate was further extended up to tax-

able income of $192,930 for a family of four by phas-

ing out the benefits from personal exemptions.) In

combination, those changes sharply reduced the mar-

riage penalty. For example, the act cut the marriage

8 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

penalty for a couple in which the husband and wife each

earned $50,000 by more than one-half in 1988, from

$2,609 to $1,284. Although the act did not eliminate

marriage penalties and bonuses, it reduced them enough

to quiet most critics.

Tax increases in 1990 and 1993 expanded the num-

ber of tax brackets from two to five, raised the maxi-

mum marginal tax rate to 39.6 percent, and sharply in-

creased the size and coverage of the earned income tax

credit. In combination, those changes imposed signifi-

cantly larger marriage penalties on both low- and high-

income families, and increased the size of the marriage

bonus for some couples. For example, a childless cou-

ple in which husband and wife each earned $300,000

faced a marriage penalty of nearly $16,000 after the

1993 act, compared with a penalty of about $1,900 un-

der the old law. At the other end of the income scale, a

one-earner couple with two children and earnings of

$10,000 saw their marriage penalty of $161 turn into a

bonus of nearly $500.

Those increases in the disparity between the taxes

paid by married couples and their unmarried counter-

parts have revived interest in reducing the effects of

marriage on tax liability. The "Contract with America

Tax Relief Act of 1995," as passed by the House of

Representatives, would have reduced marriage penalties

slightly by giving tax credits to couples incurring penal-

ties. To constrain the revenue losses from the provi-

sion, the act would have limited credits to $145. Given

that the penalty, under one broad measure, averages

nearly $1,400 for couples who incur it and can exceed

$20,000 for high-income couples, the credit would have

done little to offset the tax costs of marriage. The Sen-

ate in 1995 considered a proposal to set the standard

deduction for joint filers equal to twice that for single

filers. The change would have been phased in over a

10-year period. That proposal would have reduced or

eliminated penalties for taxpayers who do not itemize

their deductions. At the same time, it would have in-

creased bonuses for many couples and had no effect on

penalties for couples who continued to itemize deduc-

tions. The Balanced Budget Act of 1995, which in-

cluded the Senate provision, was vetoed by President

Clinton.

The United States is among a minority of devel-

oped nations that tax married couples. More than two-

thirds of the countries that are members of the Organi-

zation for Economic Cooperation and Development

(OECD) tax married couples as individuals. Only three

other developed countries tax couples jointly as does

the United States, and four others tax all family mem-

bers as a single entity (see Appendix A). Furthermore,

the trend has been moving away from joint taxation:

over the last two decades, 10 OECD members switched

from joint to individual taxation of married couples.

During the same period, no member country changed in

the other direction.

States vary more widely in taxing the incomes of

married couples (see Appendix B). Of the 42 states

that tax incomes, only 12 tax couples in ways that gen-

erate significant marriage penalties. Another 15 states

tax couples jointly, but use rate schedules that substan-

tially reduce or eliminate penalties. Nine states and the

District of Columbia allow couples to file individually

on a single return, thus providing them with tax savings

available to unmarried taxpayers. Finally, six states

impose a single tax rate that, in combination with stan-

dard deductions and exemptions, precludes all marriage

penalties.

The history of the changing taxation of couples and

single people demonstrates the tension between impos-

ing higher taxes on one group or the other with a pro-

gressive rate structure. Taxing individuals, as was the

case when the income tax began, avoids the problem of

different taxes based on marital status but runs afoul of

the principle of taxing married couples equally and of

state community property laws. The 1948 "solution" of

joint taxation dealt with both of those problems but

necessarily imposed higher taxes on single people in

relation to married couples, thus violating the principle

of marriage neutrality. The last five decades have wit-

nessed periodic movements between those two poles,

depending on the demands for fairness toward singles

or couples. As long as the Congress pursues the three

mutually incompatible goals of marriage neutrality,

equal treatment of couples with similar incomes, and

progressive tax rates, the tension between the two posi-

tions will continue.

3. Joint Committee on Taxation,

General Explanation of the Tax Re-

form Act of 1986

, Joint Committee Print JCS-10-87 (May 4, 1987),

p. 19.

4. Daniel R. Feenberg and Harvey S. Rosen, "Recent Developments in

the Marriage Tax,"

National Tax Journal

, vol. 47, no. 1 (March

1995), p. 94.

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 9

The Si

nificance of Marria

Penalties and Bonuses

Does it matter that marriage penalties and bonuses arise

from a tax code designed to meet the goals of equal tax

liabilities for married couples with the same incomes

and progressive taxation? Three issues are paramount

in answering that question: the fairness of tax levies

among couples, the effects of taxes on how much peo-

ple choose to work, and the effects of taxes on marriage

and divorce.

Fairness

Equal treatment under the law is a basic concept of fair-

ness. Horizontal equity in taxation stipulates that peo-

ple who have equal ability to pay receive equal treat-

ment and therefore bear the same tax burden. Al-

though determining ability to pay may be difficult in

practice, the concept of horizontal equity helps in eval-

uating the conflicting goals of the tax system that lead

to marriage penalties and bonuses. On the one hand, if

the act of marriage has no effect on a couple's ability to

pay, horizontal equity implies that taxes be marriage

neutral

that is, that couples pay the same taxes

whether or not they are married. On the other hand, if

they are otherwise similar, married couples with the

same total income must pay the same taxes to satisfy

the criterion of equity. In the face of progressive tax

rates, however, the two goals conflict (see "Defining

Marriage Penalties and Bonuses" above). Reconciling

them demands investigation of marriage penalties and

bonuses.

One issue that must precede the question of hori-

zontal equity involves the appropriate unit on which to

levy taxes. If individuals provide their own resources

and do not share with others, the most appropriate unit

would be the individual. If, however, people share re-

sources and consume goods and services in groups,

those groups are the appropriate units for the purpose

of taxation. The federal income tax code assumes that

married couples combine their incomes and other re-

sources to support themselves and their dependents,

and therefore imposes taxes on the joint incomes of

spouses. The code stops short of defining all tax units

on the basis of shared incomes. On the one hand, mar-

ried couples are presumed to combine resources and

thus must file jointly. On the other hand, people in re-

lationships not based on legal marriage or other family

relationships must pay taxes on their separate incomes,

even if they consume together. Unmarried couples or

two members of the same family sharing a home are

taxed as individuals.

Despite the implicit assumptions built into the tax

code, marriage may or may not affect a couple's ability

to pay, depending on whether the couple's costs of liv-

ing change. If the couple lived together before mar-

riage, simply getting married would have little or no

effect on their costs and hence on their ability to pay. If

they create a new household after marriage, their costs

could rise or fall. If, for example, each person lived

with his or her parents before the wedding, marriage

could increase their costs, but if each had maintained a

separate household, combining those households into

one could reduce costs. Because living arrangements

before and after marriage need not differ, the tax sys-

tem might logically assume that marriage has no intrin-

sic effect on a couple's ability to pay and therefore that

marriage neutrality would be desirable. The conclusion

that marriage should not affect a couple's tax bill says

nothing about the fairness of taxation of either married

or unmarried couples.

Whether imposing the same taxes on married cou-

ples with equal incomes satisfies the principle of hori-

zontal equity is more problematic. Two childless cou-

ples who have the same total income, earned in each

case by two full-time workers, may also have the same

ability to pay, regardless of the division of income be-

tween spouses. If one couple had only one worker,

however, the couples would have quite different abili-

ties to pay because the nonworking spouse could pro-

vide services in kind that are not as readily available to

the two-earner couple. But if the working spouse of the

one-earner couple held two different jobs, the couple's

ability to pay would be different yet. In essence, as is

more generally true, income alone provides a poor mea-

sure of ability to pay. Requiring that married couples

with a given amount of total income pay the same taxes

may therefore not satisfy the principle of horizontal

equity.

Children complicate the issue. Tax rules can be

totally indifferent to the presence of children, treating

the decision to have and rear children as entirely the

10 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

responsibility of the parents. If, however, society val-

ues children and chooses to subsidize parents who rear

them, tax rules, as in the current tax code, may be used

to provide tax relief to families with children. Further-

more, the government may wish to encourage couples

to have children by not taxing any resources that are

provided by parents who stay home to care for them.

Under that assumption, equal taxation of one- and two-

earner couples who have the same total incomes may be

appropriate.

The movement of people through a variety of tax

statuses during their lifetimes and their consequent lia-

bility for different amounts of tax over time further

complicate issues of fairness. A couple may change

from being two unmarried individuals with earnings, to

a two-earner childless married couple, to a one-earner

couple with one spouse caring for children, to a two-

earner couple purchasing child care, to a two-earner

couple with older children who can care for themselves,

and back to a two-earner childless couple. The individ-

ual income tax would treat the couple differently in

each period, whether or not their ability to pay changes.

Depending on the division of earnings between spouses,

the couple may pay higher or lower taxes because they

are married, even if they experience no change in ability

to pay. The various features of the tax code affect peo-

ple differently throughout their lives, sometimes reduc-

ing their tax liability and other times increasing it. As-

sessing whether they are taxed fairly depends on

whether their situation is considered in any given year

or over their entire lifetime as a couple. What may ap-

pear to be unfair in one year may appear quite different

when viewed in a lifetime context.

The conflict between the two goals of marriage

neutrality and taxing couples with equal incomes the

same would matter little if the marital penalties and

bonuses generated were small. To determine the impor-

tance of those effects in terms of both their size and

the number of couples affected requires a close exami-

nation of who pays the penalties and who receives the

bonuses.

Effects on the Supply of Labor

Taxing married couples may significantly affect the

amount of paid work done by the second worker in each

couple because a husband's and wife's incomes are

combined to determine tax liability. For example, if a

man earns $75,000 and his wife decides to take a job,

the couple will pay 28 percent of the wife's earnings to

the federal government as income taxes, a further 15.3

percent in Social Security taxes (counting both em-

ployer and employee shares), and, possibly, a state or

local income tax. Because the husband's income has

already moved the couple into the 28 percent income

tax bracket, the first dollar of the wife's income bears a

total tax rate greater than 43 percent. If she was single,

the wife would pay no income tax on the first $6,550

that she earned and only a 15 percent income tax on the

next $24,000. Including the 15.3 percent Social Secu-

rity payroll tax, her marginal tax rate would be either

15.3 percent or 30.3 percent (not counting any state or

local taxes), 28 percentage points or 13 percentage

points lower than the 43.3 percent rate she incurs as a

married person. The higher initial tax rate she faces

when married reduces the value of her work and thus

may induce her to work fewer hours each week, fewer

weeks each year, or even not to work at all. At the

same time, the husband's income may be taxed at a

lower marginal rate than it would be if he was not mar-

ried, yielding a higher after-tax wage that might lead

him to work more than he otherwise would choose to

do. If he was single, the husband would face a marginal

income tax rate of 31 percent, higher than the 28 per-

cent rate he faces when married. However, his average

tax rate is also lower in marriage, raising his after-tax

income and thus encouraging him to work less. The net

effect on his labor supply is uncertain.

Taxes exert two countervailing influences on an

individual's decision about how much to work. Com-

pared with a situation in which there are no taxes on

earnings, a tax reduces both the after-tax hourly wage

of workers and their total after-tax income. Because

people will choose to work more hours as long as their

net hourly wage exceeds their value of an extra hour of

leisure, the tax's lowering of net wages will lead indi-

viduals to work fewer hours. At the same time, because

people's income falls at any number of hours worked

5. For further detail on the effects of taxes on the work patterns of men

and women, see Congressional Budget Office,

Labor Supply and

Taxes

, CBO Memorandum (January 1996).

6. Most economists believe that the couple bears the employer's share of

the payroll tax on the wife's earnings in the form of wages that are

lower than they would otherwise be. That tax thus affects after-tax

wages, the basis on which the wife may decide whether to go to work.

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 11

as a result of the tax, they will want to consume less

leisure and work more hours. The net effect depends on

the relative magnitudes of the two offsetting pressures.

Economists refer to those influences as the substitution

effect and the income effect, respectively.

Research into the sensitivity of people's labor sup-

ply to their after-tax wages offers a mixed picture of

how higher marginal tax rates on married couples affect

their decisions about whether and how much to work.

Economist James Heckman emphasized the need to

distinguish between the effects of increased net wages

on the hours worked by people already employed and

the decision by people not working to take a job. Rob-

ert Triest found that the decisions of both husbands and

wives about the number of hours worked depend little

on after-tax wages once the individual has decided to

work at all. But the decision to work appears to be

much more responsive to changes in net wages, and

more so for wives than for husbands. Studies of these

issues encounter a range of methodological and mea-

surement problems, however, and the magnitude of the

effects is uncertain.

Some workers appear to react more strongly to

changes in their net wages than others. Economists

Chinhui Juhn, Kevin Murphy, and Robert Topel found,

for example, that low-wage workers are much more

sensitive to their net pay than are higher-wage work-

ers. Their data suggest that a 10 percent increase in

after-tax wages would lead workers in the bottom one-

fifth of the wage distribution to increase the hours they

worked by about 2.5 percent, and that those in the top

two-fifths would work only about 0.5 percent more. By

contrast, at the top end of the income scale, economist

Nada Eissa found evidence that the labor supply of

wives in upper-income families is highly sensitive to

after-tax wages. Her analysis suggested that wives in

the top 1 percent of income distribution would increase

the hours they worked by 8 percent in response to a 10

percent increase in net wages. Neither of those results,

however, applies to married couples as a group, and

studies that have attempted to measure the magnitude

of the effects of wages on work have identified a wide

range of values.

An international comparison of the labor supply of

wives under joint and individual taxation indicates that

joint taxation reduces the participation of women in the

labor force. Economist Siv Gustafsson compared the

labor supply among married women under Germany's

joint and Sweden's individual income tax systems.

She estimated that imposing the German joint taxation

approach on Sweden would have caused the participa-

tion rate of Swedish wives in the labor force to fall by

one-fourth, whereas subjecting German wives to a

Swedish-style tax system would have raised the parti-

cipation rate of married German women in the labor

force by one-fifth. That conclusion was bolstered by

Gustafsson's finding that participation of Swedish

wives in the labor force rose markedly following the

replacement of joint with individual taxation in 1971.

Although other factors surely contributed to the in-

crease, and although the rise occurred gradually over a

number of years as families adapted to the new tax sys-

tem, the rate of participation of married women in the

Swedish labor force climbed from 53 percent in 1969 to

74 percent a decade later. By comparison, Germany,

which had joint taxation throughout the period, experi-

enced one-third the increase of Sweden, from 39 per-

cent to 46 percent.

Filing a joint tax return in the United States can

result in higher, lower, or unchanged marginal tax rates

for both husband and wife, depending on their total in-

7. James J. Heckman, "What Has Been Learned About Labor Supply in

the Past Twenty Years?"

American Economic Review Papers and

Proceedings

, vol. 83, no. 2 (May 1993), pp. 116-121.

8. Robert K. Triest, "The Effect of Income Taxation on Labor Supply in

the United States,"

The Journal of Human Resources

, vol. XXV, no.

3 (1990), pp. 491-516.

9. Randolph and Rogers discuss these issues and the reasons they are

unsettled in William C. Randolph and Diane Lim Rogers, "The Impli-

cations for Tax Policy of Uncertainty About Labor Supply and Sav-

ings Responses,"

National Tax Journal

, vol. 48, no. 3 (September

1995), pp. 429-446.

10. Chinhui Juhn, Kevin M. Murphy, and Robert H. Topel, "Why Has the

Natural Rate of Unemployment Increased Over Time?"

Brookings

Papers on Economic Activity

, no. 2 (1991), pp. 75-126. In the same

publication, however, Janet Yellen countered that their estimates were

too high because they failed to account for changes in government

transfer programs (pp. 127-133).

11. Nada Eissa,

Taxation and Labor Supply of Married Women: The

Tax Reform Act of 1986 As a Natural Experiment

, Working Paper

No. 5023 (Cambridge, Mass.: National Bureau of Economic Re-

search, February 1995). Eissa found that about one-half of the rise

resulted from nonworking wives entering the labor force and one-half

from employed wives working longer hours.

12. Siv Gustafsson, "Separate Taxation and Married Women's Labor Sup-

ply: A Comparison of West Germany and Sweden,"

Journal of Popu-

lation Economics

, vol. 5, no. 1 (February 1992), pp. 61-85.

12 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table 1.

Effects on Marginal Tax Rates of Married Couples Filing Jointly Rather than Individually,

Projected 1996 (In percentage of married couples)

Marginal Tax Rate

of Spouse with Marginal Tax Rate of Spouse with Lower Earnings

Higher Earnings Lower Unchanged Higher All Couples

Lower 3 3 18 24

Unchanged 7 14 43 64

Higher 1

1 11 12

All Couples 11 17 72 100

SOURCE: Congressional Budget Office simulations.

come and how much is earned by each spouse. Esti-

mates for 1996 suggest that just 14 percent of married

couples filing jointly will face the same marginal tax

rate that each spouse would incur if they could file as

individuals (see Table 1). Filing jointly rather than in-

dividually will lead to reduced marginal tax rates for the

higher-earning spouse in 24 percent of all couples, un-

changed marginal rates in 64 percent, and higher mar-

ginal rates in 12 percent. At the same time, joint versus

individual filing raises marginal tax rates of lower-earn-

ing spouses in 72 percent of couples, lowers rates in 11

percent of couples, and has no effect in 17 percent.

The Congressional Budget Office (CBO) used

microeconomic simulations to estimate the effects of

requiring joint filing on the work behavior of husbands

and wives. Using values for substitution and income

elasticities from the economic literature on labor sup-

ply, CBO estimated how much each husband and wife

would work if they faced individual marginal tax rates

based on their own incomes rather than the rates that

apply when they file joint tax returns.

The simulations indicated that joint tax rates moti-

vate higher-earning spouses to work between 0.1 per-

cent and 0.3 percent

than they would if the tax

code required them to file as individuals. That result is

consistent with the observation that joint filing is more

likely to lower tax rates for primary earners than to

raise them. The reverse holds for secondary workers.

Simulations indicated that generally higher tax rates for

lower-earning spouses prompt them to work between 4

percent and 7 percent

less

than they would if they could

file individually. Overall, requiring couples to file joint

tax returns induces them to work less. As a result, their

total earnings are between 0.7 percent and 1.2 percent

less than they would otherwise be.

Effects on Marriage and Divorce

Couples decide to marry or divorce on the basis of the

trade-offs between a broad array of benefits and costs

of marriage. Social pressures, the desire for compan-

ionship, love, wanting to rear children, and economic

considerations may induce people to marry, whereas

issues of privacy and psychological and monetary costs

may keep couples from marrying or induce them to seek

divorce. The tax costs or benefits of marriage are of

highly uncertain but almost surely slight impor-

tance. For most couples, other factors almost cer-

tainly dominate tax considerations in the marriage deci-

13. Elasticities used for estimates of labor supply response were taken

from Congressional Budget Office,

Labor Supply and Taxes

. The

ranges used were primary worker substitution elasticity, 0.1 to 0.2;

primary worker income elasticity, 0 to -0.1; secondary worker substi-

tution elasticity, 0.6 to 0.9; and secondary worker income elasticity,

-0.2 to -0.3.

14. The potential loss of means-tested benefits that many low-income

mothers receive may be much more important. If marriage disqualifies

a family for food stamps, housing assistance, and Medicaid benefits,

the dollar cost is substantial. Especially in communities in which un-

married mothers are common, those costs could well outweigh other

factors and generate significant disincentives to marry.

CHAPTER ONE MARRIAGE PENALTIES AND BONUSES 13

sion, as evidenced by the substantial numbers of cou-

ples who bear large marriage penalties in the form of

higher taxes. Even so, couples for whom the decision

to marry is close that is, at the economist's margin,

where minor factors can carry the day may decide that

being married is not worth having to pay thousands of

dollars of additional taxes every year. Other couples

may face bonuses that provide just enough additional

inducement to lead them down the aisle. Similar issues

may enter into couples' decisions to divorce.

Economists investigating the impact of tax penal-

ties on marriage have found only small or no effects.

One study, for example, found that a 10 percent in-

crease in all marriage penalties would decrease the frac-

tion of adult women under 45 years old who are married

by less than 0.5 percent. Another study, focusing on

the likelihood of single adult women marrying in a

given year, detected no marriage penalty effect, but did

find indications that some couples appeared to delay

marriages for a few months to avoid paying higher

taxes for a single year.

One test of the effects of marriage penalties on

whether couples wed is to compare the bonuses and

penalties for married couples with those that unmarried

couples would incur if they were to marry. Such a com-

parison, based on data from the March 1996 Current

Population Survey, indicates that unmarried couples

would be more likely than married couples to incur pen-

alties if they were married, although only marginally so

Table 2.

Marriage Penalties and Bonuses for Married

and Unmarried Couples, Simulated 1995

Married Unmarried

Couples Couples

Number of Couples (Thousands)

With Penalties 26,800 1,500

Unaffected 6,400 300

With Bonuses 21,500

900

Total 54,700 2,700

Couples in Category (Percent)

With Penalties 49 56

Unaffected 12 11

With Bonuses 39

Total 100 100

Average Penalty

In Dollars 1,200 770

As a Percentage of Income 1.9 1.6

Average Bonus

In Dollars 1,140 1,020

As a Percentage of Income 2.3 2.7

SOURCE: Congressional Budget Office simulations based on data

from the March 1996 Current Population Survey.

a. Unmarried couples are those living together and reporting as part-

ners.

(see Table 2). In 1995, 56 percent of unmarried cou-

ples would have incurred penalties if they had been

married, compared with 49 percent of married couples.

Respective values for marriage bonuses would have

been 33 percent and 39 percent.

The foregoing comparisons might suggest that

some couples may elect not to marry if doing so would

bring them tax penalties. However, average penalties

for married couples are much greater than the hypothet-

15. Leslie Whittington and James Alm found small effects of income taxes

on decisions to divorce, particularly for women. Whittington and Alm,

"'Til Death Do Us Part: The Effect of Income Taxation on Divorce,"

forthcoming in the

Journal of Human Resources

(Spring 1997).

16. James Alm and Leslie A. Whittington, "Does the Income Tax Affect

Marital Decisions?"

National Tax Journal

, vol. 48, no. 4 (December

1995), pp. 565-572. In more recent work, the authors find small but

statistically significant effects of marriage penalties in reducing the

likelihood of marriage for women but not for men. Alm and

Whittington, "For Love or Money: The Impact of Income-Taxes on

Marriage" (unpublished working paper, University of Colorado at

Boulder, February 1997).

17. David L. Sjoquist and Mary Beth Walker, "The Marriage Tax and the

Rate and Timing of Marriage,"

National Tax Journal

, vol. XLVIII,

no. 4 (December 1995), pp. 547-558. James Alm and Leslie

Whittington also found small but significant effects of penalties delay-

ing marriage. Alm and Whittington, "Income Taxes and the Timing of

Marital Decisions," forthcoming in the

Journal of Public Economics

(1997).

18. Unmarried couples are unmarried adults who identified themselves to

Current Population Survey interviewers as partners. Couples could

be of the same or opposite sexes. Only 3 percent were same-sex cou-

ples. See Appendix C.

14 FOR BETTR OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

ical penalties for unmarried couples would be $1,200

versus $770 although penalties for the two groups

measured as a percentage of income are less diver-

gent 1.9 percent versus 1.6 percent. In part, the dif-

ference may arise from the different characteristics of

the two groups (see Appendix C). Unmarried couples,

with a median age of 35, are younger than their married

counterparts (median age 47) and have a lower median

income: $35,000 versus $45,000. Because their in-

comes are lower, unmarried couples may not be as will-

ing to incur marriage penalties as wealthier couples,

even if those penalties represent a smaller fraction of

their incomes. In addition, younger couples may be less

averse to living together outside of marriage than older

couples and thus may feel less pressure to marry. For

most couples, however, potential marriage penalties

and bonuses are not sufficient cause to alter marriage

decisions.

Chapter Two

Sources of Marriage

Penalties and Bonuses

arried couples incur tax penalties or receive

tax bonuses because the tax code treats them

differently from unmarried taxpayers.

Whether a couple has a penalty or a bonus depends on

the level and sources of their income, the division of

income between husband and wife, and the presence or

absence of children. Understanding marriage penalties

and bonuses requires an examination of the factors

within the law that generate differences in the taxes

paid by married and unmarried couples.

Factors Givin

Rise to

Marria

e Penalties and

Bonuses

Marriage penalties and bonuses result from factors that

affect either the level of taxable income or the tax rate

applied to that income. A penalty or a bonus occurs

whenever a couple's exemptions, deductions, tax brack-

ets, or credits differ from those of two otherwise similar

single people (see Table 3). Such differences occur

because the individual income tax sets parameters in

dollars. Those dollar amounts for married couples of-

ten differ from the sums that would apply to spouses if

they were single.

The various components of the income tax system

interact in determining whether couples incur penalties

or bonuses. They also determine the size of those tax

differentials. Certain parts of the tax system may im-

pose potentially large marriage penalties, but may do so

for relatively few couples. The following analysis ex-

amines only the hypothetical impact of each component

on the taxes married couples pay, without regard to

how many couples each component affects (see Appen-

dix Table D-1).

The Standard Deduction

In 1996, single filers could claim a standard deduction

of $4,000, married couples $6,700, and heads of house-

hold $5,900. Most taxpayers claimed the standard de-

duction, presumably because its value exceeded the de-

ductions they could itemize, such as state taxes or home

mortgage interest payments. In 1994, only 28 percent

of all individual tax returns itemized their deductions.

If each can use the full deduction, two single filers

claiming standard deductions could reduce their 1996

taxable income by $8,000 $1,300 more than the

$6,700 deduction for joint filers. Differences in the

standard deduction among filing statuses would thus

impose a marriage penalty on such a couple that is

equal to the tax on that additional taxable income, the

amount of which depends on the couple's marginal tax

1. For a comprehensive discussion of provisions of the federal individual

income tax code that give rise to marriage penalties and bonuses, see

General Accounting Office,

Income Tax Treatment of Married and

Single Individuals

, GAO/GGD-96-175 (September 1996).

16 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table 3.

Factors Determining Whether Couples Face Marriage Penalties or Bonuses, 1996

Conditions Leading Conditions Leading

Tax Parameter or Feature to Marriage Penalty to Marriage Bonus

Personal Exemptions

($2,550 for all individuals,

regardless of marital status.)

None. One spouse cannot use full sin-

gle exemption but other spouse

would have positive taxable in-

come if taxed as an individual.

Standard Deduction

($4,000 for singles,

$6,700 for couples.)

Combined use of two single

deductions exceeds value of

married deduction.

One spouse cannot use full sin-

gle deduction but other spouse

would have positive taxable in-

come if taxed as an individual.

Tax Brackets

(Lower brackets for singles are

60 percent as wide as those for

couples; top bracket starts at

same income for all.)

Spouses have more nearly equal

incomes; as married couple,

more of combined income taxed

at higher rate; high earners have

more income subject to top tax

rate.

Spouses have unequal incomes;

as singles, income of higher-

earning spouse taxed at higher

rate.

Earned Income Tax Credit

(Parameters same regardless of

filing status.)

Low-earning parent married to

spouse whose income causes

loss of some or all of earned

income tax credit.

Low-earning childless person

married to parent with no or very

low earnings.

Phaseout of Personal

Exemptions (Starting income for

singles equals two-thirds of that

for couples.)

Spouses have more nearly equal

incomes; as married couple,

more of total income falls in

phaseout range.

Spouses have unequal incomes;

as singles, more income of

higher-earning spouse subject to

phaseout.

Limitation on Itemized Deduc-

tions (Starting point same

regardless of filing status.)

Spouses have more nearly equal

incomes; as married couple,

more of total income falls in limi-

tation range.

None.

Other Fixed Dollar Limitations

(For example, income limit for

individual retirement accounts,

thresholds for taxation of Social

Security.)

Either marriage does not in-

crease limit or increase is less

than spouse adds to measure

subject to limit.

Marriage increases limit and one

spouse adds less to measure

subject to limit than the increase

in limit.

SOURCE: Congressional Budget Office.

CHAPTER TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 17

rate. A couple in the 28 percent tax bracket pays $364

(28 percent of $1,300) more in taxes than two compa-

rable single filers, whereas a couple in the 15 percent

bracket pays $195 more. The standard deduction can-

not create a marriage penalty for couples who itemize

their deductions and would do so even if they were not

married.

The decision to itemize deductions can depend,

however, on whether a couple is married. Consider, for

example, two taxpayers who can each itemize $3,600

worth of deductions if they are not married and file sin-

gly. Because the standard deduction $4,000 exceeds

their itemizable deductions, each would choose to take

the standard deduction and the total value of their de-

ductions would equal $8,000. If they were married,

their total itemized deductions would be $7,200, more

than the standard deduction of $6,700 allowed for joint

returns, and they would elect to itemize their deduc-

tions. Even though they itemize, they incur a marriage

penalty because being married reduces their total de-

ductions by $800. That penalty, however, results not

from itemizing but from the difference in standard de-

ductions for singles and couples.

The effect of standard deductions is greater if either

spouse, were the couple not married, could file as head

of household. The 1996 standard deduction for heads

of household is $5,900, therefore, two heads of house-

hold could claim total deductions of $11,800 $5,100

more than the $6,700 available to joint filers.

The standard deduction can create marriage bo-

nuses for couples in which one spouse has a low income

and cannot use the full value of the standard deduction

as a single filer. That spouse brings to a marriage his

or her unused deduction, which can serve to reduce the

couple's tax liability.

Personal Exemptions

Because their value does not vary by filing status, per-

sonal exemptions cannot contribute to marriage penal-

ties. They can, however, provide marriage bonuses for

spouses who could not use the full value of their ex-

emptions if they were not married and thus had to file

separately. In 1996, for example, the personal exemp-

tion of $2,550 per taxpayer reduced the taxable income

of a one-earner childless couple by $2,550 because they

filed jointly rather than as single individuals. If they

were not married, the nonworker would pay no taxes

with or without the exemption and would therefore re-

ceive no value from it.

Tax Brackets

Because tax brackets for joint tax filers are wider than

those for either single filers or heads of household but

not twice as wide they can generate either marriage

penalties or bonuses, depending on the division of in-

come between husband and wife. Couples in which one

spouse earns most of the income generally receive mar-

riage bonuses because the wider tax brackets for joint

filers impose lower taxes on the income of the higher-

earning spouse than do the narrower brackets for indi-

viduals filing as single or head of household (see Box

3). In 1996, those bonuses exceeded $2,000 for a cou-

ple with $40,100 of taxable income and nearly $4,600

for couples with taxable income of $147,700 or more.

Couples in which husband and wife earn more

nearly equal shares of total income generally incur mar-

riage penalties because the joint tax brackets are less

than twice the width of single brackets (see Box 4). In

1996, married couples with an equally divided taxable

income of $48,000 paid about $1,000 more in taxes

than they would have if they had not filed joint returns.

The penalty rises to just over $1,600 for couples with a

taxable income of $116,300. Two-earner couples who

have very high and equal incomes pay the largest mar-

riage penalties, since the top rate bracket begins at the

same level of taxable income for all filing statuses

$263,750 in 1996. As a result, as much as $263,750 of

joint income can be taxed at 39.6 percent rather than 36

percent, raising a couple's taxes by a maximum of al-

most $9,500 more.

The Earned Income Tax Credit

Couples who have low earnings may incur a marriage

penalty or bonus because the size of the earned income

tax credit does not depend on a family's filing status but

does vary with the number of children in the family. As

a result, combining the incomes and children of both

spouses on one return can increase or reduce the credit

As a Percentage of

Taxable Income

(Right scale)

Marriage

Bonus

(Dollars)

Marriage

Bonus

(As a

percentage

of taxable

income)

Taxable Income (Dollars)

24,000

40,100

58,150

96,900

121,300

147,700

263,750 600,000

Filing Individually

Filing Jointly

Tax Brackets and Rates (In percent)

15 28

39.615 28

39.6

5,000

10,000

15,000

5 Percentage

Point

Differential

3 Percentage

Point

Differential

Dollars

(Left scale)

13 Percentage

Point

Differential

18 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box 3.

1996 Marriage Bonuses Resulting from Tax Brackets for One-Earner Couples

Tax brackets create marriage bonuses when the mar-

ginal tax rate of an individual return exceeds that for a

joint return with the same taxable income. For example,

when a couple's taxable income is between $24,000 and

$40,100, all earned by one spouse, the marginal tax rate

that is, the tax rate on the last dollar of income on a

joint return is 15 percent, whereas that on an individual

return is 28 percent. The 13 percentage point differen-

tial yields a marriage bonus of 13 cents for each dollar

of taxable income in that range, over which the bonus,

measured in dollars, rises from zero to $2,093 (13 per-

cent of $40,100 minus $24,000).

The bonus also rises when taxable income falls be-

tween $58,150 and $96,900 (when the differential be-

tween tax rates is 3 percentage points) and between

$121,300 and $147,700 (when the differential is 5 per-

centage points). In all other ranges of taxable income,

marginal tax rates do not differ for joint and single re-

turns and marriage bonuses do not change.

SOURCE: Congressional Budget Office.

they would receive in relation to what they would get if

they could file single or head-of-household returns.

The credit follows three schedules for childless

taxpayers, those with one child, and those with two or

more children that do not depend on filing status; that

is, the credit is the same for joint filers, heads of house-

hold, and single filers. Under each schedule, the credit

increases with earnings to a maximum, is constant at

that maximum for a range of earnings, and then de-

clines to zero (see Figure 2). Beginning with the 1995

tax year, tax units claiming the EITC also had to satisfy

an asset test that denies the credit to families with more

than $2,350 of income from interest, dividends, and

39.6

Filing Individually

Filing Jointly

Tax Brackets and Rates (In percent)

15 28

39.615 28

3.6 Percentage

Point

Differential

13 Percentage

Point

Differential

3 Percentage

Point

Differential

5 Percentage

Point

Differential

As a Percentage of

Taxable Income

(Right scale)

Marriage

Penalty

(Dollars)

Marriage

Penalty

(As a

percentage

of taxable

income)

Taxable Income (Dollars)

5,000

10,000

15,000

Dollars

(Left scale)

48,000 116,300 242,600

600,000

40,100 96,900 147,700 263,750 527,500

CHAPTER TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 19

Box 4.

1996 Marriage Penalties Resulting from Tax Brackets for

Two-Earner Couples in Which Spouses Have Equal Earnings

Tax brackets create marriage penalties when the mar-

ginal tax rate of a joint return exceeds that for an indi-

vidual return with the same taxable income. For exam-

ple, when a couple's taxable income is between $40,100

and $48,000, earned equally by both spouses, the mar-

ginal tax rate

that is, the tax on the last dollar of in-

come

on a joint return is 28 percent, whereas that on

two individual returns is 15 percent. The 13 percentage

point differential yields a marriage penalty of 13 cents

for each dollar of taxable income in that range, over

which the penalty, measured in dollars, rises from zero

to $1,027 (13 percent of $48,000-$40,100).

The penalty also rises when taxable income falls be-

tween $96,900 and $116,300 (when the differential be-

tween tax rates is 3 percentage points), between

$147,700 and $242,600 (when the differential is 5 per-

centage points), and between $263,750 and $527,500

(when the differential is 3.6 percentage points). In all

other ranges of taxable income, tax rates do not differ

for joint and individual returns and marriage bonuses do

not change.

SOURCE: Congressional Budget Office.

rent. That limitation could impose a marriage penalty

as well.

In 1996, the credit was largest for taxpayers with

two or more children, reaching a maximum of $3,556.

Families with one child could receive up to $2,152 and

0102030

1,000

2,000

3,000

4,000

Earned Income Credit (In dollars)

Two or More Children

One Child

No Children

Taxpayer's Earnings or Adjusted Gross Income

(In thousands of dollars)

20 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

childless taxpayers up to $323. For families with two

children, the credit equaled 40 percent of earnings up to

$8,890, was constant at $3,556 when earnings fell be-

tween $8,890 and $11,610, and declined by 21.06 per-

cent of AGI exceeding $11,610, disappearing entirely

when AGI was $28,495 or greater. (All parameters

apply to 1996; dollar values are adjusted for inflation

annually.) For families with one child, the maximum

credit of $2,152 applied for earnings between $6,330

and $11,610, and phased out completely when AGI

reached $25,078. Childless families received the maxi-

mum credit of $323 when their earnings fell between

$4,220 and $5,280 and received no credit when their

AGI exceeded $9,500.

Two aspects of the EITC lead to marriage penalties

and bonuses: the credit is the same for all filing sta-

tuses but differs according to the number of children.

Consider first the effect of fixed credit parameters

across filing statuses. Spouses who could qualify indi-

vidually for credits could lose some or all of them when

filing jointly because their combined income would

Figure 2.

Earned Income Tax Credit by Number of Children

and Earnings, 1996 Tax Law

SOURCE: Congressional Budget Office.

move them closer to or beyond the income at which the

credit becomes zero. In addition, when filing individu-

ally, spouses could qualify for two separate credits, not

just the one they can receive as joint filers. In 1996, for

example, a couple with two children in which both hus-

band and wife earn $11,000 incurred a marriage penalty

of more than $2,900 from the EITC alone (see Box 5).

Filing jointly in 1996, the couple qualified for a single

credit of $1,368. If the couple was not married and

each spouse filed as a head of household with one child,

each would have received the maximum EITC of

$2,152. Two factors cause the reduction in the credit,

which was more than 13 percent of the couple's earn-

ings. First, combining the husband's and wife's earn-

ings puts their total income into the range over which

the credit declines. Second, the couple would qualify

for two separate credits if they could file individually

but can receive only one credit as joint filers.

The EITC can also impose a marriage penalty on

higher-income two-earner couples if the division of

earnings between spouses is very different. A couple

with two children and $50,000 in earnings, split 80 per-

cent and 20 percent between husband and wife, would

receive no EITC when filing jointly and pay total taxes

of $4,965. If they were not married, however, the

lower-earning member of the couple could claim the

maximum one-child credit of $2,152, based on having

only $10,000 of earnings. Furthermore, each member

could file as head of household, and differences in us-

able standard deductions and personal exemptions

would further reduce the couple's separate taxes by

$615 below the amount they would pay if they filed

jointly, yielding a total marriage penalty of $2,767.

The higher EITC for families with more children

can yield marriage bonuses for couples with low earn-

ings. Combining two tax units can allow the resulting

couple to claim an EITC that is greater because taxpay-

ers with one child receive a larger percentage credit

than those with no children and families with two or

more children receive larger percentages than those

with only one child. For example, if a couple had one

child, the husband earned $6,000 in 1996, and the wife

did not work, they would qualify for an EITC of

$2,040. If they were not married and the child was as-

sociated with the mother, their EITC would be $268, all

going to the husband. Joint filing thus yields a bonus

of nearly $1,800 30 percent of their earnings. That

difference occurs because taxpayers with no children

CHAPTER TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 21

Box 5.

A Marriage Penalty for a Low-Income Couple with Children

A couple with two children and $22,000 in total earn-

ings, split evenly between the husband and the wife,

would have incurred a marriage penalty of $3,701 under

1996 tax law. The penalty results from two factors.

First, if they were not married, both the husband and the

wife could file as heads of household, each claiming one

child as a dependent. As heads of household, their com-

bined standard deductions would have been $11,800,

$5,100 more than the $6,700 standard deduction avail-

able on a joint return. At the couple's marginal tax rate

of 15 percent, the lower deduction would have increased

the couple's tax liability by $765 (15 percent of $5,100).

Second, filing separate returns, the husband and wife

each could have claimed the maximum earned income

tax credit (EITC) for a filer with one child, $2,152. Fil-

ing jointly, the couple would have received only one,

smaller EITC of $1,368. Thus, filing jointly the couple

would have received a payment of $603, about $3,700

less than the $4,304 they would have gotten if they

could have filed separately.

Husband

Wife Couple

Adjusted Gross Income $11,000 $11,000 $22,000

Less personal exemptions 5,100 5,100 10,200

Less standard deduction 5,900 5,900 6,700

Equals taxable income 0 0 5,100

Tax (At 15 percent) 0 0 765

Less earned income tax credit 2,152 2,152 1,368

Tax Liability -2,152 -2,152 -603

Marriage Penalty $3,701

As a Percentage of Adjusted Gross Income 16.8

SOURCE: Congressional Budget Office.

a. Baseline for calculating marriage penalty assumes that both husband and wife file as head of household with one child.

receive a credit equal to 7.65 percent of earnings in the

phase-in range, whereas those with one child get a 34

percent credit, regardless of tax filing status. The vari-

ance of the EITC with the number of children in a tax

unit cannot impose a marriage penalty: combining chil-

dren into one unit does not lower either credit percent-

ages or the income ranges over which the credit in-

creases or decreases. Given that the credit does not

increase for tax units with additional children beyond

two means, however, some couples with more than two

children could incur a penalty. That penalty results

from the limitation of a couple to one credit, whereas

each member could receive a credit if they were not

married.

Families who have children and earnings below

$10,000 generally receive a marriage bonus because of

the phase-in of the EITC (see Figure 3). As incomes

rise above $10,000, two-earner couples with children

incur a marriage penalty as their total incomes lead to a

phaseout of their EITC. The penalty is greatest 15

percent of earnings for two-earner families with chil-

dren and total earnings between $20,000 and $30,000,

the range in which the EITC disappears for joint filers

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

Childless Couples and

Couples with One Child

Couples with Two Children

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Couples with One Earner

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Couples with One Spouse Contributing 75 Percent of Couple's Earnings

Childless Couples

Couples with One Child

Couples with Two Children

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Couples with Both Spouses Contributing Half of Couple's Earnings

Childless Couples

Couples with One Child

Couples with Two Children

22 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Figure 3.

Marriage Penalties and Bonuses Resulting from the Earned Income Tax Credit, 1996 Tax Law

SOURCE: Congressional Budget Office.

NOTE: Positive values indicate penalties; negative values indicate bonuses.

CHAPTER TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 23

but would still exist if the couple were not married and

earnings were split between parents. In 1996, the larg-

est penalty fell on couples with four or more children in

which each spouse earned $11,610. Filing jointly, such

a couple would receive a credit of $1,111. If the couple

were not married and each parent filed as a head of

household with two children, each would receive a

credit of $3,556, or a total of $7,112. The marriage

penalty attributable to the EITC would thus equal

$6,001, more than one-fourth of their earnings. Finally,

childless couples never get a tax bonus from the EITC,

but the credit may impose a penalty on those couples

with two earners whose combined income exceeds

$5,280.

The Phaseout of Personal Exemptions

High-income couples can receive marriage bonuses or

incur marriage penalties through a provision of the in-

come tax code that denies them some or all of the value

of their personal exemptions. In 1996, the phaseout

reduced the personal exemptions taxpayers could claim

by 2 percent for every $2,500 (or fraction of $2,500) by

which their adjusted gross income exceeds $117,950

for single filers, $147,450 for heads of household, or

$176,950 for joint filers. Because the threshold income

level for single people was two-thirds of the level for

married couples, unmarried couples with two high in-

comes could have combined incomes up to $58,950

higher than if they were married without being affected

by the phaseout.

That difference can result in single filers preserving

up to 46 percent of their personal exemptions, com-

pared with married couples. For a childless couple in

the 31 percent tax bracket, the difference would trans-

late into a marriage penalty of more than $700. If mar-

ried, the couple would lose 46 percent of their $5,000 in

personal exemptions $2,300. At a 31 percent tax rate,

the loss would increase their taxes by $713 31 percent

of $2,300. Couples with children could bear an even

larger penalty because, if they were not married, they

could file as heads of household and protect their ex-

emptions until their individual incomes reached

$147,450. The penalty borne by such couples could

exceed $2,900.

Conversely, couples who have one high-earner and

one non- or low-earner can receive marriage bonuses

because the phaseout of exemptions for couples starts

at a higher income level. In practice, the phaseout of

personal exemptions affects few couples: less than 3

percent of joint filers have incomes above the phaseout

threshold and not all of them incur marriage penalties.

The Limitation on Itemized Deductions

High-income couples can incur a marriage penalty be-

cause the tax code limits their itemized deductions more

than if they were not married. Regardless of filing sta-

tus, the limitation reduces the itemized deductions a

taxpayer can claim by 3 percent of the amount by which

adjusted gross income exceeds an indexed thresh-

old $117,950 in 1996. Because the threshold is the

same for both single and joint filing statuses, a couple

in which husband and wife each has AGI at the thresh-

old could claim their full itemized deductions if they

were not married but could deduct $3,539 (3 percent of

$117,950) less if they were married. The limitation,

however, cannot generate marriage bonuses. Again,

few couples are affected; less than 7 percent of joint

filers are subject to the limitation.

Other Fixed Dollar Limitations

Other features of the tax code that use fixed dollar val-

ues to limit reductions in taxable income or allowable

tax credits can generate marriage penalties or bonuses.

Limitations that do not differ by filing status, or that

are not high enough to offset for joint filers the effects

of combining a couple's income, impose higher tax lia-

bilities on married couples than if they were not mar-

ried. Such limitations include the dependent care credit,

the income limit for deductible individual retirement

accounts (IRAs), and the thresholds for taxation of So-

cial Security benefits, among others.

The dependent care credit, for example, declines

from 30 percent of allowed expenses for tax units with

incomes below $10,000 to 20 percent for those with in-

comes above $28,000, whereas allowed expenses can-

not exceed $2,400 for one child and $4,800 for two or

more children. A married couple with four children in

which the husband and wife each earn $24,000 could

receive a credit of $960, but both spouses could claim a

credit of $1,104 if they were not married. Most of the

decrease in the credit resulting from marriage derives

from limiting the total creditable child care expenses to

$4,800. Of the $1,248 penalty in the example, $144

results from the reduction in the percentage rate of the

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

Couples with One Spouse

Contributing 75 Percent of Earnings

Couples with One Earner

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Childless Couples

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

0 5 10 15 20 25 30 35 40 45 50

-20

-10

Couple's Earnings (In thousands of dollars)

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Couples with One Child

Marriage Penalty or Bonus (As a percentage of couple's earnings)

Couples with Two Children

Couples with Both Spouses

Contributing Half of Earnings

Couples with One Spouse

Contributing 75 Percent of Earnings

Couples with Both Spouses

Contributing Half of Earnings

Couples with One Earner

Couples with Both Spouses

Contributing Half of Earnings

Couples with One Spouse

Contributing 75 Percent of Earnings

Couples with One Earner

24 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Figure 4.

Marriage Penalties and Bonuses by Number of Children and Earnings Split, 1996 Tax Law

SOURCE: Congressional Budget Office simulations.

NOTE: Positive values indicate penalties; negative values indicate bonuses.

CHAPTER TWO SOURCES OF MARRIAGE PENALTIES AND BONUSES 25

credit and $1,104 comes from the limit on creditable

expenses.

Fixed-dollar limitations generate marriage bonuses

when the dollar limitation is higher for a couple than for

a single person and when one spouse adds little to the

measure that is subject to the limitation. For example,

a single earner can deduct up to $2,000 in contributions

to an IRA if he or she has earnings of at least $2,000

and if his or her adjusted gross income is not more than

$25,000, whereas a couple can do so if its income does

not exceed $40,000. A couple with both spouses earn-

ing $25,000 could take no IRA deduction, but each

member could deduct $2,000 if they were not married.

At the same time, a one-earner couple with a $40,000

income could take the full deduction, but neither spouse

could deduct any retirement contributions if he or she

were not married. Furthermore, the one-earner couple

could increase their deductible IRA contribution to

$4,000 because a spouse with no earnings can deduct

up to $2,000.

Which Cou

les Pa

the

Marria

e Tax?

Three factors have the greatest influence on whether a

couple bears a marriage penalty or receives a marriage

bonus: the couple's total income, the division of that

income between husband and wife, and the presence

and number of children (and other dependents) that de-

termine the filing status of unmarried individuals and

qualify taxpayers for the EITC and personal exemp-

tions.

The following analysis calculates marriage penal-

ties and bonuses as the differences in tax for a married

couple filing a joint return versus their filing two sin-

gle or, if applicable, head-of-household returns. All

income is assumed to be from earnings and each spouse

would claim the standard deduction or itemized deduc-

tions equal to 20 percent of earnings, whichever is

greater. That simplified calculation differs from the

more complete model used in Chapter 3 to estimate

marriage penalties and bonuses.

Couples with just one earner (and no unearned in-

come) never incur a marriage penalty and receive a bo-

nus at all but the lowest income levels. At low incomes,

that bonus results from the additional personal exemp-

tions and standard deduction brought to the marriage by

the nonworking spouse and, for couples with children,

larger earned income tax credits (see Figure 4). Upper-

income couples receive their bonuses primarily from the

tax-rate brackets that are wider for joint filers than for

single people.

The largest marriage bonuses, measured as a per-

centage of income, occur in two distinct cases. First,

two-earner couples with one child and very low incomes

split equally between spouses receive a larger EITC as

the credit phases in and thus receive a bonus of up to 13

percent of their income. Second, low-income single-

earner couples in which each spouse has one child, and

for whom combining children into one tax unit in-

creases the size of the EITC, receive bonuses of up to

11 percent of income (see Figure 4). The largest

bonuses in dollar terms

more than $5,600 go to

childless one-earner couples with incomes between

$180,000 and $190,000.

Marriage penalties generally increase as the divi-

sion of income between husband and wife becomes

more equal. At high incomes the penalties result pri-

marily from the different tax-rate brackets for each fil-

ing status, whereas the EITC and differences between

standard deductions for joint filers and single people or

heads of household dominate for couples with low in-

comes. Once again, the penalties, measured as a per-

centage of income, are greatest for low-income couples

who have several children and an equal division of in-

come between spouses; the loss of EITC on joint re-

turns can cost such families up to 18 percent of income.

In dollar terms, the penalty resulting from differences in

tax brackets, limitations on itemized deductions, and

the phaseout of personal exemptions combine to im-

pose the maximum penalty more than $21,500 on

couples whose income is equally divided between

spouses and whose taxable income exceeds $527,500.

Less than 1 percent of married couples have incomes

that high, however, and few of those are likely to have

income divided evenly enough between spouses to gen-

erate such large penalties.

Chapter Three

Magnitudes of Marriage

Penalties and Bonuses

early all married couples incur penalties or

bonuses that shift billions of dollars of income

taxes among taxpayers. Depending on how

marriage penalties are calculated, between 14 million

and 23 million married taxpayers incurred penalties

totaling between $8 billion and $40 billion in 1996. In

that year, between 24 million and 31 million other cou-

ples benefitted from between $32 billion and $45 bil-

lion in bonuses. Average penalties and bonuses in-

crease with income and more than half of the total value

of each falls on tax units with incomes above $50,000.

Only couples with no tax liability and those few with

incomes falling in specific ranges incur neither a pen-

alty nor a bonus. Less than 10 percent of all joint tax

filers are unaffected under any of the four measures of

bonuses and penalties described below.

Measurin

Marria

e Penalties

and Bonuses

The size and imposition of marriage penalties and bo-

nuses depend on the benchmark used for comparison

against actual tax liabilities. A marriage penalty equals

the amount by which a couple's tax liability under joint

filing exceeds the taxes they would pay if the income

tax code allowed them to file as individuals. If they

were allowed to file individually, however, it is not clear

whether the tax code would provide the same set of ex-

emptions, deductions, and credits currently available to

unmarried taxpayers. For example, without joint filing,

the code might not allow taxpayers to file as heads of

household and take advantage of the tax brackets avail-

able under that filing status. If joint filing was not gen-

erally required for married couples, the Congress might

elect to change dollar limits on various tax provisions,

thus raising or lowering the taxes owed by individual

filers. In sum, it is not clear exactly what provisions of

the tax code should provide the benchmark against

which to estimate marriage penalties and bonuses.

The aspects of the tax code used to calculate penal-

ties and bonuses determine both the size of those mea-

sures and which couples incur them. In general, mea-

sures that incorporate more aspects of the tax code gen-

erate larger penalties affecting more couples and

smaller bonuses for fewer couples. This analysis fo-

cuses on one basic measure of penalties and bonuses,

but it also compares their size and affected populations

under that measure with similar values calculated under

one broader and three more limited measures. The

choice of measure affects both the total amounts of

penalties and bonuses and which couples bear them.

Furthermore, the estimates of marriage penalties

and bonuses made in this analysis understate the full

costs of joint filing because they are derived from ob-

servations of couples' incomes under the existing tax

code. Thus, they do not account for behavioral changes

made in response to that code. If the high marginal tax

1. The estimates given in this chapter are based on data from the Internal

Revenue Service's Statistics of Income. See Appendix E for a descrip-

tion of those data.

28 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

rates facing second earners under joint tax filing induce

them to work less than they would under lower rates,

the estimated costs of requiring joint returns are too low

and the values assigned to bonuses too high. Overall,

joint filing requirements may reduce the labor supply of

married couples by about 1 percent. The forgone earn-

ings caused by the tax code impose costs not only on

the workers in the form of lower incomes but also on

society in the form of lost output. In addition, some

couples who appear to benefit from joint filing by re-

ceiving bonuses may actually be worse off: their tax

savings may result from their working less in response

to taxes rather than from the labor supply choices they

would make without taxes.

Finally, regardless of which aspects of the tax code

are used, estimates of penalties and bonuses must make

assumptions about the division of incomes, deductible

expenses and the assignment of dependents. One alter-

native would be a "divorce" model that would set the

following conditions of a divorce: spouses would re-

port their own incomes and itemize their own deduct-

ible expenses, and children would likely go with the

mother. A divorce model would be appropriate if the

relevant baseline consists of the taxes that a couple

would owe if they were to dissolve their marriage.

Alternatively, deductions, exemptions, and children

might be allocated so as to minimize each couple's joint

tax liability. A third possibility would rely on specific

rules for assignment. Estimates of penalties and bo-

nuses using either of those alternatives, rather than a

divorce model, would more likely reflect possible

changes that the Congress might make to alter the tax

treatment of married couples. Consequently, this study

relies on specific assumptions about the division of in-

comes, expenses, and dependents to quantify marriage

penalties and bonuses (see Appendix F for estimates of

penalties and bonuses under a divorce model).

The estimates of marriage penalties and bonuses

below are just that: estimates. Because they ignore

behavioral responses, they are not measures of how tax

revenues would change if married couples could file in-

dividual tax returns. Nor are they estimates of the full

economic costs of joint filing. They are simply the cal-

culated differences between the taxes married couples

pay under current law and the taxes they would pay if

they had to file individual tax returns given specific

assumptions about the division of incomes, itemizable

deductions, and dependent exemptions between

spouses.

This study's basic measure of penalties and bo-

nuses calculates the penalty or bonus a couple bears as

the difference between the taxes they would pay when

filing jointly under 1996 law and the taxes they would

pay if they filed singly given the following assump-

tions:

o A couple's income from investments is divided be-

tween spouses in proportion to their earnings.

Each spouse pays taxes on his or her earnings and

share of investment income. Couples with no earn-

ings are treated as one-earner couples.

o Spouses claim the couple's itemized deductions in

proportion to their individual incomes. A spouse

claims the appropriate standard deduction if it ex-

ceeds his or her share of the couple's itemized de-

ductions.

o All childless couples file two single returns.

o In couples with one child, the higher-earning

spouse claims the child as a dependent and files as

a head of household. The other spouse files a sin-

gle return.

o If a couple has two or more children, the lower-

earning spouse claims one child as a dependent and

files as a head of household. The other spouse

claims all other children and also files as head of

household.

o Both spouses can claim the earned income tax

credit if eligible.

To show how the measure used to assess penalties

and bonuses affects their magnitudes, the analysis also

considers one broader and three more restrictive alter-

natives (see Box 6 for a summary of all five measures).

The broader one is identical to the basic measure dis-

cussed above except that the higher-earning spouse

claims all of the couple's itemized deductions, and the

lower-earning spouse claims the appropriate standard

2. See "Effects on the Supply of Labor" in Chapter 1 of this study for a

discussion of behavioral responses to increased marginal tax rates.

CHAPTER THREE MAGNITUDES OF MARRIAGE PENALTIES AND BONUSES 29

Box 6.

Alternative Measures of Marriage Penalties and Bonuses

A marriage penalty or bonus equals the difference be-

tween the tax liability of a couple filing jointly and their

liability if they could file as individuals. The calculation

depends on assumptions about the taxes they would pay

if they filed individually. Calculations differ depending

on the division between spouses for tax purposes of un-

earned income, itemized deductions, and children,

whether individuals can file as heads of household, and

whether individuals can claim amounts of earned in-

come tax credits (EITCs) different from what they

would receive on joint returns. The following measures

differ in their assumptions about how couples would file

tax returns if they were allowed to file separately.

Basic Measure

: Spouses divide unearned income and

itemized deductions in proportion to their earnings. The

first child is assigned to the spouse with higher earnings,

the second to the lower-earning spouse, and all others to

the higher earner. If eligible, both spouses can file as

head of household and claim the earned income tax

credit.

Broader Measure

: Same as basic measure except that

the spouse with higher earnings claims all itemized de-

ductions and the lower earner takes the appropriate

standard deduction.

Basic Measure Less Head of Household

: Same as

basic measure except that spouses must file single re-

turns and cannot file as heads of household.

Basic Measure Less EITC

: Same as basic measure

except that penalty or bonus calculation excludes any

changes in eligibility for or amounts of the earned in-

come tax credit.

Basic Measure Less Head of Household and EITC

Same as basic measure except that spouses must file

single returns and cannot file as heads of household and

the penalty or bonus calculation excludes any changes in

eligibility for or amounts of the earned income tax

credit.

deduction. In general, that broader measure reduces

the tax liability that the couple would incur if spouses

filed individually and therefore increases the calculated

size and incidence of marriage penalties (it has the re-

verse effect on bonuses).

The more restrictive measures result in smaller pen-

alties and larger bonuses. The first does not allow par-

ents to file as heads of household but requires that they

file single returns. That restriction increases the base-

line tax liabilities used to calculate penalties and bo-

nuses and thus reduces the size of the penalties and the

population they affect. It also increases the size of bo-

nuses and the number of couples they affect.

A second more restrictive measure allows parents

to file as heads of household but excludes any effects of

changes in amounts of earned income tax credit that

they would be entitled to claim if they were allowed to

file separate returns. Omitting EITC effects has the

same impact on estimated penalties and bonuses that

denying head-of-household filing status does: the base-

line tax liabilities against which penalties and bonuses

are assessed increase, fewer couples incur penalties,

and more couples receive bonuses.

The final measure excludes both head-of-household

filing status and the savings that spouses would receive

from separately claiming the EITC. Combining the two

limitations increases the number of couples receiving

bonuses with a commensurate reduction in the number

of couples incurring penalties.

Total Marria

e Penalties

and Bonuses

Under the basic measure of penalties and bonuses,

nearly 21 million couples incurred marriage penalties

averaging nearly $1,400 apiece in 1996, raising their

taxes by a total of about $29 billion. At the same time,

another 25 million couples received bonuses averaging

3. This definition is proposed by Daniel R. Feenberg and Harvey S.

Rosen, most recently in "Recent Developments in the Marriage Tax,"

National Tax Journal

, vol. 47, no. 1 (March 1995), pp. 91-101.

30 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

$1,300 apiece, saving them a total of about $33 billion

in taxes (see Table 4). An additional 3 million couples

incurred neither penalties nor bonuses.

Thus, about 51 percent of all married couples paid

lower taxes in 1996 because they filed joint returns, and

42 percent paid more. Only 6 percent were unaffected.

Under the basic measure of the marriage penalty, cou-

ples paid a total of about $4 billion less in taxes than

they would have if they were required to file as individ-

uals.

The broader measure of marriage penalties and bo-

nuses results in larger estimates of the number and size

of penalties and smaller values for the number and size

of bonuses. Compared with the effects of the basic

measure, the broader concept would mean that an addi-

tional 2 million couples would incur penalties and the

Table 4.

Total Value of and Tax Returns with Marriage Penalties and Bonuses Under

Alternative Measures, Projected 1996

Basic

Broader Basic Basic Basic Measure

Measure Measure Measure Measure Less Head of

of Penalty of Penalty Less Head Less Household

and Bonus and Bonus of Household EITC and EITC

Joint Returns (Millions)

With Penalties 23.0 20.9 18.5 17.8 13.9

Unaffected 2.4 3.1 3.0 4.1 3.9

With Bonuses 23.9

25.3 27.7 27.3 31.4

Total 49.3 49.3 49.3 49.3 49.3

Joint Returns (Percent)

With Penalties4742383628

Unaffected56688

With Bonuses 49

51 56 55 64

Total 100 100 100 100 100

Total Value for Joint Returns (Billions of dollars

)

Penalties 40.2 28.8 18.1 17.0 8.5

Bonuses 32.2 32.9 42.2 33.2 44.7

Net Effect on

Tax Liability 7.9 -4.1 -24.1 -16.1 -36.2

Average Penalty or Bonus for Joint Returns (Dollars)

Penalties 1,750 1,380 980 960 610

Bonuses 1,350 1,300 1,520 1,210 1,420

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: EITC = earned income tax credit.

a. Positive values indicate net penalties; negative values indicate net bonuses.

CHAPTER THREE MAGNITUDES OF MARRIAGE PENALTIES AND BONUSES 31

average penalty would rise to nearly $1,750, roughly a

25 percent increase. As a result, total penalties would

be 40 percent higher about $40 billion. More than 1

million fewer couples would receive bonuses, and the

total bonus would fall by less than $1 billion. Under

this measure, couples as a group would pay an addi-

tional $8 billion in taxes because they had filed joint

tax returns.

More restrictive measures would have opposite

effects. Denying the head of household filing status

would shift more than 2 million couples from penalties

to bonuses, reduce the total penalty by nearly $11 bil-

lion, and increase the total bonus by roughly $9 billion.

Under that measure, joint filing cuts the taxes paid by

couples by about $24 billion. Alternatively, restricting

the measure to exclude tax changes resulting from cou-

ples claiming additional EITC would reduce the number

of penalized couples by 3 million and shift 2 million

couples into bonus status. Compared with the basic

measure, it would reduce total penalties by more than

40 percent, but would have little effect on the total

value of bonuses.

Combining both restrictions that is, excluding

both head-of-household filing status and EITC differ-

ences would remove an additional 4 million couples

from the penalty category and give bonuses to about

4 million more couples. That tighter measure would cut

the total penalty to less than one-third of that estimated

under the basic definition. Overall, under this measure,

joint filing would reduce the total tax liability of mar-

Table 5.

Couples Receiving Marriage Penalties and Bonuses by Adjusted Gross Income, Projected 1996

Less Than $20,000- $50,000- $100,000 All

$20,000 $50,000 $100,000 and Over Incomes

Tax Returns (Thousands)

With Penalties 1,100 8,100 9,000 2,700 20,900

Unaffected 2,300 300 500 0 3,100

With Bonuses 5,800

10,000 7,200 2,300 25,300

Total 9,200 18,400 16,700 5,000 49,300

Percentage of All Returns in Income Category

With Penalties 12 44 54 54 42

Unaffected 25 1 3 0 6

With Bonuses 63

55 43 46 51

Total 100 100 100 100 100

Percentage of All Returns in Penalty,

Bonus, or Unaffected Category

With Penalties 5 39 43 13 100

Unaffected 74 10 16 0 100

With Bonuses 23 40 28 9 100

All Returns19373410100

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: Marriage penalties and bonuses are calculated using the basic measure outlined in Box 6.

32 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

ried couples by about $36 billion below what they

would have owed if they had been required to file indi-

vidual returns.

The Distribution of Marria

Penalties and Bonuses

Marriage penalties and bonuses are not apportioned

evenly across the income distribution (see Table 5 on

page 31). Instead, lower-income couples are more

likely to receive bonuses, whereas higher-income cou-

ples are about equally likely to incur penalties or re-

ceive bonuses. In 1996, couples with an adjusted gross

income of less than $20,000 19 percent of all joint

filers made up 23 percent of bonus recipients but only

5 percent of couples incurring penalties. By contrast,

couples with an AGI between $50,000 and $100,000

34 percent of all couples represented 43 percent of

those bearing penalties and 28 percent of those receiv-

ing bonuses. Three-fourths of couples filing tax returns

and incurring neither bonuses nor penalties had an AGI

below $20,000. Those couples generally owed no

taxes, whether they filed jointly or singly.

In dollar terms, average marriage penalties and bo-

nuses rise with couples' incomes, but the reverse holds

when they are measured in relation to income (see

Table 6). In 1996, couples with bonuses and AGI be-

low $50,000 paid taxes averaging between $700 and

$900 less than what they would have paid if they had

filed singly, whereas those with bonuses and an AGI

above $100,000 saved an average of nearly $3,000 in

taxes by filing joint returns. Despite those large dollar

differences, the distribution of bonuses was progres-

sive. Affected couples with incomes below $20,000

received bonuses averaging 5 percent of their incomes

in 1996, double that for bonus recipients with incomes

between $20,000 and $100,000 and four times that for

those with incomes above $100,000.

Table 6.

Marriage Penalties and Bonuses by Adjusted Gross Income of Couple, Projected 1996

Less Than $20,000- $50,000- $100,000 All

$20,000 $50,000 $100,000 and Over Incomes

Total Penalty or Bonus for Affected Couples (Millions of dollars)

Penalties 900 9,600 11,100 7,200 28,800

Bonuses 3,900

8,700 13,500 6,800 32,900

Net Effect on

Tax Liability -3,000 900 -2,400 400 -4,100

Average Penalty or Bonus for Affected Couples (Dollars)

Penalties 770 1,190 1,240 2,640 1,380

Bonuses 680 870 1,880 2,970 1,300

Average Penalty or Bonus for Affected Couples (Percentage)

Penalties 7.6 3.2 1.7 1.4 2.0

Bonuses 5.0 2.6 2.8 1.2 2.3

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: Marriage penalties and bonuses are calculated using the basic measure outlined in Box 6.

a. Positive values indicate net marriage penalties; negative values indicate net bonuses.

CHAPTER THREE MAGNITUDES OF MARRIAGE PENALTIES AND BONUSES 33

Table 7.

Joint Tax Returns by Penalty or Bonus Status and Division of Earnings Between Spouses, Projected 1996

Both Spouses

Each Spouse Had Earnings

Earned and One Spouse

at Least Contributed Only One

One-Third of More Than Spouse

Couple's Two-Thirds of Had All

Earnings Couple's Earnings Earnings Couples

Joint Returns

(

Thousands)

With Penalties 10,800 10,100 0 20,900

Unaffected 600 100 2,300 3,100

With Bonuses 600

5,300 19,400 25,300

All Joint Returns 12,000 15,500 21,700 49,300

Percentage of All Joint Returns

With Penalties 22 20 0 42

Unaffected 1 b 5 6

With Bonuses 1

11 39 51

All Joint Returns 24 31 44 100

Percentage of Joint Returns in Income Shares Category

With Penalties 90 65 0 42

Unaffected 5 1 11 6

With Bonuses 5

34 89 51

All Joint Returns 100 100 100 100

Percentage of Joint Returns in Penalty or Bonus Category

With Penalties 52 48 0 100

Unaffected 21 5 75 100

With Bonuses 2 21 77 100

All Joint Returns 24 31 44 100

SOURCE: Congressional Budget Office projections based on data from the 1993 Statistics of Income.

NOTES: Marriage penalties and bonuses are calculated using the basic measure outlined in Box 6.

See Table D-2 for additional data supporting this table.

a. Couples with no earners and those with one earner incur similar marriage penalties and bonuses and are thus combined in the analysis. For

simplicity of exposition, the text refers to both types of couples as having one earner.

b. Less than 0.5 percent.

34 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Average penalties vary more widely in terms of

both dollars and percentage of income. Couples at the

bottom end of the income scale who incur penalties

paid an average of nearly $800 in additional taxes,

compared with marriage penalties averaging more than

$2,600 for couples with an AGI above $100,000. For

those in the bottom income category, however, that av-

erage penalty represented nearly 8 percent of income,

compared with less than 2 percent for couples incurring

penalties in the top category. Those couples in the

middle-income ranges who are affected namely, those

with an AGI between $20,000 and $100,000 incurred

penalties averaging about $1,200, not quite 3 percent of

their income.

In combination, marriage penalties and bonuses

raised the total taxes paid by couples with an AGI be-

tween $20,000 and $50,000 and by those with an AGI

above $100,000, and lowered taxes for couples in other

income categories. Filing joint returns increased the tax

bill of the first group by about $900 million in 1996

the difference between $9.6 billion in marriage penal-

ties and $8.7 billion in bonuses. The one-fifth of all

couples who have incomes below $20,000 benefited as

a group; about $3.9 billion in bonuses offset roughly

$900 million in penalties, yielding net tax savings of

about $3 billion because those couples file joint returns.

Couples with an AGI between $50,000 and $100,000

one-third of all joint filers paid nearly $2.4 billion less

in total taxes because they filed jointly, whereas those

in the top income group paid an additional $400 million

because they could not file as individuals. Nonetheless,

about one-half of all couples received bonuses and two-

fifths incurred penalties. The net effect within an in-

Table 8.

Marriage Penalties and Bonuses by Penalty or Bonus Status and Division of Earnings

Between Spouses, Projected 1996

Both Spouses

Each Spouse Had Earnings

Earned and One Spouse

at Least Contributed Only One

One-Third of More Than Spouse

Couple's Two-Thirds of Had All

Earnings Couple's Earnings Earnings Couples

Total Penalty or Bonus for Affected Couples (Millions of dollars)

Penalties 15,200 13,600 0 28,800

Bonuses 100

4,400 28,500 32,900

Net Effect on

Tax Liability 15,100 9,200 -28,500 -4,100

Average Penalty or Bonus for Affected Couples (Dollars)

Penalties 1,400 1,350 n.a. 1,380

Bonuses 90 830 1,470 1,300

SOURCE: Congressional Budget Office projections based on data from the 1993 Statistics of Income.

NOTES: Marriage penalties and bonuses are calculated using the basic measure outlined in Box 6.

See Table D-2 for additional data supporting this table.

n.a. = not applicable.

a. Couples with no earners and those with one earner incur similar marriage penalties and bonuses and are thus combined in the analysis. For

simplicity of exposition, the text refers to both types of couples as having one earner.

b. Positive values indicate net marriage penalties; negative values indicate net bonuses.

CHAPTER THREE MAGNITUDES OF MARRIAGE PENALTIES AND BONUSES 35

come category results from a significant redistribution

of tax liabilities from those individuals receiving bo-

nuses to those paying penalties.

Couples whose incomes are more equally divided

between husband and wife are more likely to incur mar-

riage penalties and less likely to receive bonuses than

couples in which one spouse's income is significantly

greater (see Table 7 on page 33). Couples with just one

earner made up 44 percent of all couples in 1996 but

represented 77 percent of all returns receiving bonuses,

lowering their income tax liabilities by more than $28

billion compared with what they would have owed if

they had filed individual returns (see Table 8). Virtu-

ally no one-earner couples paid marriage penalties. By

contrast, couples in which both spouses earned at least

one-third of their combined income one-quarter of all

couples were disproportionately likely to incur penal-

ties. In 1996 they filed 52 percent of all penalized re-

turns but only 2 percent of returns with bonuses. The

90 percent of such couples who incurred penalties paid

a total of $15 billion in additional taxes because they

filed jointly, an average penalty of about $1,400.

Two-earner couples whose income is less equally

divided between spouses were less likely to incur penal-

ties and more likely to receive bonuses than those with

a more equal division. Nearly two-thirds of two-earner

couples in which one spouse earned less than one-third

of total income paid penalties totaling over $13 billion,

Table 9.

Effective Individual Income Tax Rates of Married Couples Filing Individual or Joint Returns,

by Adjusted Gross Income and Division of Earnings Between Spouses, Projected 1996 (In percent)

Both Spouses

Each Spouse Had Earnings

Earned and One Spouse

at Least Contributed Only One

One-Third of More Than Spouse Nonelderly

Adjusted Gross Couple's Two-Thirds of Had Single

Income Earnings Couple's Earnings Earnings Filers

Effective Tax Rate when Couples File Joint Tax Returns

Less Than $20,000 -9.9 -12.9 -8.2 2.9

$20,000 - $50,000 8.5 7.8 7.3 10.7

$50,000 - $100,000 12.5 12.0 12.2 16.0

More Than $100,000 19.3 22.9 24.4 25.8

All Couples 13.2 15.2 15.0 11.3

Effective Tax Rate if Couples Were Required to File Individual Tax Returns

Less Than $20,000 -19.7 -13.1 -1.6 2.9

$20,000 - $50,000 5.4 5.9 10.1 10.7

$50,000 - $100,000 11.0 11.4 15.9 16.0

More Than $100,000 17.4 22.3 25.7 25.8

All Couples 11.2 14.3 17.6 11.3

SOURCE: Congressional Budget Office projections based on data from the 1993 Statistics of Income.

NOTE: Marriage penalties and bonuses are calculated using the basic measure outlined in Box 6.

a. Couples with no earners and those with one earner incur similar marriage penalties and bonuses and are thus combined in the analysis. For

simplicity of exposition, the text refers to both types of couples as having one earner.

36 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

and the other third received total bonuses of roughly $4

billion. The net effect raised taxes for all such couples

by a total of about $9 billion in 1996.

Effective Tax Rates of

Married Cou

les

Among all couples and within all income categories,

filing joint tax returns generally equalizes the effective

tax rates borne by one- and two-earner married couples,

thus roughly satisfying the goal of having couples with

equal incomes pay comparable taxes. The differences

in effective tax rates under joint and individual filing,

however, indicate that the goal of marriage neutrality is

not met. Couples who have incomes between $20,000

and $50,000, for example, paid taxes averaging about 8

percent of their incomes in 1996, regardless of how

much of their income was earned by each spouse (see

Table 9 on page 35). If couples had been required to

file individual returns, the average effective tax rate of

one-earner couples would have been 10 percent, double

the 5 percent rate paid by two-earner couples. A simi-

lar situation occurred for couples whose incomes were

between $50,000 and $100,000.

Joint filing appears to have had a less equalizing

effect on tax rates for couples whose incomes were

above $100,000 and for whom effective 1996 tax rates

ranged from 19 percent to 24 percent compared with a

range of average rates from 17 percent to 26 percent

under individual filing. The lower effective tax rate

results because two-earner couples in which both hus-

band and wife earn at least one-third of total income

had, on average, lower incomes than other high-income

couples.

A similar but reversed situation occurred for cou-

ples who had incomes below $20,000. Low-income,

two-earner couples had, on average, higher incomes

than their one-earner counterparts. Those with more

equal splits of earnings between husband and wife had

higher incomes than those for which the split was less

equal. That income inequality among types of couples

led to a significant divergence of effective tax rates un-

der joint filing (-8 percent to -13 percent), although that

spread was narrower than that which would have

occurred under individual filing (-2 percent to -20 per-

cent).

At all income levels, married couples incurred

markedly lower effective 1996 tax rates than single tax-

payers, largely because of the separate tax schedules for

the two groups and because couples generally can claim

more personal exemptions than single filers. Tax units

whose AGI was above $20,000 faced average effective

tax rates between 1 percentage point and 6 percentage

points lower if they were couples than if they were sin-

gle taxpayers. The difference was much larger 10

percentage points or more for tax units whose AGI

was under $20,000, primarily because the EITC was

more likely to go to couples than to single filers. The

generally lower incomes of single filers mean that as a

group, they pay a smaller share of their incomes in

taxes than married couples.

4. Negative tax rates imply that taxpayers receive a payment rather than

paying a tax.

5. Appendix G discusses the distribution of different types of families

across income categories. Working-age married couples are more

likely to be in higher-income categories than are families headed by

working-age single people or the elderly.

1969 1973 1977 1981 1985 1989 1993

100

Only Husband Worked

Both Husband and Wife Worked

Only Wife Worked

Neither Husband nor Wife Worked

Cumulative Percentage of Married Couples

Chapter Four

Labor Market and Demographic

Changes Affecting Marriage

Penalties and Bonuses

n the last two decades, the share of married cou-

ples who have two earners has risen sharply as

the difference between the earnings of husbands

and wives has decreased, particularly for couples with

higher incomes. Those labor market and demographic

changes, which have affected both childless couples and

those with children, have helped to increase the fraction

of couples incurring marriage penalties and drop the

proportion receiving marriage bonuses.

Between 1969 and 1995, the share of all married

couples in which both spouses worked increased from

46 percent to 60 percent, whereas the proportion with

just one worker fell from 46 percent to 25 percent (see

Figure 5). Over the same period, the fraction of cou-

ples with no workers nearly doubled from 8 percent to

15 percent, reflecting the growth of the elderly popula-

tion and the falling rates of labor force participation

among older men. Thus, although one-half of working

married couples had just one earner in 1969, that share

had dropped to less than one-third by 1995.

The increasing likelihood of married couples hav-

ing two earners was offset during the period by the de-

cline in married couples as a share of all families and

individuals (see Figure 6). For simplicity of exposition,

the following discussion refers to this group as "fami-

lies," including individuals not living with relatives as

families with a single member. In 1969, married cou-

ples made up two-thirds of all families; by 1995, that

share had dropped to one-half. Most of the shift re-

ure 5.

Work Patterns of Married Couples, 1969-1995

SOURCE: Congressional Budget Office tabulations of data from

the Current Population Survey. Data limitations required

interpolation of values for five years: 1970, 1972, 1974,

1975, and 1977.

1. The tabulations presented in this chapter are generally derived from

data from the Current Population Survey by the Bureau of the Census.

See Appendix E for a description of those data.

1969 1973 1977 1981 1985 1989 1993

100

Percentage of All Families and Unrelated Individuals

Unrelated Individuals

Families Headed by Females

Families Headed by Males

Married Couples

38 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

ure 6.

Distribution of Families and Unrelated

Individuals b

pe, 1969-1995

SOURCE: Congressional Budget Office tabulations based on data

in Bureau of the Census,

Household and Family Char-

acteristics: March 1993,

Current Population Reports

Series P20, No. 477 (June 1994), and unpublished data

from the Bureau of the Census.

sulted from sharp increases in the number of unrelated

individuals that is, one-person families whose pro-

portion rose from just over one-fifth in 1969 to more

than one-third in 1995. The fraction of all families

headed by unmarried women living with relatives,

mostly mothers and their children, grew less rapidly

over the 1969-1995 period: from 9 percent to 12 per-

cent of all units. The decline in the relative number of

married-couple families indicates that a smaller fraction

of families file joint tax returns. As a result, fewer fam-

ilies may incur marriage penalties or bonuses today

than was the case 25 years ago, even though the chang-

ing characteristics of married couples may have sub-

jected more couples to penalties.

Counts of actual tax returns filed in 1969 and 1993

provide somewhat different information but generally

confirm those changes. In 1969, 60 percent of returns

were filed by married couples, 4 percent by heads of

household, and 36 percent by single people. By 1993,

married returns had declined to 44 percent, heads of

household had tripled to 13 percent, and single returns

had risen to 42 percent. Those values differ from the

overall demographic data for two reasons: first, low-

income people need not file returns and generally do so

only to get refunds or the EITC; and second, single de-

pendents file their own returns if they have sufficient

income. Heads of household with low incomes may be

more likely to file returns than other low-income people

because they are likely to qualify for the EITC (which

did not exist in 1969).

Workin

-A

e Married

Cou

les

Since 1969, working-age married couples have become

increasingly likely to have two earners and to have a

more equal distribution of earnings between husband

and wife (see Table 10). That analysis defines working

-age couples with earnings as married couples in which

both the husband and wife are between 25 and 64 years

old; in which neither the husband nor the wife reports

negative earnings from self-employment; and in which

either the husband or the wife or both reports posi-

tive earnings.

Limiting the age range for couples restricts obser-

vations to the population most likely to be in the labor

force, thereby avoiding younger couples who may not

yet have completed their education and older people

who may have only loose ties to the labor force. Omit-

ting couples with negative earnings eliminates the con-

fusion that such uncommon units might create, and re-

quiring that couples have some earnings focuses the

analysis on couples in the labor force. For 1995, con-

straining the base population in those ways omits just

over one-fourth of all married couples. (See Appendix

H for a more complete discussion of limitations im-

posed on the observed population.)

The fraction of working-age married couples in

which both spouses held jobs rose from just under one-

half in 1969 to nearly three-fourths in 1995. Over the

same period, the fraction of couples in which both hus-

band and wife earned at least one-third of the couple's

total earnings doubled from 17 percent to 34 percent.

The near halving of the proportion of couples with just

one worker reduced the likelihood that couples would

CHAPTER FOUR LABOR MARKET AND DEMOGRAPHIC CHANGES AFFECTING MARIAGE PENALTIES AND BONUSES 39

Table 10.

Distribution of Workin

-A

e Married Couples, b

Division of Earnin

s Between Spouses in

1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnin

s and One Spouse

Earned at Least Contributed More

One-Third of Than Two-Thirds of Onl

One Spouse

Year Couple's Earnin

s Couple's Earnin

sHad Earnin

s All Couples

Working-Age Married Couples (Thousands)

1969 5,600 10,100 16,700 32,400

1979 7,500 12,700 13,700 34,000

1989 11,800 15,100 11,300 38,200

1995 13,600 15,200 11,300 40,100

Percentage of All Working-Age Married Couples

1969 17.2 31.2 51.7 100.0

1979 22.2 37.4 40.4 100.0

1989 30.9 39.5 29.5 100.0

1995 33.9 37.9 28.2 100.0

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Survey for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

receive marriage bonuses, and the greater equality of

earnings between spouses increased both the share of

couples bearing marriage penalties and the size of those

penalties. The changes in the earnings patterns of mar-

ried couples varied markedly, depending on a couple's

total earnings, age, and presence and number of chil-

dren.

Chan

es in Earnin

s Amon

Income Cate

ories

Between 1969 and 1995, the percentage of working-age

married couples in which both husband and wife

worked increased at all earnings levels, but rose more

for couples with high earnings than for those with low

earnings (see Table 11). In 1995, 31 percent of couples

with earnings below $20,000 in 1995 dollars had

two workers, compared with 29 percent in 1969. By

contrast, the share of couples with two earners whose

combined earnings were between $50,000 and

$100,000 increased over the same period from 63 per-

cent to 87 percent. Nearly 80 percent of couples in the

highest earnings group

more than $100,000 had two

earners in 1995, almost double the 43 percent of 1969.

By the end of the period, couples in the highest income

category were more likely to have two earners than

those whose incomes were below $50,000 and nearly as

likely as those whose incomes were between $50,000

and $100,000.

In part, the shift results from two earners generat-

ing more income than one, so that a couple in which

both husband and wife work is less likely to have earn-

ings low enough to fall into the bottom income cate-

gory. Having two earners increases the likelihood that

a couple's income puts them in one of the higher earn-

ings categories. Regardless of the reason for the shift,

the rise in two-earner couples over the last quarter cen-

tury was concentrated in the higher income categories,

where marriage penalties are generally larger and more

prevalent. The shift in working patterns of married

couples has thus intensified the problem of marriage

penalties.

40 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table 11.

Percenta

e of Workin

-A

e Married Couples with Two Earners, b

Total Income in 1969, 1979, 1989, and 1995

Less Than $20,000- $50,000- Than All

Year $20,000 $50,000 $100,000 $100,000 Incomes

1969 28.5 46.0 62.7 43.0 48.3

1979 29.5 56.3 73.4 56.7 59.6

1989 38.0 66.6 82.7 78.7 70.5

1995 30.9 69.7 87.0 78.6 71.8

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Survey for 1970, 1980, 1990, and 1996.

NOTES: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

Incomes are measured in 1995 dollars.

Husbands and wives now earn more nearly equal

incomes than they did 25 years ago. The change oc-

curred within all income groups, but was more pro-

nounced among those with the highest earnings (see

Table 12). Among working-age couples with total

earnings between $20,000 and $50,000 (in 1995 dol-

lars), the proportion of couples in which both spouses

earned at least one-third of the couple's total earnings

doubled from 15 percent in 1969 to 32 percent in 1995.

Couples whose total earnings were above $100,000

showed even greater movement toward equal earnings

between spouses: over the 1969-1995 period, the frac-

tion of those couples in which each spouse earned at

least one-third of the couple's earnings tripled from

10 percent to 33 percent. Even so, in 1995, an equal

division of earnings between spouses was most likely

for couples in the next lower income category (between

$50,000 and $100,000).

Both the increased likelihood of having two earners

and the greater equality of husbands' and wives' earn-

ings increased the chances that working couples would

incur a marriage penalty at income tax time. Those

shifts were more likely to occur for couples with high

earnings, which aggravated the situation because

higher-income families generally bear greater marriage

penalties in dollar terms, although not as a percentage

of income (see Table 6 in Chapter 3).

Chan

es in Earnin

s Amon

e Cate

ories

Both the portion of working-age couples with two earn-

ers and the equality of earnings between husband and

wife increased for all age groups between 1969 and

1995, but the increases were greater for couples in

which both spouses were under age 55 (see Tables 13

and 14). In 1969, roughly one-half of working couples

in each age category defined in terms of the age of the

older spouse had two earners, and couples in the 55 to

64 age group were only slightly less likely than younger

couples to have two workers. By 1995, three-quarters

of couples in each of the younger age groups had two

earners compared with less than 60 percent of couples

in the oldest age category. The trend of wives entering

the labor force appears to have been offset in the oldest

age group in which the older spouse was between 55

and 64 years old by the decreasing participation of

men over age 55.

Between 1969 and 1995 husbands and wives in

every age group moved toward greater equality of earn-

ings, and again the trend was stronger for younger cou-

ples. The share of couples in which both spouses

earned at least one-third of the couple's income more

CHAPTER FOUR LABOR MARKET AND DEMOGRAPHIC CHANGES AFFECTING MARIAGE PENALTIES AND BONUSES 41

Table 12.

Percenta

e Distribution of Workin

-A

e Married Couples, b

Division of Earnin

s Between

Spouses and Total Income in 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnin

s and One Spouse

Earned at Least Contributed More

One-Third of Than Two-Thirds of Onl

One Spouse

Year Couple's Earnin

s Couple's Earnin

s Had Earnin

All Income Cate

ories

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

Couple's Income Less Than $20,000

1969 7.1 21.4 71.5

1979 9.7 19.9 70.5

1989 12.3 25.7 62.0

1995 11.6 19.3 69.1

Couple's Income $20,000 to $50,000

1969 14.9 31.1 54.0

1979 20.3 36.0 43.7

1989 27.3 39.4 33.4

1995 32.1 37.6 30.3

Couple's Income $50,000 to $100,000

1969 27.4 35.4 37.3

1979 29.3 44.1 26.6

1989 40.2 42.5 17.3

1995 44.4 42.7 13.0

Couple's Income More Than $100,000

1969 9.9 33.1 57.0

1979 16.0 40.6 43.3

1989 32.5 46.3 21.3

1995 32.5 46.1 21.4

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTES: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative

earnings, and at least one spouse has positive earnings.

Incomes are measured in 1995 dollars.

42 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table 13.

Percenta

e of Workin

-A

e Married Couples with Two Earners, b

e of Older Spouse

in 1969, 1979, 1989, and 1995

e of Older Spouse All A

Year 25-34 35-44 45-54 55-64 Groups

1969 46.5 49.5 51.2 44.8 48.3

1979 67.7 63.5 59.8 44.8 59.6

1989 76.0 74.2 71.8 54.0 70.5

1995 76.7 74.1 74.2 57.9 71.8

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

than doubled for those under age 45, from 15 percent in

1969 to more than 35 percent in 1995. By contrast, the

oldest age group experienced little change in the equal-

ity of earnings between spouses, and the comparable

value rose only 7 percentage points from 19 percent to

26 percent. The greater equality of earnings for youn-

ger couples means that they are significantly more

likely to incur marriage penalties than their older coun-

terparts, all else being equal. Furthermore, if the ten-

dency toward greater earnings equality among the

young signals a long-term trend that will extend over

time into older age groups, the pattern may foreshadow

further increases in the number of couples who will be

subject to marriage penalties in the future.

Chan

es in Earnin

s b

Number of Children

Working-age couples were more likely to have two

earners in 1995 than in 1969, regardless of how many

children they had. By 1995, more than two-thirds of

couples in each category with none, one, and two or

more children had two earners. The increase in that

likelihood over the period was greater for those with

two or more children than for those with none or one

(see Table 15). Between 1969 and 1995, the fraction

of couples with two or more children and both parents

working rose by more than one-half, from 43 percent to

69 percent. By contrast, comparable values for child-

less couples grew about one-third, from 53 percent to

72 percent, largely because childless couples are gener-

ally older than couples with children and, as discussed

above, older couples are less likely than their younger

counterparts to have two earners.

Equality of earnings between spouses also in-

creased during the 1969-1995 period for couples, re-

gardless of the number of children, and couples with

children had larger increases than those with none.

Even so, at the end of the period, couples with two or

more children had less parity of earnings than other

couples (see Table 16). In 1969, one-fourth of child-

less couples had both spouses earning at least one-third

of their total income, compared with one-fifth of cou-

ples with one child and one-ninth of couples with two

or more children. By 1995, more than 36 percent of

couples with no children or one child showed that de-

gree of earnings equality

about 7 percentage points

more than couples with two or more children. Couples

with children are more likely than their childless coun-

terparts to incur marriage penalties, especially at lower

earnings levels, and the increased equality of earnings

between spouses served only to aggravate the situation

during the 1969-1995 period.

One rationale for the tax code's equal treatment of

one- and two-earner couples could be to provide an in-

direct subsidy for families in which one parent stays

home to care for children. Such a rationale might be

CHAPTER FOUR LABOR MARKET AND DEMOGRAPHIC CHANGES AFFECTING MARIAGE PENALTIES AND BONUSES 43

Table 14.

Percenta

e Distribution of Workin

-A

e Married Couples, b

Division of Earnin

s Between

Spouses and A

e of Older Spouse in 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnin

s and One Spouse

Earned at Least Contributed More

One-Third of Than Two-Thirds of Onl

One Spouse

Year Couple's Earnin

s Couple's Earnin

s Had Earnin

All A

e Groups

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

Older Spouse A

e 25-34

1969 15.2 31.4 53.5

1979 27.4 40.3 32.3

1989 36.2 39.8 24.0

1995 39.9 36.7 23.3

Older Spouse A

e 35-44

1969 14.7 34.8 50.5

1979 21.9 41.6 36.5

1989 32.2 42.0 25.8

1995 34.8 39.3 25.9

Older Spouse A

e 45-54

1969 19.5 31.7 48.8

1979 21.0 38.7 40.2

1989 31.0 40.8 28.2

1995 33.8 40.4 25.8

Older Spouse A

e 55-64

1969 19.3 25.4 55.2

1979 17.8 27.0 55.2

1989 21.4 32.6 46.0

1995 25.8 32.1 42.1

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

44 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

valid if families with children typically have only one

earner. That situation was common in the past but defi-

nitely does not prevail today. In 1969, nearly one-half

of all families with children were headed by couples

with just one earner; by 1995, that share had fallen

below one-fifth, whereas the fraction headed by two-

earner couples had risen above one-half (see Table 17).

Over the same period, the fraction of families with chil-

dren headed by a single parent more than doubled, from

13 percent to 30 percent. In view of such major demo-

graphic changes, equal tax treatment of one- and two-

earner couples cannot be justified on the basis of aiding

children. If the tax system has as one objective the fa-

vorable tax treatment of families with children, the tax

code can better offer such benefits directly through

higher exemptions or credits for children. Preferential

tax treatment of single-parent families, however, as

provided by the head-of-household filing status, might

offer reasonably well-targeted assistance to families

with children. Even so, a subsidy offered directly to

families with children would almost certainly provide a

more effective form of assistance.

Table 15.

Percenta

e of Workin

-A

e Married Couples

with Two Earners, b

Number of Children in

1969, 1979, 1989, and 1995

Two

No One or More All

Year Children Child Children Couples

1969 53.0 53.3 43.2 48.3

1979 59.9 65.0 56.4 59.6

1989 70.4 75.9 67.2 70.5

1995 71.6 76.9 69.1 71.8

SOURCE: Congressional Budget Office tabulations of data from

the March Current Population Survey for 1970, 1980,

1990, and 1996.

NOTE: W orking-age married couples include only those in which

both spouses are between ages 25 and 64, neither spouse

has negative earnings, and at least one spouse has positive

earnings.

Overview

Two complementary trends emerged over the last quar-

ter century: there were more working-age couples with

two earners and greater parity of earnings between hus-

bands and wives. Those trends occurred in all earnings,

age, and child categories, but were concentrated more

among couples with higher earnings, in younger age

groups, and with children. The patterns should have

given more couples marriage penalties and fewer cou-

ples bonuses, simply because penalties occur more fre-

quently for two-earner couples with a more equal divi-

sion of earnings between husband and wife than for

those whose earnings are less equally divided. The

trends in earnings since 1969 have made marriage pen-

alties more likely for all couples, particularly for those

with high earnings, those with both spouses under age

55, and those with children. At the same time, the

sharp decline in all categories in the proportion of one-

earner couples has meant a corresponding drop in the

number of couples benefiting from marriage bonuses,

which go to virtually all couples with one earner and

occur more frequently for couples who have a less

equal division of earnings.

An estimate of the effect of demographic and

earnings changes since 1969 on marriage penalties and

bonuses can be obtained by superimposing the 1969

distribution of married couples and their earnings on

the projected 1996 distribution and comparing the size

and incidence of penalties and bonuses under 1996 tax

law (see Table 18; statistics in the following discussion

include only working-age couples with some earnings).

Thus, if married couples in 1996 had work and earnings

patterns similar to those of their counterparts in 1969,

the fraction of couples incurring marriage penalties

would be markedly lower 30 percent compared with

47 percent and the fraction receiving bonuses would

be much higher 64 percent rather than the actual 48

percent. Those estimates support the conclusion that

labor market and demographic changes over the last

quarter century have increased the likelihood that cou-

ples will incur marriage penalties.

CHAPTER FOUR LABOR MARKET AND DEMOGRAPHIC CHANGES AFFECTING MARIAGE PENALTIES AND BONUSES 45

Table 16.

Percenta

e Distribution of Workin

-A

e Married Couples, b

Division of Earnin

s Between

Spouses and Number of Children in 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnin

s and One Spouse

Earned at Least Contributed More

One-Third of Than Two-Thirds of Onl

One Spouse

Year Couple's Earnin

s Couple's Earnin

s Had Earnin

All Couples

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

No Children

1969 24.4 28.6 47.0

1979 27.0 32.9 40.1

1989 34.3 36.2 29.6

1995 36.3 35.3 28.4

One Child

1969 19.3 34.1 46.7

1979 24.6 40.4 35.0

1989 34.9 41.1 24.1

1995 37.6 39.4 23.1

Two or More Children

1969 11.3 31.9 56.8

1979 16.3 40.1 43.6

1989 25.0 42.2 32.8

1995 29.2 39.9 30.9

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

Additional evidence of the effects of changing earn-

ings and demographics comes from the work of econo-

mists James Alm and Leslie Whittington, who found a

general rise since 1970 in the percentage of couples

incurring a marriage tax. Using both different as-

sumptions and data from those of the analysis dis-

cussed in Chapter 3, Alm and Whittington concluded

that the share of couples incurring penalties rose from

less than 40 percent in 1970 to nearly 60 percent in

1994, whereas the fraction receiving bonuses fell from

50 percent to less than 30 percent. Among couples in-

curring penalties, the average penalty rose through the

1970s from about $600 to more than $1,800 (in 1994

dollars) before falling to $1,000 in 1990 and then rising

2. James Alm and Leslie A. Whittington, "The Rise and Fall and Rise ...

of the Marriage Tax,"

National Tax Journal

, vol. XLIX, no. 4

(December 1996), pp. 571-589.

46 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table 17.

Percenta

e Distribution of Families with Children,

pe and Number of Earners in Married

Couples in 1969, 1979, 1989, and 1995

One-Earner Two-Earner Sin

le-

Married Married Parent

Year Couples Couples Families

1969 44.6 41.1 12.9

1979 29.8 46.9 21.1

1989 21.0 51.4 25.1

1995 18.7 50.2 29.6

SOURCE: Congressional Budget Office tabulations of data from

the March Current Population Surveys for 1970, 1980,

1990, and 1996.

NOTE: Married couples with zero or negative earnings are excluded.

again to $1,200 in 1994. Bonuses followed a similar

pattern, rising from nearly $800 in 1971 to more than

$1,600 in 1980 and then falling to just over $1,000 by

1994.

The Alm and Whittington analysis further demon-

strates that changes in earnings and demographic char-

acteristics have led to larger marriage penalties. One

simulation held tax law constant and found that

changes in the earnings of couples have caused the av-

erage effect of marriage on income tax liabilities to

change from a bonus of between $600 and $800 in

1970 to a penalty of $400 in 1994 (all in 1994 dollars).

An alternative simulation holding the makeup of the

population fixed at 1992 levels demonstrated that al-

though changes in tax law have caused average penal-

ties to vary widely over the past two decades between

more than $600 in the early 1970s to just over $200 in

the late 1980s

those changes alone have not increased

the likelihood that couples will incur penalties. Al-

though the Alm and Whittington analysis relies on dif-

ferent assumptions and data than does this study, it pro-

vides support for the conclusion that changing earnings

patterns and demographics have made marriage penal-

ties more common.

Table 18.

Simulated Effects of Demo

raphic Chan

e on

Marria

e Penalties and Bonuses, 1969 and 1996

Simulated Estimated

1969 1996

Percenta

e of Returns

with One Earner 57 38

Percenta

e of Earnin

Goin

to One-Earner

Couples 52 30

Percenta

e of

Returns with:

Penalt

30 47

Bonus 64 48

Neither penalt

nor bonus 6 5

SOURCE: Congressional Budget Office simulations based on data

from the 1993 Statistics of Income and the March Cur-

rent Population Surveys for 1970 and 1996.

NOTE: Estimates for 1969 were calculated by superimposing the

distribution of working-age couples by age and division of

earnings between husband and wife on measures of mar-

riage penalties and bonuses projected for 1996. The simu-

lations for 1969 thus implicitly assume that 1996 tax law

would apply to the 1969 distribution of married couples.

Chapter Five

Reducing Marriage Penalties

ecent growth in the size of marriage penalties

and the number of couples who are subject to

them has renewed interest in the effect of joint

filing on the taxes people pay. There are several ap-

proaches to the problem.

At one extreme lie minor changes within the current

tax system, including tax credits to offset the penalty,

changes in tax brackets and standard deduction levels,

or reenacting the two-earner deduction. For low-

income families, changing the structure of the earned

income tax credit could reduce perhaps to zero the

marriage penalties that it generates. Such proposals

would reduce the impact of the penalties but not elimi-

nate them.

At the other extreme is a reversion to the situation

that existed before 1948, in which all individuals had to

file single tax returns and pay taxes on their own in-

come. That change would totally eliminate the mar-

riage penalty but also get rid of bonuses and impose a

large redistribution of tax burdens among married cou-

ples. An intermediate approach allowing couples to

elect either joint, single, or, if appropriate, head-of-

household filing status would remove penalties but

leave bonuses in place, thus causing substantial revenue

losses and complicating the tax system. Finally, pro-

posals to replace the current income tax with either a

simpler tax on incomes or some form of consumption

tax would change the existing pattern of penalties and

bonuses in ways that will be apparent only after the

plans for reform are more specific.

Any change in the tax system for the purpose of

reducing or eliminating marriage penalties must con-

front a basic trade-off. Because marriage penalties

mean that many couples pay higher taxes than they

would if they were not required to file as a couple, low-

ering the penalties requires either that taxes on single

and head-of-household filers be raised or that taxes on

joint filers be cut. The former approach would almost

certainly bring complaints from those not filing jointly

that the tax laws again discriminated against them with

a new "singles penalty." The latter approach would

reduce tax revenues unless other changes make up for

the loss. Furthermore, the more it reduced penalties,

the greater would be the revenue loss. Changes that

reduce marriage penalties without incurring revenue

losses must redistribute the burden of taxes away from

couples now incurring penalties and toward other tax-

payers namely, either couples now receiving bonuses

or unmarried taxpayers. That basic trade-off limits

what the different options can accomplish.

A final issue concerns the relationship between

marriage penalties and bonuses. Changes in the tax

code that reduce penalties are generally blunt instru-

ments that affect all joint filers, whether they incur bo-

nuses or not. As a result, they tend not only to reduce

penalties but to increase existing bonuses and create

new ones. One important measure of the efficiency of

policies to lower penalties involves the extent to which

they also raise bonuses.

1. Jonathan Barry Forman offers a range of options in "What Can Be

Done About Marriage Penalties?"

Family Law Quarterly

, vol. 30, no.

1 (Spring 1996), pp. 1-22.

48 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box 7.

Setting Tax Brackets and Standard Deductions for Joint Filers to Twice Those for Single Filers

Simulations of the option setting tax brackets and stan-

dard deductions for joint filers equal to twice those for

single filers assumed that the 28 percent, 31 percent,

and 36 percent tax brackets for joint filers start at tax-

able incomes twice those for single filers. Doubling the

current starting point for the 39.6 percent bracket was

considered too extreme, however. Instead, the lower

end of that bracket for 1993 was taken to be 1.67 times

the appropriate starting point for single filers. Thus, the

simulation took the start of the 39.6 percent bracket for

single filers for 1993 to be $150,000, doubled that to

obtain a new 39.6 percent bracket starting income for

joint filers for 1993, and indexed both values forward to

1996. The resulting starting points for the top tax

bracket were thus assumed to be $158,250 for single

filers and $316,500 for joint filers. Those values com-

pare with $263,750 for both filing statuses under cur-

rent law. Furthermore, the simulations assumed bracket

starting points for heads of household halfway between

those for single and joint filers.

Simulations assumed standard deductions for the

three filing statuses of single filers, $4,000; heads of

household, $6,000; and joint filers, $8,000, compared

with $4,000, $5,900, and $6,700, respectively, under

current law.

The tax brackets used in the simulations were as fol-

lows:

Lowest Taxable Income for Tax Rates (In dollars)

Tax Rate Single Filers Head of Household Joint Filers

(Percent) Current Simulated Current Simulated Current Simulated

15 00 00 00

28 24,000 24,000 32,150 36,000 40,100 48,000

31 58,150 58,150 83,050 87,225 96,900 116,300

36 121,300 121,300 134,500 181,950 147,700 242,600

39.6 263,750 158,250 263,750 237,375 263,750 316,500

SOURCE: Congressional Budget Office.

If different options affect taxpayers in different

parts of the income distribution, combining options

could serve to spread the benefits of change more

equally among couples who are currently penalized.

For example, widening tax brackets and raising the

standard deduction for joint filers reduces penalties pri-

marily for couples who have incomes above $50,000,

whereas revising the EITC affects low-income couples

most. Combining the two options would reduce penal-

ties for the entire income distribution. Even so, couples

currently incurring marriage penalties would receive

varying degrees of relief. Furthermore, because options

may interact with each other, the effects of combining

two of them would not equal the sum of their individual

effects. Many combinations of options are possible.

The following analysis, however, considers only indi-

vidual options.

Widen Tax Brackets and Raise

the Standard Deduction for

Joint Filers

The cause of marriage penalties for many middle- and

high-income couples is that joint filers face tax brackets

that are less than twice as wide as those for single filers

and a standard deduction less than twice that for sin-

Under $20,000 $20,000-$50,000 $50,000-$100,000 Over $100,000

100

Adjusted Gross Income

Percentage of Marriage Penalty Removed

All

Families

Widen Brackets

($25 billion)

Expand EITC

($4 billion)

Two-Earner

Deduction

($9 billion)

Credit Against

Tax

($10 billion)

CHAPTER FIVE REDUCING MARRIAGE PENALTIES 49

gles. For example, in 1996 the 15 percent tax bracket

for joint filers applied to the first $40,100 of their tax-

able income 1.67 times the $24,000 first bracket for

single filers. Similarly, the 1996 standard deduction for

joint filers was $6,700 or 1.68 times that for single

filers. Because the ratios are less than two, married

couples in which husband and wife have roughly equal

incomes pay higher taxes than they would if they were

not married. Widening the tax brackets to twice the

width of single brackets and setting the joint standard

deduction to twice that for singles would eliminate that

source of marriage penalties. At the same time, how-

ever, it would create new marriage bonuses for many

couples and increase those for most couples now re-

ceiving them.

An alternative approach to reducing marriage pen-

alties would be to make standard deductions and tax

brackets for single filers equal to one-half of those cur-

rently available to joint filers. Parameters for heads of

household would be set midway between the two. Such

changes would increase the taxes of single and head-of-

household filers and thus increase federal tax revenues.

Because the incomes of such filers are generally lower

than those of joint filers, the revenue impact of the

change would be smaller.

ure 7.

Pro

ected 1996 Marria

e Penalt

Relief Under Alternative Proposals, Basic Measure of Marria

e Penalt

SOURCE: Congressional Budget Office projections based on data from the 1993 Statistics of Income.

NOTES: The basic measure of penalty and bonus assumes that income is divided between spouses based on the division of earnings; itemized

deductions are divided in proportion to income; childless couples file two single returns; in couples with one child, the higher-earning

spouse files as head of household and the lower-earning spouse files as single; in couples with two or more children, both spouses file as

head of household and the higher-earning spouse claims all but one child; and all filers claim earned income tax credit (EITC) if eligible.

The 1996 revenue cost of each proposal is shown in parentheses.

Widen Tax Brackets Two-Earner Deduction

Expand EITC Credit Against Tax

51 Percent to Couples

with Bonuses

7 Percent in Excess

of Penalty

43 Percent Offsettin

Penalties

80 Percent Offsettin

Penalties

10 Percent to Couples

with Bonuses

7 Percent in Excess of Penalty

3 Percent to

Unaffected Couples

4 Percent to

Unaffected Couples

3 Percent to Couples

with Bonuses

11 Percent in Excess of Penalty

82 Percent Offsettin

Penalties

88 Percent Offsettin

Penalties

11 Percent in Excess

of Penalty

1 Percent to

Couples with

Bonuses

50 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Setting tax brackets and standard deductions for

joint filers at twice those of single filers would reduce

federal tax revenues by an estimated $25 billion mea-

sured at 1996 income levels (see Box 7 on page 48 for

details of this option). The tax reduction would be

shared almost equally between couples now incurring

marriage penalties and those currently receiving bo-

nuses. On the whole, the changes would reduce mar-

riage penalties by 44 percent (see Figure 7 on page 49).

That reduction would go primarily to couples with in-

comes above $50,000, whose total penalty would fall

by two-thirds; lower-income taxpayers would have their

penalties reduced by just over one-tenth. At the same

time, because their top tax bracket would begin at a

ure 8.

Pro

ected 1996 Tar

et Efficienc

of Alternative Proposals to Relieve Marria

e Penalties,

Basic Measure of Marria

e Penalt

SOURCE: Congressional Budget Office projections based on data frzom the 1993 Statistics of Income.

NOTE: The basic measure of the penalty and bonus assumes that income is divided between spouses based on the division of earnings; itemized

deductions are divided in proportion to income; childless couples file two single returns; in couples with one child, the higher-earning spouse

files as head of household and the lower-earning spouse files as single; in couples with two or more children, both spouses file as head of

household and the higher-earning spouse claims all but one child; and all filers claim the earned income tax credit (EITC) if eligible.

CHAPTER FIVE REDUCING MARRIAGE PENALTIES 51

lower taxable income than under current law, high-

income taxpayers filing singly or as heads of household

would pay an additional $140 million in taxes.

Reducing marriage penalties by widening tax

brackets and increasing the standard deduction for joint

filers is highly inefficient because one-half of the reve-

nue cost to the government would go to couples already

receiving marriage bonuses (see Figure 8). Further-

more, more than five-sixths of the roughly $13 billion

in reduced taxes for current bonus recipients would go

to couples with incomes above $50,000, who incur less

than two-thirds of all current penalties.

Setting tax brackets and standard deductions for

joint filers at twice those for single filers would largely

return the tax bracket relationships to what they were

before 1969. As indicated in the historical discussion

above, the Congress changed the tax code in 1969 to

reduce the "singles penalty," the higher taxes that many

taxpayers incurred because they were not married.

Changes made that year drew a compromise between

marriage and singles penalties, creating the former in

the process of reducing the latter. Reversing that action

could rekindle the complaints of single filers that the

tax code discriminates against them. Even if changes

left their taxes unaffected, single filers might object that

lowering the taxes of married couples would leave sin-

gles worse off in relative terms, again paying higher

taxes because they are single than they would if they

were married.

Exem

t from Taxes Some

Income of Lower-Earnin

ouse

One cause of marriage penalties is that the first dollar

of taxable income of the lower-earning spouse is taxed

at the marginal tax rate of the higher-earning spouse.

Although the wider tax brackets for joint filers reduce

the taxes paid on the higher-earning spouse's income,

the higher initial tax rate faced by the lower-earning

spouse creates a disincentive for that spouse to work at

all. Furthermore, it results in a marriage penalty if that

spouse earns a large enough share of the couple's total

income. Exempting part of the lower-earning spouse's

income from taxation would alleviate this problem.

One approach to such an exemption would be to restore

the second-earner deduction.

Between 1982 and 1986, two-earner couples were

allowed to deduct from taxable income 10 percent (5

percent in 1982) of the earnings of the lower-earning

spouse, up to a maximum of $3,000. The Tax Reform

Act of 1986 eliminated the provision, in part because

other changes in the tax law had reduced the size of

marriage penalties, but also to offset the act’s revenue

losses. Restoring the two-earner deduction would re-

duce or eliminate the marriage penalty for many cou-

ples. For example, a couple with one spouse earning

$30,000 and the other earning $20,000 would be

shifted from incurring a $195 penalty to receiving a

$105 bonus (see Box 8). Overall, restoring the deduc-

tion would lower the total marriage penalty levied in

1996 by roughly one-third at an estimated cost of $9

billion in lost revenues.

The relief provided by the two-earner deduction

would be greatest for couples with incomes between

$50,000 and $100,000, for whom marriage penalties

would fall by nearly one-half. By contrast, couples at

the bottom of the income distribution would gain an

average of less than one-tenth of their current marriage

penalties, largely because their penalties arise from the

standard deduction and the earned income tax credit,

whereas the two-earner deduction reduces penalties de-

riving from the tax brackets. Because of the $3,000

limit on the deduction, couples at the top of the income

distribution would recoup only one-third of their current

penalties.

The two-earner deduction would benefit not only

couples currently incurring marriage penalties but also

many families now in a bonus position. For example, a

two-earner childless couple in which the lower-earning

spouse earns $5,000 of the couple's $50,000 in earn-

ings would pay $75 less in 1996 taxes because of the

deduction, even though they would already receive a

marriage bonus of more than $1,900. About one-fifth

of all couples who currently receive bonuses would ben-

efit from the two-earner deduction with average tax

savings of about $170. As a consequence, 20 percent

of the revenues lost because of the deduction would

increase existing bonuses or create new ones (see Fig-

ure 8). Just one-half of this amount would go to cou-

ples now receiving bonuses. About one-sixth of that

52 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box 8.

The Two-Earner Deduction

A couple with $50,000 in total earnings

$30,000

earned by the husband and $20,000 by the wife

would

have incurred a marriage penalty of $195 under 1996

tax law. The penalty results from a smaller standard

deduction under joint filing than under individual filing.

If the couple could also have claimed a two-earner de-

duction equal to 10 percent of the wife's earnings, their

taxable income would have been reduced by $2,000 (10

percent of $20,000), their tax liability would have fallen

by $300 (15 percent of $2,000), and their $195 mar-

riage penalty would have become a bonus of $105.

Couple

With

Under Two-Earner

Husband

Wife Current Law Deduction

Adjusted Gross Income $30,000 $20,000 $50,000 $50,000

Less personal exemptions 2,550 2,550 5,100 5,100

Less standard deduction 4,000 4,000 6,700 6,700

Less two-earner deduction 0 0 0 2,000

Equals taxable income 23,450 13,450 38,200 36,200

Taxable at 15 percent 23,450 13,450 38,200 36,200

Tax Liability 3,518 2,018 5,730 5,430

Marriage Penalty (+) or Bonus (-) $195 -$105

As a Percentage of Adjusted Gross Income 0.4 -0.2

SOURCE: Congressional Budget Office.

amount would go to the roughly 1 million couples now

neither paying penalties nor receiving bonuses who

would also benefit from the two-earner deduction. The

final one-third would go to couples incurring penalties

but would exceed the amount of penalty they pay.

Restoring the two-earner deduction could induce

some married couples to change whether and how much

they work. Depending on the earnings of the lower-

earning spouse, the change would affect couples in one

of three ways. First, for one-earner couples, the deduc-

tion would lower the marginal tax rate that nonworking

spouses would face if they entered the labor force and

would thus increase their after-tax wages. That in-

crease could induce them to go to work.

Second, two-earner couples in which the lower-

earning spouse earned less than $30,000 and thus for

whom the $3,000 cap would not limit the deduc-

tion would experience a similar increase in the after-

tax wages of lower-earning spouses, which could lead

them to work more hours. At the same time, however,

because the deduction would raise the couple's after-tax

income even if neither spouse changed hours of work,

such a couple would want to work less. The two in-

ducements offset one another and the net effect on labor

supply is indeterminate.

Finally, the deduction would increase the after-tax

income of two-earner couples whose spouses both earn

more than $30,000, but would have no effect on their

CHAPTER FIVE REDUCING MARRIAGE PENALTIES 53

after-tax wage rate because additional earnings would

not generate additional deductions. Those couples

would tend to work fewer hours because of the gain in

income and would have no offsetting effect of higher

net wages. The total effect on the labor supply of re-

storing the two-earner deduction could be up or down,

depending on the relative numbers of those three types

of couples and how much each would change its work

patterns.

Modif

Earned Income Tax

Credit to Reflect Number of

Adult Earners

Low-income families with children can incur significant

marriage penalties because of limits on the earned in-

come tax credit. If both husband and wife work, the

phaseout of the EITC on the basis of their combined

income can lead to the loss of some or all of the EITC

to which one of them might be entitled individually.

(See Box 5 for an example in which the loss of EITC

benefits would cost a couple one-eighth of their pretax

income, compared with what would happen if they

could file as single taxpayers.) Modifying the EITC to

take account of whether one or both spouses work

would alleviate this marriage penalty.

A wide range of options for modifying the EITC to

provide larger credits for two-earner couples could re-

duce marriage penalties but would also encounter prob-

lems of cost and complexity. At one extreme, each par-

ent could receive the EITC on the basis of the current

credit parameters and their earnings and total incomes.

That liberalization would remove completely any mar-

riage penalties associated with the EITC, but would

eliminate the existing limits on family income for re-

ceiving the credit, reduce federal revenues and increase

spending by a total of about $14 billion in 1996, and

grant EITC to more than 11 million families not now

eligible for the credit. Couples in which one spouse had

earnings low enough to qualify for the EITC under cur-

rent credit parameters would receive the credit, regard-

less of how much the other spouse earned. As a result,

one-third of the tax reduction would go to couples with

incomes above $50,000.

Modifying that approach to phase out individual

credits based on family rather than individual income

but setting higher levels for couples would limit the

credit to lower-income families and hold down revenue

losses. Each spouse would be entitled to an EITC

based on individual earnings, but the couple's EITC

would be reduced by a percentage of the amount by

which their total income exceeds a given level. The rate

of reduction and the income above which it becomes

effective would determine the maximum income a fam-

ily could have without losing the credit entirely. For

example, if couples were limited to collecting EITC

only if their total adjusted gross income were less than

twice the current limits, the combined revenue loss and

spending increase would drop to about $10 billion in

1996, only 10 percent of which would go to couples

with incomes above $50,000.

Both revenue losses and spending increases would

be smaller under an alternative that set EITC parame-

ters for couples equal to twice those for single and

head-of-household tax filers. That option would effec-

tively require spouses to share earnings equally and

then qualify for the EITC as individuals using the cur-

rent credit parameters. Compared with the previous

approach, this option would reduce the amount of EITC

going to couples in which husband and wife have sub-

stantially different incomes. Combined revenue losses

and spending increases would have totaled nearly $4

billion in 1996, virtually all of which would have gone

to couples with incomes below $50,000. Marriage pen-

alties resulting from the EITC would decline or disap-

pear for about one-third of couples currently penalized,

reducing those penalties by about one-fourth. The ap-

proach would be highly effective in targeting couples

now incurring EITC marriage penalties; less than 1 per-

cent of total revenue losses and spending increases

would go to couples currently receiving EITC marriage

bonuses. Couples with incomes between $20,000 and

$50,000 would receive about 90 percent of the mar-

riage penalty relief, and almost all of the rest would go

to couples with incomes below $20,000. As would be

the case for each of the previous EITC options, this

approach would extend the credit to couples with in-

comes far above current limits $28,495 in 1996.

More than 3.7 million couples would become newly

eligible for the EITC.

More restrictive liberalizations of the EITC could

ease the marriage penalties created by the credit while

54 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

generating smaller total revenue losses and spending

increases and without extending the credit to high-in-

come couples. The more restrictive the changes, how-

ever, the fewer couples now penalized would get relief

and the smaller the fraction of current penalties would

be removed. Granting the credit to couples based on

individual earnings but phasing the credit out more

quickly than under current law would hold down reve-

nue losses, but would be much less effective in reducing

marriage penalties and the disincentives for low-income

people to marry and work.

An alternative to modifying the EITC would be

replacing it with a taxable child allowance. If the al-

lowance was available to all taxpayers with children,

receipt would be unaffected by marriage and the allow-

ance would generate neither penalties nor bonuses. Al-

though making child allowances taxable would main-

tain progressivity, extending them to all families, re-

gardless of income, would require markedly increasing

outlays or reducing benefits to low-income families

substantially below current levels of the EITC. In any

case, this alternative would involve major changes in

the nature of the redistribution currently accomplished

by the EITC.

Restore the Re

uirement That

ouses File Individual

Returns

Before 1948, the individual income tax imposed no

marriage penalties because husbands and wives had to

file their own individual returns. Although shifting to a

joint filing requirement might have fit the times when

most couples had just one worker, the preponderance of

two-earner couples in the 1990s may call for a return to

individual taxation. Such a change would not only re-

move all marriage penalties but would also eliminate

marriage bonuses. The net effect would be a shift in

the burden of income taxes away from two-earner cou-

ples and toward one-earner couples.

By eliminating differences in couples' tax liabilities

brought about by marriage, the change would improve

the fairness of the tax system between married couples

and otherwise similar pairs who are not married and

reduce the disincentives to work caused by joint income

taxation with progressive rates. Revenue gains from

removing marriage bonuses would more than offset

revenue losses resulting from the elimination of penal-

ties and would raise net federal revenues by an esti-

mated $4 billion measured at 1996 income levels. That

revenue gain assumes that, if eligible, each spouse

could file as a head of household and claim the EITC.

Denying either of those options would generate greater

revenue gains.

Requiring individual filing would, however, cause a

large redistribution of tax liabilities from those now in-

curring penalties to those now receiving bonuses. As-

suming that there would be no changes in marital status

or earnings because of the changed tax law, nearly 21

million couples now incurring penalties would save an

average of nearly $1,400 in taxes, and another 25 mil-

lion couples currently benefiting from bonuses would,

on average, pay about $1,300 more in taxes than they

do under current law.

Requiring all people to file individual tax returns

would mean that couples with identical incomes but

different divisions of those incomes between husband

and wife would pay different taxes, an outcome that

some people would view as unfair. Differences in the

way couples earn their incomes may, however, warrant

varying levels of taxes. A couple in which only one

spouse works is better off than a two-earner couple

with the same income if the first couple's nonworking

spouse is free to provide services that the second couple

might have to purchase. Taxing couples on the basis of

their ability to pay would require that the first couple

pay more in taxes than the second, which is what would

happen under a system of individual taxation.

At the same time, taxing one-earner couples more

than two-earner couples with the same dollar income

could be considered antifamily because it could induce

more couples to have two earners and thus require child

care facilities outside the family. Furthermore, two

couples in which all spouses work full time could have

markedly different divisions of income between hus-

band and wife and therefore pay different amounts of

income tax, even though both their time and monetary

resources would be identical. Only by further com-

plicating the tax code, however, could the Congress

CHAPTER FIVE REDUCING MARRIAGE PENALTIES 55

eliminate such inequities; even then, additional issues

affecting ability to pay would remain.

Returning to individual filing for married couples

would cause other problems. First, the change would

have offsetting effects on work incentives. On the one

hand, by lowering the tax rate imposed on the first dol-

lar earned by a lower-earning spouse, the shift would

reduce the work disincentives created by the tax system.

On the other, the marginal tax on the income of the

higher-earning spouse could be raised, imposing work

disincentives that could lead to a reduction in the num-

ber of hours that spouse would work. The net effect is

ambiguous.

A second issue involves the conflict between indi-

vidual filing requirements and state community prop-

erty laws. As noted earlier in the historical discussion,

community property states hold that a couple's income

is divided evenly between husband and wife, therefore

each spouse should pay taxes on one-half of the cou-

ple's income. Requiring that spouses pay individual

taxes on their own incomes would run afoul of the 1930

Supreme Court decision affirming that couples in com-

munity property states can split their total incomes

evenly for tax purposes. A return to mandated individ-

ual filing would have to deal with the community prop-

erty issue.

Finally, individual filing by spouses would require

rules governing the allocation of jointly received in-

comes and of deductions and exemptions. One ap-

proach would allow couples to divide those items in any

way they wanted, although allowing total freedom

would make tax filing more complicated as couples

tried to determine their tax-minimizing allocations.

Alternatively, the law could require a joint return only

for income received jointly, a practice followed by

many other industrialized countries. That requirement

would raise the costs of complying with the tax code

and would impose additional burdens on the Internal

Revenue Service. Finally, the law could mandate par-

ticular divisions of joint incomes, deductions, and ex-

emptions between spouses, thereby limiting the number

of calculations a couple would have to make but none-

theless complicating the tax code.

Allow Cou

les Choice

of Filin

Status

Giving couples the choice of filing jointly, as individu-

als, or, if eligible, as heads of household, would elimi-

nate all marriage penalties at the cost of large losses of

revenue. Because spouses could choose to file individ-

ually or as household heads, marriage could not, by it-

self, increase a couple's tax liability. However, allow-

ing joint tax returns would retain current marriage bo-

nuses. Couples receiving bonuses would continue to do

so, whereas those now incurring penalties would shift

to individual or head-of-household returns. In 1996,

this option would have reduced federal tax revenues by

an estimated $29 billion, the total value of all marriage

penalties. Disallowing the option of filing as head-of-

household would decrease the revenue loss. The ap-

proach would also encounter the same difficulties with

dividing joint incomes, deductions, and exemptions be-

tween spouses as the option to require individual filing.

Under a less extreme approach, couples would be

required to file joint returns but would be given, on the

basis of a simplified tax calculation, a credit equal to all

or part of their marriage penalty. Using a simplified

calculation would increase the administrative feasibility

of that option and hold down its costs for example,

counting only earned income and allowing only stan-

dard deductions and personal exemptions. Even so, re-

quiring that a couple seeking relief from marriage pen-

alties prepare two individual tax calculations in addi-

tion to completing their full joint return would impose

significant compliance costs on filers and large ad-

ministrative costs on the Internal Revenue Service.

Furthermore, if the marriage penalty incurred by a cou-

ple does not derive from differences in tax brackets and

standard deductions for single and joint filers, then bas-

ing credits on a simplified tax calculation would not

offset all penalties. For example, the marriage penal-

ties incurred by low-income families with children re-

sult principally from the EITC; only if credit calcula-

tions included the EITC would this approach provide

significant relief from marriage penalties for low-

income families.

56 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

A credit based only on earnings, standard deduc-

tions, personal exemptions, and joint-versus-single fil-

ing statuses would reduce the total marriage penalty by

about one-third at a cost of about $10 billion in forgone

revenues. Because it would offset most of the penalties

deriving from tax brackets, this approach would lower

penalties most for higher-income families; on average,

couples with incomes above $50,000 would see their

marriage penalties fall by more than one-half.

Conversely, because the credit would not take the

EITC into account the major cause of marriage penal-

ties for low-income couples families with incomes

below $20,000 would have less than one-tenth of their

penalty removed. In addition, some couples currently

receiving marriage bonuses would qualify for a credit.

Roughly one-sixth of the total credit would increase

existing bonuses or create new ones (see Figure 8).

Fundamental Tax Reform

In recent years, calls for a fundamental restructuring of

the federal tax system have increased, ranging from a

flatter income tax to substituting consumption for in-

come as the object of taxation. The precise effects of

any such reforms on marriage penalties depend on the

specific details of the new tax, but it is possible to pre-

dict some general outcomes as a result of particular

kinds of changes.

Flat Taxes

One approach to fundamental tax reform would replace

the current individual income tax with a flat (single

rate) tax on labor and business incomes. Under one

formulation, individual taxpayers could deduct only a

personal allowance defined by tax filing status and a

fixed personal allowance for each dependent before

imposing a single tax rate on the remaining (taxable)

income. Business incomes would be taxed separately.

A flat tax would reduce and possibly eliminate

marriage penalties. Because just one tax rate would

apply, there could be no penalty from tax brackets.

Furthermore, if the tax eliminated the EITC, that credit

would no longer impose marriage penalties. Only dif-

ferences in personal allowances for tax filers would

generate penalties. If the allowance for a single filer

was exactly one-half that for a couple, marriage penal-

ties would occur only if unmarried taxpayers with de-

pendents could file as heads of household with larger

personal allowances than they would receive as single

filers. Other versions of the flat tax that would elimi-

nate all exemptions and deductions beyond personal

allowances would have similar effects

Marriage bonuses would continue under a flat tax if

one spouse was unable to use the full value of personal

allowances to reduce taxable income. Any couple in

which only one member paid positive taxes before mar-

riage would receive a marriage bonus because the pair

could reduce their taxable income by up to the differ-

ence between joint and single personal allowances and

thus pay lower taxes. Only if neither spouse owed any

taxes before marriage or both spouses paid positive

taxes before marriage, and marriage did not change

their total deductions and exemptions, would they nei-

ther incur a marriage penalty nor receive a marriage

bonus.

Retail Sales Taxes and

Value-Added Taxes

Replacing the current individual income tax with a re-

tail sales tax or a value-added tax would eliminate all

marriage penalties and bonuses. Because a retail sales

tax (RST) or a value-added tax (VAT) would tax con-

sumption and involve no personal deductions or exemp-

tions, only differences in levels of consumption would

produce differences in tax liability. Under an RST or

VAT, couples would pay higher or lower taxes after

marriage than before only if their consumption

changed. Marriage itself would have no effect on their

taxes. Most proposals for an RST or a VAT include

provisions for tax relief for low-income families

through either a rebate or an income-based credit. Re-

bates or credits, depending on how they were struc-

tured, could reintroduce marriage penalties and bo-

nuses.

Appendixes

Appendix A

Tax Treatment of Married Couples

in Other Countries

ost industrialized countries tax married cou-

ples differently than does the United States.

In 1993, of the 27 countries in the Organiza-

tion for Economic Cooperation and Development

(OECD), 19 imposed income taxes separately on hus-

bands and wives, as the United States did before 1948

(see Box A-1). In those countries, marriage has little

effect on a couple's tax liability on earnings, but cou-

ples with identical incomes can owe substantially dif-

ferent amounts of taxes, depending on the division of

earnings between husband and wife. In general, how-

ever, the 19 countries tax spouses' combined invest-

ment earnings, either at the higher of their marginal tax

rates or through a legally determined division of asset

income between them for tax purposes.

Four countries France, Luxembourg, Portugal,

and Switzerland follow an opposite approach. They

grant marriage bonuses to virtually all families and pro-

vide child bonuses as well. For example, the French

system requires taxpayers to combine all income earned

by parents and their children and then to split the in-

come among all family members. Children are

weighted half as much as parents. Individual tax sched-

ules then apply to each of the split incomes. A couple

with two children would thus pay the same tax rate as a

single person with one-third of their family income.

Under progressive tax rates, such income-splitting

nearly always results in a marriage bonus.

Among the OECD countries, only three tax couples

on their combined incomes as the United States does.

Germany allows complete income-splitting between

spouses, just as the United States did between 1948 and

1969. Because both singles and couples face the same

tax rates, German couples generally receive a marriage

bonus. Ireland and Norway, like the United States, im-

pose taxes on each couple's combined income using

separate tax schedules that mitigate the effects of joint

taxation and can result in marriage penalties or bo-

nuses.

Individual taxation has dominated only since 1970.

Before that year, only six OECD countries taxed

spouses separately. By 1980, seven countries had

shifted away from joint or family taxation to individual

taxation, and three more countries followed suit in 1989

and 1990. In the same period, no countries moved in

the other direction, and any reversal in this pattern is

highly unlikely.

Individual taxation generally applies only to earn-

ings. Countries that require couples to file individual

returns treat property income in one of two ways. Some

countries ignore the division of ownership of assets

between spouses and tax income from all of a couple's

assets as income to the higher-earning spouse. Conse-

quently, the higher marginal tax rate of the spouse who

earns more applies to all property income. Other coun-

tries require a specific division of property income be-

tween spouses, either evenly or in proportion to earn-

ings. The marginal tax rate of each spouse thus applies

to part of the couple's property income. In either case,

legal ownership of assets does not matter in determin-

ing tax liability and couples gain no advantage from

manipulating ownership to minimize their tax bills.

60 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box A-1.

Tax Treatment of Earned Income in OECD Countries, 1993

Individual Taxation Joint Taxation Family Taxation

Australia Germany France

Austria Ireland Luxembourg

Belgium Norway Portugal

Canada United States Switzerland

Czech Republic

Denmark

Finland

Greece

Hungary

Iceland

Italy

Japan

Mexico

Netherlands

New Zealand

Spain

Sweden

Turkey

United Kingdom

SOURCE: Congressional Budget Office, derived from Organization for Economic Cooperation and Development, part IV, "Description of the

Tax/Benefit Systems,"

The Tax/Benefit Position of Production Workers, Annual Report, 1990-1993

(Paris: Organization for Economic

Cooperation and Development, 1994), pp. 107-228.

a. Taxes are based on total income from all family members and composition of family.

Appendix B

Treatment of Married Couples Under

State Income Taxes

he states vary widely in the way they tax the

incomes of couples (see Box B-1). Eight states

have no income tax and therefore do not face

the issue. Six additional states impose a single tax rate

on all tax units, regardless of income level or filing sta-

tus. Because marriage does not affect the levels of de-

duction and personal exemption in any of those states,

couples incur no marriage penalties and receive bonuses

only if one spouse has earnings below the level of indi-

vidual exemptions and deductions. The remaining 36

states and the District of Columbia fall into three

groups that treat the taxation of married couples differ-

ently.

States with Joint Return

Rate Schedules

Fifteen states have schedules that reduce or eliminate

any tax penalty on marriage. Eight of those states cre-

ate tax brackets for married couples by doubling the

width of the rate brackets for individual filers, follow-

ing the pattern of the federal schedules used between

1948 and 1969. That approach creates marriage bo-

nuses for virtually all couples. The other seven states

widen but do not double the brackets for married cou-

ples, thereby lowering penalties below what they would

be if couples had to use the rate brackets for individual

filers but leaving marriage penalties in place for some

couples. The last approach mirrors the current federal

income tax, under which the width of tax brackets for

married couples is generally two-thirds larger than that

for single filers.

States Allowin

Cou

les to

File Se

arate Returns

on a Sin

le Form

Nine states and the District of Columbia allow hus-

bands and wives to use one return but pay taxes on

their separate incomes as if they were single. Because

couples have the choice of whether to file separately,

they can only face a marriage bonus, which generally

occurs when the earnings of the husband and wife are

substantially different. Such an option was effectively

available on the federal level to couples in community

property states before 1948; couples in those states

could elect either to report their individual incomes or

to divide their total income equally between spouses.

The 10 jurisdictions allowing spouses to file state

tax returns individually on a single form face a prob-

lem namely, how to tax property income. Because of

the ease with which couples can switch ownership of

assets between spouses, at least some of those jurisdic-

tions allow couples to choose the division of income

from jointly owned property between themselves if they

elect to file separately on a single return. Even so, sep-

arate filers must report income from their individually

owned assets as that of the owner. In such cases, cou-

62 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Box B-1.

Treatment of Married Couples Under State Income Taxes, 1993

States with

Joint Rate

Schedules for

States Allowing Married Couples, States in Which States with Flat

Married Couples to Reducing or Married Couples Tax Rates and

File Separately on a Eliminating May Face a Therefore No States with No

Single Tax Return Marriage Penalty Marriage Penalty Marriage Penalty Income Tax

Arkansas Alabama Georgia Connecticut Alaska

Delaware Arizona Kansas Illinois Florida

District of Columbia California Maryland Indiana Nevada

Iowa Colorado New Jersey Massachusetts New Hampshire

Kentucky Hawaii New Mexico Michigan South Dakota

Mississippi Idaho North Dakota Pennsylvania Texas

Missouri Louisiana Ohio Washington

Montana Maine Oklahoma Wyoming

Tennessee Minnesota Rhode Island

Virginia Nebraska South Carolina

New York Vermont

North Carolina West Virginia

Oregon

Utah

Wisconsin

SOURCE: Derived from Advisory Commission on Intergovernmental Relations, "Budget Processes and Tax Systems,"

Significant Features of Fiscal

Federalism 1994

, vol. 1 (Washington, D.C.: Advisory Commission on Intergovernmental Relations, June 1994), Tables 16 and 17, pp.

54-62.

a. Standard deductions and exemptions for couples equal the sum of those of spouses filing as individuals and hence do not impose marriage penalties.

b. These states fully eliminate any marriage penalty by allowing married couples to use rate brackets double the width of those available to single

filers.

c. North Dakota, Rhode Island, and Vermont levy state income taxes as a percentage of federal income tax liability and therefore impose marriage

penalties proportional to those at the federal level.

ples could still shift asset ownership to the spouse with

the lower marginal tax rate to minimize their combined

tax liability.

States in Which Cou

les Ma

Face Marria

e Penalties

Twelve states impose marriage penalties and bonuses

similar to those at the federal level. Because North

Dakota, Rhode Island, and Vermont levy taxes as a per-

centage of federal liability, couples filing in those states

face penalties or bonuses proportional to those in the

federal income tax. Couples in other states may incur

marriage penalties because they must use tax schedules

the same as or not very different from those used by

single people. In general, those states require couples

who file joint federal returns also to file joint state re-

turns.

Appendix C

Characteristics of Married and

Unmarried Couples, 1995

he Current Population Survey (CPS) provides

data that allow comparison of married couples

and self-identified unmarried couples. Since

1995, the CPS has asked respondents whether they had

unmarried partners and were therefore couples who

were not married but living together in relationships

that have many of the characteristics of marriages. The

March 1996 CPS identified nearly 2.7 million such

couples, 97 percent of which were opposite-sex cou-

ples. The survey also recorded 54.7 million married

couples. Self-identified unmarried couples thus consti-

tuted about 5 percent of all couples.

An alternative method used for many years by the

census to identify unmarried couples defines them as

unrelated adults of opposite sexes living together in a

household with no other adults present. This definition

finds many more unmarried couples than does self-

identification: 3.8 million in March 1996 versus 2.7

million self-identified couples. The number of unmar-

ried couples identified under this definition has in-

creased rapidly over the past quarter century from 0.5

million in 1970 to 1.2 million in 1980, 2.9 million in

1990, and 3.8 million in 1996.

Demo

hic Differences

As a group, unmarried couples earn lower incomes, are

younger, and have fewer children than married couples.

Unmarried couples had markedly lower incomes in

1995 than their married counterparts: 43 percent of

unmarried couples and 30 percent of married couples

had incomes below $30,000. In the same year, 42 per-

cent of married couples and 30 percent of those not

married had incomes above $50,000 (see Table C-1).

In 1995, one-half of all unmarried couples were

under age 35, compared with only one-fifth of married

couples. At the other end of the age spectrum, one-

third of married couples but less than one-tenth of un-

married couples were 55 years old or older (see Table

C-2).

The two groups were more similar in terms of num-

bers of children: 46 percent of married couples and 42

percent of unmarried couples had children in 1995.

Among those who had children, however, more than 60

percent of married couples had two or more children

compared with less than half of unmarried couples (see

Table C-3).

Differences in Simulated

Marria

e Penalties and

Bonuses

Given the differences between married and unmarried

couples, it is not surprising that simulations of marriage

penalties and bonuses for the two groups are dissimilar

(see Table C-4).

64 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Based on the March 1996 CPS, the simulations

indicate that 49 percent of married couples incur penal-

ties and 39 percent have bonuses. By contrast, 56 per-

cent of unmarried couples would bear penalties and 33

percent would receive bonuses if they were required to

file joint tax returns. Estimated average bonuses for the

two groups are similar $1,140 and $1,020 but aver-

age penalties for married couples are more than 50 per-

cent larger than those for their unmarried counter-

parts nearly $1,200 versus about $770. Much of the

difference probably occurs because unmarried couples

as a group have substantially lower incomes than mar-

ried couples.

APPENDIX C CHARACTERISTICS OF MARRIED AND UNMARRIED COUPLES, 1995 65

Table C-1.

Distribution of Married and Unmarried Couples by Total Income, 1995

Less Than $10,000- $20,000- $30,000- $50,000- $75,000- More Than

$10,000 $20,000 $30,000 $50,000 $75,000 $100,000 $100,000 Total

Thousands of Couples

Married 2,400 6,500 7,850 14,610 12,290 5,610 5,390 54,670

Unmarried 190 420 530 740 470 170 150 2,650

Percentage of Couples

Married 4121427221010100

Unmarried 716202818 6 6100

SOURCE: Congressional Budget Office tabulations of data from the March 1996 Current Population Survey.

a. Unmarried couples are those living together and reporting as partners.

66 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table C-2.

Distribution of Married and Unmarried Couples by Age of Older Spouse, 1995

Under 65 and All

Age 25 25-34 35-44 45-54 55-64 Over Ages

Thousands of Couples

Married 1,150 9,230 14,290 11,820 8,100 10,070 54,670

Unmarried 350 980 680 400 150 90 2,650

Percentage of Couples

Married 21726221518100

Unmarried 13 37 26 15 6 3 100

SOURCE: Congressional Budget Office tabulations of data from the March 1996 Current Population Survey.

a. Unmarried couples are those living together and reporting as partners.

APPENDIX C CHARACTERISTICS OF MARRIED AND UNMARRIED COUPLES, 1995 67

Table C-3.

Distribution of Married and Unmarried Couples by Number of Children, 1995

Two or

No One More All

Children Child Children Couples

Thousands of Couples

Married 29,300 9,570 15,800 54,670

Unmarried 1,530 570 550 2,650

Percentage of Couples

Married 54 18 29 100

Unmarried 58 22 21 100

SOURCE: Congressional Budget Office tabulations of data from the March 1996 Current Population Survey.

a. Unmarried couples are those living together and reporting as partners.

68 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table C-4.

Simulated Marriage Penalties and Bonuses for Married and Unmarried Couples, 1995

Married Unmarried

Couples Couples

Number of Couples (Thousands)

With penalties 26,800 1,480

Unaffected 6,350 300

With bonuses 21,520

870

Total 54,670 2,650

Percentage of Couples

With penalties 49 56

Unaffected 12 11

With bonuses 39

Total 100 100

Average Penalty 1,200 770

Average Bonus 1,140 1,020

Total Penalties (Millions of dollars) 32,040 1,140

Total Bonuses (Millions of dollars) 24,540 890

SOURCE: Congressional Budget Office simulations based on data from the March 1996 Current Population Survey.

a. Unmarried couples are those living together and reporting as partners.

Appendix D

Additional Supporting Tables

he following four tables D-1 through D-4

provide additional information about marriage

penalties and bonuses. Table D-1 shows the

relative importance of factors that affect the tax liabili-

ties of married couples. Table D-2 illustrates the distri-

bution of married couples by the level of earnings of

each spouse. Table D-3 demonstrates how the division

of earnings between spouses influences the likelihood

that a couple would incur a penalty or receive a bonus,

as well as the size of the penalty or bonus. Table

D-4 supporting Table 7 and Table 8 in Chapter 3

reveals how penalties and bonuses vary with the num-

ber of earners that a couple has.

70 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table D-1.

Factors Determining Marriage Penalties and Bonuses by Adjusted Gross Income, Simulated 1996

All Less Than $20,000- $50,000- More Than

Return Category Incomes $20,000 $50,000 $100,000 $100,000

Thousands of Returns

With Penalty 20,880 1,120 8,050 8,990 2,730

Tax penalty only 10,760 30 2,360 5,970 2,400

EITC penalty only 880 490 240 150 a

Both tax and EITC penalties 6,840 300 4,150 2,190 210

Tax penalty exceeds EITC bonus 00000

EITC penalty exceeds tax bonus 2,400 290 1,310 680 120

With Neither Penalty nor Bonus 3,130 2,320 260 550 10

With Bonus 25,260 5,760 10,040 7,170 2,290

Tax bonus only 20,030 4,160 8,100 5,800 1,970

EITC bonus only 680 670 10 0 0

Both tax and EITC bonuses 1,100 580 520 0 0

Tax bonus exceeds EITC penalty 3,450 350 1,410 1,370 320

EITC bonus exceeds tax penalty 00000

All Returns 49,270 9,190 18,351 16,700 5,020

Percentage of Returns in Income Category

With Penalty 4212445454

Tax penalty only 22 0 13 36 48

EITC penalty only 25110

Both tax and EITC penalties 14 3 23 13 4

Tax penalty exceeds EITC bonus 00000

EITC penalty exceeds tax bonus 53742

With Neither Penalty nor Bonus 6 25 1 3 0

With Bonus 51 63 55 43 46

Tax bonus only 41 45 44 35 39

EITC bonus only 17000

Both tax and EITC bonuses 26300

Tax bonus exceeds EITC penalty 74886

EITC bonus exceeds tax penalty 00000

All Returns 100 100 100 100 100

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: EITC = earned income tax credit.

a. Less than 5,000.

APPENDIX D ADDITIONAL SUPPORTING TABLES 71

Table D-2.

Married Couples by Earnings of Each Spouse, Simulated 1996 (Percentage of all married couples)

Earnings of Higher-Earning Spouse

Earnings of All

Lower-Earning No Less Than $10,000- $20,000- $30,000- $50,000- $75,000- More Than Earnings

Spouse Earnings $10,000 $20,000 $30,000 $50,000 $75,000 $100,000 $100,000 Levels

No Earnings 10765741244

Less Than $10,000 * 344631121

$10,000 - $20,000 * * 35621a15

$20,000 - $30,000 * * * 3 6 2 a a 11

$30,000 - $50,000 ****421a7

$50,000 - $75,000 *****1aa1

$75,000 - $100,000 ******aaa

More Than $100,000 *******aa

All Earnings Levels 10 10 13 17 28 14 4 5 100

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: * = Situation cannot occur because lower-earning spouse must have earnings less than those of higher-earning spouse.

a. Less than 0.5 percent.

72 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table D-3.

Married Couples with Marriage Penalties or Bonuses and Average Penalty or

Bonus by Earnings of Each Spouse, Simulated 1996

Earnings of Higher-Earning Spouse

All

No Less Than $10,000- $20,000- $30,000- $50,000- $75,000- More Than Earnings

Measure Earnings $10,000 $20,000 $30,000 $50,000 $75,000 $100,000 $100,000 Levels

Lower-Earning Spouse Had No Earnings

Percentage with Penalties000000000

Average Penalty (Dollars) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Percentage with Bonuses 86 68 88 98 99 99 99 99 89

Average Bonus (Dollars) 1,410 1,060 1,010 1,010 1,270 2,210 2,550 3,270 1,470

Lower-Earning Spouse Earned Less Than $10,000

Percentage with Penalties * 59 71 70 58 38 25 24 58

Average Penalty (Dollars) * 760 1,110 1,110 1,090 1,070 1,030 1,360 1,060

Percentage with Bonuses *3929304262757642

Average Bonus (Dollars) * 340 580 580 800 1,260 1,350 2,100 890

Lower-Earning Spouse Earned $10,000 to $20,000

Percentage with Penalties * *96927274716782

Average Penalty (Dollars) * * 1,760 1,130 1,410 1,460 1,550 1,760 1,400

Percentage with Bonuses * * 1 4 24 26 29 33 15

Average Bonus (Dollars) * * a 60 390 570 820 1,390 490

Lower-Earning Spouse Earned $20,000 to $30,000

Percentage with Penalties * * * 82 85 94 82 84 86

Average Penalty (Dollars) * * * 720 950 1,200 1,290 2,270 1,000

Percentage with Bonuses * * * 5 7 6 17 15 7

Average Bonus (Dollars) * * * 10 70 240 330 510 140

Lower-Earning Spouse Earned $30,000 to $50,000

Percentage with Penalties****9399979896

Average Penalty (Dollars)****1,5701,8901,9303,4501,830

Percentage with Bonuses****20311

Average Bonus (Dollars)****an.a.11064050

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

APPENDIX D ADDITIONAL SUPPORTING TABLES 73

Table D-3.

Continued

Earnings of Higher-Earning Spouse

All

No Less Than $10,000- $20,000- $30,000- $50,000- $75,000- More Than Earnings

Measure Earnings $10,000 $20,000 $30,000 $50,000 $75,000 $100,000 $100,000 Levels

Lower-Earning Spouse Earned $50,000 to $75,000

Percentage with Penalties*****99999999

Average Penalty (Dollars)*****2,2802,8405,2303,090

Percentage with Bonuses*****0bbb

Average Bonus (Dollars)*****n.a.ccc

Lower-Earning Spouse Earned $75,000 to $100,000

Percentage with Penalties******979999

Average Penalty (Dollars)******4,0707,3006,120

Percentage with Bonuses******0bb

Average Bonus (Dollars)******n.a.cc

Lower-Earning Spouse Earned More Than $100,000

Percentage with Penalties*******9999

Average Penalty (Dollars)*******12,05012,050

Percentage with Bonuses*******bb

Average Bonus (Dollars)*******cc

All Earnings Levels of Lower-Earning Spouse

Percentage with Penalties 0 15 41 56 55 52 46 35 42

Average Penalty (Dollars) n.a. 760 1,400 1,030 1,240 1,550 1,850 3,900 1,380

Percentage with Bonuses 86 60 53 40 41 48 53 64 51

Average Bonus (Dollars) 1,410 940 930 890 1,020 1,800 2,030 2,930 1,300

NOTES: Percentages in each earnings category may not add to 100 percent because some couples neither incur penalties nor receive bonuses.

n.a. = not applicable.

* = Situation cannot occur because lower-earning spouse must have earnings less than those of higher-earning spouse.

a. Less than 50 cents.

b. Less than 0.5 percent.

c. Too few cases to provide meaningful measure.

74 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table D-4.

Distribution of Marriage Penalties and Bonuses by Number of Earners, Simulated 1996

Number of Earners in Couple All

None One Two Returns

Joint Returns (Millions)

With Penalty 0 0 20,900 20,900

Unaffected 700 1,600 800 3,100

With Bonus 4,100

15,300 5,900 25,300

All Returns 4,800 16,900 27,500 49,300

Joint Returns (Percent)

With Penalty 0 0 76 42

Unaffected 14 10 3 6

With Bonus 86

90 21 51

All Returns 100 100 100 100

Total Value for Joint Returns (Millions of dollars

)

Penalty 0 0 28,800 28,800

Bonus -5,800 -22,600 -4,400 -32,900

Average Penalty or Bonus for Joint Returns (Dollars)

Penalty n.a. n.a. 1,380 1,380

Bonus -1,410 -1,480 -760 -1,300

Net Effect on Tax Liability -5,800 -22,600 24,300 -4,100

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

NOTE: n.a. = not applicable.

a. Positive values indicate net penalties; negative values indicate net bonuses.

Appendix E

Sources of Data

he analysis reported in this study uses two

sources of data about married couples: the

Internal Revenue Service's Statistics of In-

come (SOI) and the Current Population Survey (CPS)

from the Bureau of the Census. The SOI provides the

basis for estimating the magnitudes of marriage penal-

ties and bonuses in Chapter 3 and the effects of the op-

tions described in Chapter 5. The CPS was used to ex-

amine labor market and demographic changes since

1969, discussed in Chapter 4.

Statistics of Income

This study uses data from the 1993 SOI to estimate

marriage penalties and bonuses. The SOI is an annual

database created by the Internal Revenue Service (IRS).

The 1993 file consists of extensive data reported on the

individual federal income tax returns filed by a

nonrandom sample of nearly 105,000 taxpayers during

1994. The IRS selects sample returns based on their

usefulness in analyzing the income tax and in modeling

tax policy. In practice, the likelihood of the IRS select-

ing a return is inversely related to the amount of income

reported; that is, the file contains relatively more re-

turns with large positive or negative incomes than re-

turns with smaller incomes. For example, the 1993

SOI includes all individual tax returns reporting income

greater than $10 million or losses greater than $5 mil-

lion, but only one out of every 3,000 returns with posi-

tive income less than $30,000. The IRS assigns a

weight to each sample return equal to the inverse of its

probability of selection. Weighted counts of returns

thus represent all returns filed during 1994.

For each selected tax return, the SOI reports each

entry on the basic return (1040, 1040A, or 1040EZ, as

appropriate) as well as most of the information in-

cluded on accompanying schedules. In addition, the

1993 SOI includes information taken from W-2 forms

reporting the wage and salary incomes of husbands and

wives separately. (Joint tax returns show only the com-

bined earnings of husband and wife.) The SOI thus

provides most of the information about incomes needed

to simulate the taxes that spouses would pay if they

could file individual rather than joint returns. Note,

however, that the file does not indicate how the couple

would divide up dependents, itemized deductions, or

income from assets. The simulations used in the analy-

sis must therefore rely on assumptions about those divi-

sions, as described in Chapter 3.

Besides the basic limitations inherent in any data

set that only samples a portion of the entire population,

the SOI is further limited because it does not represent

the entire population. Because it samples only taxpay-

ers who filed returns during a given year, the SOI ex-

cludes nonfilers. In particular, the file underrepresents

people with incomes too low to owe taxes, other than

those who file tax returns to claim the earned income

tax credit or refunds of taxes withheld by employers.

Any analysis that examines the entire population must

thus rely on another source of data.

Current Po

ulation Surve

The CPS is an annual microdata file created by the Bu-

reau of the Census that represents the noninstitu-

76 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

tionalized U.S. population. Each month the Census

Bureau interviews members of approximately 60,000

households concerning their demographic and employ-

ment characteristics, primarily to obtain information for

estimating unemployment statistics. In March of each

year, the bureau supplements the basic CPS with ques-

tions about incomes from a variety of sources received

during the previous calendar year by each member of

sampled households. Each reporting unit (household,

family, and individual) is assigned a weight equal to the

inverse of the probability of that unit's selection for in-

terviewing. Weighted tabulations of variables included

in the March CPS thus provide representative estimates

of values for the entire noninstitutionalized population.

Data files are available for the annual March CPS be-

ginning in the late 1960s.

The CPS has a variety of shortcomings that limit its

usefulness for analysis. First, incomes reported to the

CPS are substantially less than the values that indepen-

dent sources of data would predict, particularly for un-

earned income such as transfer income, interest, rents,

and dividends. Second, the Census Bureau "topcodes"

income values, reporting high incomes only as being

above a fixed level. Thus, for example, the CPS shows

earnings from a single job only up to $99,999. In the

March 1996 CPS, values above the topcoding level are

reported as the mean of all topcoded values. In earlier

years, the CPS only showed the topcode value for

amounts above that level. For all years, the CPS thus

provides only limited information about units with the

highest incomes.

Topcoding makes it impossible to determine the

precise division of earnings between spouses for high-

income couples. In those cases less than 1 percent of

couples with incomes above $100,000 in 1995 the

analysis reported in Chapter 4 categorized both spouses

as having more than one-third of the couple's combined

earnings. Finally, because the CPS has changed over

time, the data collected in different years are not fully

comparable. Changes in data collection methods, the

questionnaire used, and levels of topcoding all create

differences that can be only partially corrected.

Despite its shortcomings, the Current Population

Survey provided the information necessary to examine

demographic changes in working-age married couples

over time, presented in Chapter 4.

Appendix F

Estimated Marriage Penalties and

Bonuses Under a Divorce Model

he estimated magnitudes of marriage penalties

and bonuses provided in Chapter 3 assume

that the higher-earning spouse would claim

most of a couple's children as dependents if the couple

were allowed to file individual tax returns. The lower-

earning spouse would claim at most one child and that

is only if the couple had at least two children. That as-

signment of children for the purpose of tax filing gen-

erally but not always results in a smaller tax liability

for the couple under individual filing.

An alternative approach would assign all children

to the lower-earning spouse, more closely approximat-

ing the fact that children generally go with the mother

when couples divorce and that, on average, women earn

less than men. If the simulations reported in Chapter 3

used this alternative assumption to estimate the taxes

couples would owe if they could file individual returns,

they would find fewer couples incurring penalties and

more receiving bonuses (see Table F-1). Simulated

total bonuses would be two-thirds larger up from $33

billion to $55 billion and total penalties would drop

from $29 billion to $25 billion. The estimated net bo-

nus resulting from requiring married couples to file

joint tax returns would thus jump from $4 billion to

$30 billion. In addition, the estimated average penalty

incurred by penalized couples would fall from $1,380

to $1,300, whereas the average bonus for couples re-

ceiving one would increase by half from $1,300 to

$1,960.

Two points are worth noting. First, the changes in

estimated penalties and bonuses as a result of reassign-

ing children occur because shifting children from the

tax return of the higher-earning spouse to that of the

lower-earner raises the couple's tax liability under indi-

vidual filing. The higher earner generally loses the op-

tion of filing a head-of-household return and thus faces

a higher tax rate and loses personal exemptions for the

children. Although some lower-earning spouses would

lose the benefits of the earned income tax credit under

the alternative assumption, others would have exemp-

tions they could not use fully and possibly lower tax

rates that would apply to low or no taxable income.

When higher individual tax liabilities provide the base-

line against which penalty or bonus status is deter-

mined, penalties shrink and bonuses grow.

Second, although assigning all children to the

lower-earning spouse may better represent the way chil-

dren would be allocated if the couple were to divorce,

estimates of penalties and bonuses should use that as-

signment only if the policy alternative to joint tax filing

is divorce and consequent individual filing. If the rele-

vant alternative is not divorce but individual filing un-

der which the couple chooses how to allocate children

between returns, the assignment used in the simulations

for Chapter 3 is more appropriate.

78 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table F-1.

Marriage Penalties and Bonuses for Married Couples Under Alternative Assumptions

about Distribution of Children Between Spouses, Simulated 1996

All Children

Children Assigned

Assigned to Lower-

as in Earning

Simulations Spouse

Number of Couples (Thousands)

With Penalties 20,900 19,300

Unaffected 3,100 1,800

With Bonuses 25,300

28,100

Total 49,300 49,300

Percentage of Couples

With Penalties 42 39

Unaffected 6 4

With Bonuses 51

Total 100 100

Total Value (Millions of dollars)

Penalties 28,800 25,200

Bonuses 32,900 55,200

Average Penalty or Bonus (Dollars)

Penalty 1,380 1,300

Bonus 1,300 1,960

SOURCE: Congressional Budget Office simulations based on data from the 1993 Statistics of Income.

a. The couple's first child is assigned to the higher-earning spouse, the second child to the lower-earning spouse, and all other children to the higher-

earning spouse.

Appendix G

The Income Distribution

of Families

orking-age married couples generally have

higher incomes than their unmarried counter-

parts and elderly families, and thus tend to

face higher tax rates because of the progressivity of the

federal income tax. Data from the March 1996 Current

Population Survey indicate that just over one-third of

all families had incomes below $20,000 in 1995, a sim-

ilar fraction had incomes between $20,000 and

$50,000, about one-fifth had incomes between $50,000

and $100,000, and about one-sixteenth had incomes ex-

ceeding $100,000 (see Table G-1).

Working-age single parents those 25 through 64

years old had the lowest incomes in 1995, with 55

percent receiving less than $20,000 annually and only 8

percent receiving $50,000 or more. Working-age,

childless, single people are somewhat better off: 43

percent received less than $20,000, and 13 percent re-

ceived $50,000 or more. Elderly families had incomes

similar to those of working-age single parents.

Largely because nearly three-fourths of them had at

least two incomes, working-age married couples had the

highest incomes. One-seventh of those without chil-

dren had incomes of $100,000 or greater, and 42 per-

cent had incomes between $50,000 and $100,000.

Only one-tenth had incomes below $20,000. Their

counterparts with children were nearly as well off, with

one in eight families receiving at least $100,000 and

two-fifths receiving between $50,000 and $100,000.

The last quarter century has seen a greater concen-

tration of working-age couples who have incomes

above $50,000 (in 1995 dollars) and a smaller fraction

of couples with incomes below that level (see Table G-

2). Between 1969 and 1995, couples with total income

above $50,000 (in 1995 dollars) became relatively

more common, increasing from 38 percent to 56 per-

cent of all couples. Most of the shift in income shares

took place during the 1970s. There was movement out

of the middle-income ranges to both higher- and lower-

income categories during the 1980s, but little change

has occurred in the distribution since 1989. Those

shifts reflect two offsetting factors. First, the earnings

of younger workers declined markedly between 1969

and 1995, leading to lower incomes for younger fami-

lies. At the same time, the aging of the baby-boom

generation and the rising proportion of two-earner cou-

ples have increased the incomes of many couples. Many

observers have suggested that the increasing number of

two-earner couples has been a response to declining

wages.

80 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table G-1.

Percentage Distribution of Families by Income Category and Family Type, 1995

Working-Age

Working-Age Single Working-Age

Childless Working-Age Individuals Couples

Income of Single Childless with with Elderly All

Taxable Unit Individuals Couples Children Children Families Families

Less Than $20,000 43 10 55 10 53 36

$20,000 - $50,000 43 33 36 38 34 37

$50,000 - $100,000 11 42 7 40 10 21

More Than $100,000 2

15 1 12 3 6

All Families 100 100 100 100 100 100

SOURCE: Congressional Budget Office tabulations of data from the March 1996 Current Population Survey.

NOTES: Working-age individuals are people between ages 25 and 64, inclusive. Working-age couples are those in which both spouses are in that

age range.

Families are groups of related individuals living together. Individuals not living with relatives are included as one-person families.

APPENDIX G THE INCOME DISTRIBUTION OF FAMILIES 81

Table G-2.

Distribution of Working-Age Married Couples with Earnings by Income Category

in 1969, 1979, 1989, and 1995

Couple's Total Income

(1995 dollars) 1969 1979 1989 1995

Less Than $20,000 8 6 7 8

$20,000 -$50,000 54 42 37 37

$50,000 -$100,000 33 43 43 42

More Than $100,000 4 9 13 14

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

Appendix H

Comparison of Demographic Changes for

Working-Age Couples and All Married Couples

he demographic analysis reported in this study

relies primarily on data from the March Cur-

rent Population Survey, issued annually by the

Bureau of the Census. Those data include demographic

and income information for a sample of about 150,000

people, along with weights to allow the calculation of

statistics representing the noninstitutional population of

the United States. The analysis focuses on married

couples during the 1969-1993 period.

Because marriage penalties and bonuses primarily

affect couples who work, the analysis covers only

working-age couples with earnings. That limitation

imposed the following three requirements on couples:

both spouses must be between ages 25 and 64, inclu-

sive; neither spouse can report negative earnings (in

general, negative earnings result from losses in self-em-

ployment activities); and at least one spouse has to

have positive earnings.

Those limitations eliminated 27 percent of all mar-

ried couples from the analyzed 1993 population (see

Table H-1). Of the 27 percent, 12 percent violated the

age limits but not the earnings requirements, another 12

percent satisfied neither the age nor income limits, and

3 percent failed to meet the income requirements in

cases in which both spouses were in the appropriate age

range.

In the case of more than one-half of the couples

who were removed from the analysis because of age,

both spouses were 65 years old or older. Another one-

fifth had elderly husbands. Among nearly all of the

other excluded couples, either the wife or both spouses

were under age 25. Just 5 percent of the omitted cou-

ples were dropped from the analysis because the hus-

band was too young or the wife too old. Of the 3 per-

cent of couples excised solely on the basis of earnings,

nearly all had no earnings.

Limiting the analyzed population to working-age

couples increased the proportion of couples with two

earners, largely because many of those omitted were

elderly couples who did not work (see Table H-2).

Both populations, however, had substantial increases in

proportions of two-earner couples over the 1969-1995

period.

Text tables 9 through 15 show the demographic

changes that occurred between 1969 and 1995 for

working-age married couples with earnings. Tables H-

3 through H-9 repeat those tables, adding comparable

data for all married couples to allow comparison of the

changes for the smaller and the more inclusive groups.

The tables show similar qualitative trends with differ-

ences that would be expected when older, younger, and

nonworking couples were excluded. For example, Ta-

ble H-3 shows that all married couples followed the

same trend of declining shares of one-earner couples

and greater equality of earnings between spouses as

was observed for working-age couples with earnings.

Including couples with spouses over age 64 or under

age 25 increases the share of couples with one earner,

but does not alter the fact that such couples were sub-

stantially less common in 1995 than in 1969. Similar

observations hold for other demographic changes.

84 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-1.

Selection of Married Couples for Analysis Based on Age and Earnings, 1995

Number Distribution

Category (Thousands) (Percent)

All Married Couples 54,650 100

Couples Excluded by Age but Not Earnings

Both spouses age 65 or over 1,400 3

Husband age 65 or over, wife age 25 through 64 1,550 3

Husband age 25 through 64, wife under age 25 1,250 2

Both spouses under age 25 1,150 2

Other 500 1

Couples Excluded by Earnings but Not Age

Both spouses have zero earnings 1,650 3

At least one spouse reports negative earnings 100 a

Couples Excluded by Both Age and Earnings

Both spouses age 65 or over 5,650 10

Husband age 65 or over, wife age 25 through 64 1,100 2

Husband age 25 through 64, wife under age 25 50 a

Both spouses under age 25 50 a

Other 200 a

Married Couples Selected for Analysis 40,100 73

SOURCE: Congressional Budget Office tabulations of data from the March 1996 Current Population Survey.

a. Less than 0.5 percent.

b. Both spouses age 25 through 64, neither spouse has negative earnings, and at least one spouse has positive earnings.

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 85

Table H-2.

Percentage of Couples with Two Earners: All Married Couples and

Working-Age Married Couples with Positive Earnings in 1969, 1979, 1989, and 1995

Population Group 1969 1979 1989 1995

All Married Couples

with Earnings 45.2 51.7 58.0 58.4

Working-Age Married Couples

with Positive Earnings 48.3 59.6 70.5 71.8

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

86 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-3.

Distribution of Working-Age and All Married Couples by Division

of Earnings Between Spouses in 1969, 1979, 1989, and 1995

Lower-Earning

Each Spouse Spouse Had Some

Earned at Least Earnings but Less Neither Spouse or

One-Third of Than One-Third of Only One

Year Couple's Earnings Couple's Earnings Had Earnings All Couples

Thousands of Working-Age Married Couples

1969 5,600 10,100 16,700 32,400

1979 7,500 12,700 13,700 34,000

1989 11,800 15,100 11,300 38,200

1995 13,600 15,200 11,300 40,100

Percentage of All Working-Age Married Couples

1969 17.2 31.2 51.7 100.0

1979 22.2 37.4 40.4 100.0

1989 30.9 39.5 29.5 100.0

1995 33.9 37.9 28.2 100.0

Thousands of Married Couples

1969 7,200 13,200 24,700 45,100

1979 9,300 15,900 23,600 48,800

1989 13,300 17,600 22,400 53,300

1995 14,800 17,100 22,800 54,700

Percentage of All Married Couples

1969 16.0 29.2 54.8 100.0

1979 19.2 32.6 48.3 100.0

1989 25.0 33.0 42.0 100.0

1995 27.1 31.3 41.6 100.0

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 87

Table H-4.

Percentage Distribution of Working-Age and All Married Couples

with Two Earners by Total Income in 1969, 1979, 1989, and 1995

Couple's Income (1995 dollars)

Less Than $20,000- $50,000- More Than

Year $20,000 $50,000 $100,000 $100,000 All Incomes

Working-Age Married Couples

1969 28.5 46.0 62.7 43.0 48.3

1979 29.5 56.3 73.4 56.7 59.6

1989 38.0 66.6 82.7 78.7 70.5

1995 30.9 69.7 87.0 78.6 71.8

All Married Couples

1969 21.5 49.1 63.5 42.7 45.2

1979 17.7 58.8 73.6 56.3 51.7

1989 18.3 67.4 82.3 78.1 58.0

1995 15.4 69.6 86.4 78.0 58.4

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

88 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-5.

Percentage Distribution of Working-Age and All Married Couples by Division of Earnings

Between Spouses and Total Earningsin 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

Working-Age Married Couples

All Income Categories

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

Income Less Than $20,000

1969 7.1 21.4 71.5

1979 9.7 19.9 70.5

1989 12.3 25.7 62.0

1995 11.6 19.3 69.1

Income $20,000 to $50,000

1969 14.9 31.1 54.0

1979 20.3 36.0 43.7

1989 27.3 39.4 33.4

1995 32.1 37.6 30.3

Income $50,000 to $100,000

1969 27.4 35.4 37.3

1979 29.3 44.1 26.6

1989 40.2 42.5 17.3

1995 44.4 42.7 13.0

Income More Than $100,000

1969 9.9 33.1 57.0

1979 16.0 40.6 43.3

1989 32.5 46.3 21.3

1995 32.5 46.1 21.4

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

(Continued)

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 89

Table H-5.

Continued

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

All Married Couples

All Income Categories

1969 16.0 29.2 54.8

1979 19.2 32.6 48.3

1989 25.0 33.0 42.0

1995 27.1 31.3 41.6

Income Less Than $20,000

1969 5.3 16.2 78.5

1979 5.1 12.6 82.3

1989 5.5 12.8 81.7

1995 5.3 10.1 84.6

Income $20,000 to $50,000

1969 16.3 32.7 50.9

1979 21.6 37.2 41.2

1989 27.7 39.7 32.6

1995 31.9 37.6 30.4

Income $50,000 to $100,000

1969 28.3 35.2 36.5

1979 29.7 44.0 26.4

1989 39.9 42.5 17.7

1995 43.8 42.6 13.6

Income More Than $100,000

1969 9.7 33.0 57.3

1979 15.7 40.6 43.7

1989 31.8 46.3 21.9

1995 31.4 46.6 22.0

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTES: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

Incomes measured in 1995 dollars.

90 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-6.

Percentage Distribution of Working-Age and All Married Couples with Two Earners

by Age of Older Spouse in 1969, 1979, 1989, and 1995

Age of Older Spouse All Age

Year 25-34 35-44 45-54 55-64 Groups

Working-Age Married Couples

1969 46.5 49.5 51.2 44.8 48.3

1979 67.7 63.5 59.8 44.8 59.6

1989 76.0 74.2 71.8 54.0 70.5

1995 76.7 74.1 74.2 57.9 71.8

All Married Couples

1969 54.4 49.5 50.9 29.4 45.2

1979 69.0 63.3 58.8 25.5 51.7

1989 74.6 73.4 70.5 27.4 58.0

1995 74.9 72.8 72.6 28.3 58.4

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 91

Table H-7.

Percentage Distribution of Working-Age and All Married Couples by Division of Earnings

Between Spouses and Age of Older Spouse in 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

Working-Age Married Couples

All Age Groups

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

Older Spouse 25-34 Years Old

1969 15.2 31.4 53.5

1979 27.4 40.3 32.3

1989 36.2 39.8 24.0

1995 39.9 36.7 23.3

Older Spouse 35-44 Years Old

1969 14.7 34.8 50.5

1979 21.9 41.6 36.5

1989 32.2 42.0 25.8

1995 34.8 39.3 25.9

Older Spouse 45-54 Years Old

1969 19.5 31.7 48.8

1979 21.0 38.7 40.2

1989 31.0 40.8 28.2

1995 33.8 40.4 25.8

Older Spouse 55-64 Years Old

1969 19.3 25.4 55.2

1979 17.8 27.0 55.2

1989 21.4 32.6 46.0

1995 25.8 32.1 42.1

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

(Continued)

92 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-7.

Continued

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

All Married Couples

All Age Groups

1969 16.0 29.2 54.8

1979 19.2 32.6 48.3

1989 25.0 33.0 42.0

1995 27.1 31.3 41.6

Older Spouse 25-34 Years Old

1969 19.2 35.2 45.6

1979 27.4 41.6 31.0

1989 34.7 40.0 25.4

1995 37.0 37.9 25.1

Older Spouse 35-44 Years Old

1969 14.5 35.0 50.5

1979 21.5 41.8 36.7

1989 31.4 42.0 26.6

1995 34.1 38.7 27.2

Older Spouse 45-54 Years Old

1969 19.1 31.8 49.1

1979 20.2 38.6 41.2

1989 30.0 40.6 29.5

1995 33.0 39.6 27.4

Older Spouse 55-64 Years Old

1969 11.8 17.6 70.6

1979 9.9 15.6 74.5

1989 10.4 17.0 72.6

1995 12.0 16.3 71.7

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 93

Table H-8.

Percentage Distribution of Working-Age and All Married Couples with Two Earners

by Number of Children in 1969, 1979, 1989, and 1995

Two or More

Year No Children One Child Children All Couples

Working-Age Married Couples

1969 53.0 53.3 43.2 48.3

1979 59.9 65.0 56.4 59.6

1989 70.4 75.9 67.2 70.5

1995 71.6 76.9 69.1 71.8

All Married Couples

1969 43.4 54.0 43.0 45.2

1979 44.5 63.9 55.2 51.7

1989 48.4 72.8 65.3 58.0

1995 48.4 74.2 67.4 58.4

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.

94 FOR BETTER OR FOR WORSE: MARRIAGE AND THE FEDERAL INCOME TAX June 1997

Table H-9.

Percentage Distribution of Working-Age and All Married Couples by Division of Earnings

Between Spouses and Number of Children in 1969, 1979, 1989, and 1995

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

Working-Age Married Couples

All Couples

1969 17.2 31.2 51.7

1979 22.2 37.4 40.4

1989 30.9 39.5 29.5

1995 33.9 37.9 28.2

No Children

1969 24.4 28.6 47.0

1979 27.0 32.9 40.1

1989 34.3 36.2 29.6

1995 36.3 35.3 28.4

One Child

1969 19.3 34.1 46.7

1979 24.6 40.4 35.0

1989 34.9 41.1 24.1

1995 37.6 39.4 23.1

Two or More Children

1969 11.3 31.9 56.8

1979 16.3 40.1 43.6

1989 25.0 42.2 32.8

1995 29.2 39.9 30.9

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

(Continued)

APPENDIX H COMPARISON OF DEMOGRAPHIC CHANGES FOR WORKING-AGE COUPLES AND ALL MARRIED COUPLES 95

Table H-9.

Continued

Both Spouses Had

Each Spouse Earnings and One Spouse

Earned at Least Contributed More Neither Spouse or

One-Third of Than Two-Thirds of Only One

Year Couple's Earnings Couple's Earnings Had Earnings

All Married Couples

All Couples

1969 16.0 29.2 54.8

1979 19.2 32.6 48.3

1989 25.0 33.0 42.0

1995 27.1 31.3 41.6

No Children

1969 19.7 23.7 56.6

1979 19.9 24.7 55.5

1989 23.0 25.4 51.6

1995 23.8 24.6 51.6

One Child

1969 18.0 36.0 46.0

1979 22.8 41.1 36.1

1989 32.1 40.8 27.2

1995 35.5 38.7 25.8

Two or More Children

1969 10.9 32.1 57.0

1979 15.7 39.6 44.8

1989 23.9 41.4 34.7

1995 28.2 39.2 32.6

SOURCE: Congressional Budget Office tabulations of data from the March Current Population Surveys for 1970, 1980, 1990, and 1996.

NOTE: Working-age married couples include only those in which both spouses are between ages 25 and 64, neither spouse has negative earnings,

and at least one spouse has positive earnings.