Examining the Effect of the Earned Income Tax Credit on the Labor Market Participation of Families on Welfare

Examining the Effect of the Earned Income Tax Credit
on the Labor Market Participation of Families on Welfare





V. Joseph Hotz
Department of Economics and NBER
University of California, Los Angeles
E-mail: hotz@ucla.edu

Charles H. Mullin
Bates White
E-mail: charlie.mullin@bateswhite.com

John Karl Scholz
Department of Economics, Institute for Research on Poverty and NBER
University of Wisconsin–Madison
E-mail: jkscholz@facstaff.wisc.edu





December 2005












We thank Glen Cain, Liz Davis, Janet Holtzblatt, Tom MaCurdy, Bruce Meyer, and numerous seminar
participants for comments, and Dawn Duren, Eduardo Fajnzylber, Webb Hester, Doug Houston, Jacob
Klerman, Doug McKee, Oscar Mitnik, Don Oellerich, Juan Pantano, George Ramsey, Pat Ruggles, and
Werner Schink for their help with this project. We also thank colleagues at the California Department of
Social Services; the California Franchise Tax Board; RAND; the National Science Foundation; and the
Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation,
who collectively helped support this work. None of these people or organizations, however, are
responsible for the views expressed in this paper.

---

Abstract
This paper examines the employment effects of the earned income tax credit (EITC). We use a
unique dataset, created by matching administrative data from public assistance records, unemployment
insurance records, and federal tax returns for a sample of California residents. We conduct a set of four
tests to assess our ability to isolate the causal effects of the EITC on employment.
The first test is based on the intuition that if the EITC alters employment, all else being equal,
employment rates for two-or-more child families should grow relative to the employment rates of one-
child families, as credit amounts available to these groups of families diverged over the 1990s. The
second test examines whether or not people eligible for the EITC actually file tax returns and claim it.
The third test is based on the intuition that, if the EITC, and not other factors such as the strong economy
in the 1990s, is causing employment differences between families with two or more children relative to
those with one child, we should expect to see no employment differences (after conditioning on other
characteristics) between families with two children and families with three or more children, since the
EITC did not change differentially for the latter two groups. The fourth test conditions the sample on
those who do not file tax returns and again examines employment changes in the 1990s for families with
two or more children relative to families with one child.
Using fixed-effects empirical employment models estimated on a sample of single-parent
families, our coefficient estimates are consistent with the EITC having a substantial, positive effect on the
employment of families who have used or will use welfare.

---

Examining the Effect of the Earned Income Tax Credit
on the Labor Market Participation of Families on Welfare
Between 1990 and 1999, real spending on the earned income tax credit (EITC) increased to $31.9
billion from $9.6 billion (in 1999 dollars). It is by far the largest cash or near-cash antipoverty policy in
the United States. Moreover, with phased-in increases enacted in 1990 and in 1993, it became the most
rapidly growing (substantial) item in the federal budget.
Employment rates of single women with children also rose sharply over this period, and welfare
caseloads fell precipitously. Between March 1990 and March 2000, employment rates of single women
with children rose to 73.9 percent from 55.2 percent. Welfare caseloads fell to 2.3 million families from
4.1 million (though caseloads rose until 1994, when they peaked at 5.0 million families).1
The coincident timing of these trends raises a question: did the EITC play a substantial role in the
increase in labor force participation of single women with children?2 Not surprisingly, given the important
role the EITC plays in the nation’s safety net, a number of previous studies examine this topic.3 These
papers come to a consistent conclusion: the EITC has a significant, empirically large effect on labor force
participation of single women with children.4

1Employment rates and caseload numbers come from Tables No. 577 and No. 544, respectively, from
Statistical Abstract of the United States 2000.
2Standard labor-leisure models unambiguously predict that EITC increases (or increases in other wage
subsidies) will increase employment among low-skilled workers.
3For published work, see Dickert, Houser, and Scholz (1995); Eissa and Leibman (1996); Keane and
Moffitt (1998); Ellwood (2000); Meyer and Rosenbaum (2000, 2001); Grogger (2003); and Eissa and Hoynes
(2004). Hotz and Scholz (2003) survey EITC research.
4Hotz and Scholz (2003) compute employment elasticities with respect to net income (associated with
EITC changes) from selected previous studies that range from 0.69 to 1.16. Grogger’s (2003) estimates were not
included in our earlier study, but he concludes that the EITC “may be the single most important policy measure for
explaining the decrease in welfare and the rise in work and earnings among female-headed families in recent years”
(p. 408). Eissa and Hoynes (2004) focus on the employment and hours decisions of secondary workers in married
families and find small, negative effects of the credit on work. Cancian and Levinson (2003) use data from the
National Survey of America’s Families and focus on Wisconsin’s supplemental EITC for families with three or
more children. Unlike other studies on the topic, they do not find statistically significant EITC employment effects.

---

2
At least three concerns arise with previous work examining the EITC’s employment effects. The
identification strategy in most papers is to compare employment changes of groups eligible for the EITC
(generally single mothers) with employment changes of groups ineligible for the EITC (generally single
women without children) before and after EITC increases.5 But other factors changed over the periods in
which the EITC increased, and these other factors might have different employment effects for women
with and without children. Many states, for example, made extensive changes to welfare both before and
after the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) abolished AFDC
in 1996. And aggregate GDP increased 120 consecutive months beginning in March 1991, resulting in the
longest economic expansion in U.S. history.
A second issue arises with the difference-in-difference approach: namely, that the composition of
the groups may change over time. There are a priori reasons to be concerned about potential
compositional bias in these repeated cross-sectional studies.6 Between 1984 and 1996, for example, the
period examined in Meyer and Rosenbaum (2001), the number of single mothers in the Current
Population Survey (CPS) increased 30.5 percent, to 7.7 million families from 5.9 million families.7 The
difference-in-difference identification strategy will be biased in indeterminate directions if employment
propensities of the sample changed over time after accounting for observed characteristics (Heckman,
1996).

5Keane and Moffitt (1998) estimate utility function parameters of households that are consistent with
participation decisions in multiple welfare programs, accounting for the precise budget sets that occur in the 1984
SIPP. Dickert, Houser, and Scholz (1995) do not observe EITC variation in their cross-sectional study and
consequently they infer the potential effects of the EITC from observing the correlation between employment and
differences in the return to work across states. These employment differences, however, may be due to unobserved
state-specific factors unrelated to differences in the return to work generated by the tax and transfer systems.
6All previous studies use repeated cross-sectional data from the Current Population Survey (CPS) and the
Survey of Income and Program Participation (SIPP) except for Cancian and Levinson (2003).
7As discussed in Hotz and Scholz (2002), the ratio of AFDC recipients reported in the CPS to
administrative counts of recipients fell from 86.7 percent in 1990 to 79.6 percent in 1996. The ratio of AFDC dollars
to administrative totals was 78.4 percent in the 1984 CPS and 67.7 percent by 1996. By all accounts the SIPP and
CPS are of high quality, but these trends raise a concern about the ability of studies using the CPS or SIPP to
accurately characterize the tax and transfer environment facing low-income families.

---

3
Third, all previous EITC employment studies have been indirect in the sense that they have
inferred the EITC’s effects based on the relationship between employment and EITC policy changes. But
if EITC changes cause employment changes, we should see similar changes in EITC claims. Because of
data limitations, no previous study has directly examined EITC claiming and employment.
This paper addresses these and other concerns. We use data only from California to mitigate the
influence of changes in welfare and other social policies on employment and to better account for local
labor market conditions.8 Using data with county identifiers, we are able to include considerably more
detailed covariates to proxy for local labor market conditions than have previous studies.9 Moreover,
California also made relatively modest changes to AFDC and TANF compared to other states in the
1990s. We nevertheless also condition on several features of the county-administered welfare programs in
California.
Ours is also the first EITC employment study to use longitudinal data, which allow us to account
for unobserved, household-specific, time-invariant factors that affect employment. Longitudinal data also
help us avoid bias due to (unobserved) changes in the composition of “treatment” and “comparison”
groups in repeated cross-sectional studies.
At the core of our empirical analysis is a set of four tests that we use to assess the causal effects
of the EITC on employment. The first test examines employment using a different identification strategy
than most previous studies that typically compare employment rates of women with children to other
groups. Beginning in 1991, the maximum EITC available to families with two or more children increased
relative to the EITC available for one-child families. While the difference in the maximum credit by
number of children was initially modest, starting in 1994, as a result of changes in the EITC included in

8See Ellwood (2000) for more on these issues.
9Bartik and Eberts (1999) criticize studies that include only unemployment rates to account for labor
market conditions, writing, the “… unemployment rate by itself may be a woefully incomplete measure of economic
conditions affecting potential welfare recipients.” They argue that more textured measures of the economic
environment facing welfare families matter in understanding caseload changes.

---

4
the 1993 federal budget act, this differential increased substantially, the maximum credit rising from $77
in 1993 to $490 in 1993 to $1,404 in 1996 (see Table 1). If the EITC alters employment, all else being
equal, employment rates for families with two or more children should grow relative to the employment
rates of families with one child, as credit amounts available to these groups of families diverge.10 Our
empirical approach formalizes this strategy as a “difference-in-difference” estimator, controlling for
covariates and household-specific fixed effects, to identify the effects of the EITC on employment.
The second test makes use of a unique match of administrative data on welfare receipt,
unemployment records, and the federal tax returns filed by California residents, which allow us to
examine directly the employment and EITC-claiming behavior of low-income families. No previous
EITC employment study has used data on whether or not people eligible for the EITC actually file tax
returns and claim it. If the employment patterns previous studies attribute to the EITC are, in fact, caused
by the credit, those who are affected by the credit should file and claim it.
If the EITC, and not other factors such as the strong economy in the 1990s, is causing
employment differences between families with two or more children relative to those with one child, we
should expect to see no employment differences (after conditioning on other characteristics) between
families with two children and families with three or more children, since the EITC did not change
differentially for the latter two groups. Our third test examines this proposition.
Our fourth test conditions the sample on those who do not file tax returns, and again examines
employment changes in the 1990s for families with two or more children relative to families with one
child. If the EITC is the factor causing employment changes in the data, we would expect to see no

10In a sensitivity analysis section, Meyer and Rosenbaum (2001) examine differences in labor market
effects generated by the EITC for families with two or more children relative to families with one child, as is done in
our paper. They find much smaller EITC effects in this alternative specification and the coefficients are insignificant
in their larger, CPS Outgoing Rotation Group sample. Cancian and Levinson (2003) use a similar approach and find
insignificant EITC employment effects.

---

5
differential employment changes for the two groups (again, after conditioning on other characteristics)
among those not filing tax returns (and hence not claiming the credit).
Using fixed-effects empirical employment models estimated on a sample of single-parent
families, our coefficient estimates are consistent with the EITC having a substantial, positive effect on the
employment of families who have or will use welfare. Patterns of EITC claiming mirror the employment
patterns. Moreover, the differential employment responses occur only for families with two or more
children relative to one child, and not for families with three or more children relative to families with
exactly two children. Finally, there is no differential expansion of employment for families with two or
more children relative to one-child families for those who do not file tax returns. If factors other than the
EITC were driving the employment patterns observed in the data, it is unlikely that they would only affect
those who file tax returns and not affect the employment of families with three or more children relative
to families with two children.
We also discuss sensitivity analyses that show our results are robust to alternative modeling
choices. The most striking aspect of the sensitivity analysis, however, is that the EITC has no discernible
effect on employment for AFDC-UP (two-parent) families.
The remainder of the paper is organized as follows. In Section 1 we describe the structure and
history of the EITC, the various data sources we use and how we construct our analysis sample of
households who were on welfare in California for some period of time during 1990s. In Section 2 we
present unadjusted difference-in-difference estimates of the annual effects of the EITC on employment
rates and rates of EITC claiming for the period 1992–2000. In Section 3 we present refined estimates for
both of these outcomes using multivariate regression methods. In Section 4 we present the results for the
four tests described above as well as the results of several additional sensitivity analyses we conducted.
We conclude the paper with a summary of our findings and a discussion of how they compare with those
found in previous studies.

---

6
1.
THE EITC AND CALIFORNIA ADMINISTRATIVE DATA
In 1999, taxpayers with two or more children could receive an EITC of 40 percent of income up
to $9,540, for a maximum credit of $3,816. Taxpayers (with two or more children) with earnings between
$9,540 and $12,460 receive the maximum credit. Their credit is reduced by 21.06 percent of earnings
between $12,460 and $30,585. Table 1 shows the complete evolution of income eligibility thresholds,
credit rates, and phase-out (or implicit tax) rates.
To receive the credit, taxpayers file their regular tax return and fill out the six-line Schedule EIC
that gathers information about qualifying children. The EITC is refundable, meaning that the Treasury
Department pays it out regardless of whether the taxpayer has any federal income tax liability. There are
several basic tests for EITC eligibility. The taxpayer’s earned income and adjusted gross income must be
below a threshold that varies by year and by family size. To receive the credit available to families with
children,11 the qualifying child must be younger than 19, younger than 24 if a full-time student, or any age
if totally disabled. The claimant must be the parent, the grandparent, or foster parent of the child.12 The
qualifying child must live with the taxpayer at least six months during the year.
A key development for the purposes of this paper was put in place as part of the 1990 EITC
expansions. After 1990, for the first time, families with two or more children were able to receive a larger
EITC than they could if they had only one child. The difference through 1993, however, never exceeded
$77. As part of the 1993 EITC expansion, the differences became much larger: the maximum difference
was $490 in 1994, $1,016 in 1995, and $1,404 in 1996 (and indexed for inflation thereafter). As noted in
the introduction, if the EITC alters employment, all else being equal, employment rates for two-or-more-

11A small credit available for childless taxpayers between the ages 24 in 65 with very low incomes was
added in 1994. The credit rate for these taxpayers is 7.65 percent and the maximum credit in 1999 is $347.
12Until late 1999, a foster child was any child for whom the claimant cared for “as if the child is his or her
own.” Now the caring stipulation still holds, but the child must also be placed in the home by an authorized
placement agency.

---

7
Table 1
Earned Income Tax Credit Parameters, 1987–2000 (in nominal dollars)
Diff. in Max
Credit:
Phase-In
Phase-In
Max.
2+ - 1 Child
Phase-Out
Phase-Out
Year
Rate (%)
Range ($)
Credit ($)
($)
Rate (%)
Range ($)
1987 14.0 0–6,080 851
10.0 6,920–15,432
1988 14.0 0–6,240 874
10.0 9,840–18,576
1989 14.0 0–6,500 910
10.0 10,240–19,340
1990 14.0 0–6,810 953
10.0 10,730–20,264
1991a 16.71
0–7,140 1,192

11.93
11,250–21,250
17.32
1,235
43
12.36
11,250–21,250
1992a 17.61
0–7,520 1,324

12.57
11,840–22,370
18.42
1,384
60
13.14
11,840–22,370
1993a 18.51
0–7,750 1,434

13.21
12,200–23,050
19.52
1,511
77
13.93
12,200–23,050
1994 23.61
0–7,750
2,038

15.98
11,000–23,755
30.02
0–8,245
2,528
490
17.68
11,000–25,296
7.653
0–4,000
306
7.65
5,000–9,000
1995 34.01
0–6,160
2,094

15.98
11,290–24,396
36.02
0–8,640
3,110
1,016
20.22
11,290–26,673
7.653
0–4,100
314
7.65
5,130–9,230
1996 34.01
0–6,330
2,152

15.98
11,610–25,078
40.02
0–8,890
3,556
1,404
21.06
11,610–28,495
7.653
0–4,220
323
7.65
5,280–9,500
1997 34.01
0–6,500
2,210

15.98
11,930–25,750
40.02
0–9,140
3,656
1,446
21.06
11,930–29,290
7.653
0–4,340
332
7.65
5,430–9,770
1998 34.01
0–6,680
2,271

15.98
12,260–26,473
40.02
0–9,390
3,756
1,485
21.06
12,260–30,095
7.653
0–4,460
341
7.65
5,570–10,030
1999 34.01
0–6,800
2,312

15.98
12,460–26,928
40.02
0–9,540
3,816
1,504
21.06
12,460–30,580
7.653
0–4,530
347
7.65
5,670–10,200
2000 34.01
0–6,920
2,353

15.98
12,690–27,413
40.02
0–9,720
3,888
1,535
21.06
12,690–31,152
7.653
0–4,610
353
7.65
5,770–10,380
Source: 1998 Green Book, Committee on Ways and Means, U.S. House of Representatives, U.S.
Government Printing Office, page 867; 1998 through 2000 parameters come from Publication 596,
Internal Revenue Service.
aBasic credit only. Does not include supplemental young child or health insurance credits.
1Taxpayers with one qualifying child.
2Taxpayers with more than one qualifying child.
3Childless taxpayers.

---

8
child families should grow relative to the employment rates of one-child families starting in 1994, as
credit amounts available to these groups of families diverge.
For those out of the labor market, the EITC provides an unambiguous, positive incentive to work.
We focus on this feature of the credit in this paper.13
1.1
The Core Administrative Data Sources
California’s Medi-Cal Eligibility Data System (MEDS) provides information on program
participation and demographic characteristics of participants in Medi-Cal, the California state Medicaid
program, and AFDC/CalWORKS, the California state welfare program.14 MEDS also provides AFDC
participation histories of individuals in our samples starting in 1986.
We measure labor force participation using quarterly data from 1991 through 2000 on
employment (and earnings) from the California Employment Development Department (EDD) Base
Wage Files. The EDD Base Wage File contains employer-reported taxable wage payments for jobs
covered by unemployment insurance (UI) and disability insurance (DI).15 Hotz and Scholz (2002) survey
studies that examine the accuracy and coverage of unemployment insurance data for the low-income
population. In brief, UI data exclude some groups of workers, such as the self-employed, military, federal
employees, and independent contractors. Nevertheless, employment rates derived from UI data appear to
be similar to those that result from survey data. We expect UI-based employment rates to be lower
because of coverage problems with flexible workers and independent contractors. But surveys suffer from
nonresponse, so undercounts in both data sources typically appear to be similar in practice.

13See Hotz and Scholz (2003) for a summary of papers examining the effects of the EITC on hours of work.
14More generally, the MEDS is a statewide administrative system that contains information on monthly
participation in the state’s Medicaid program (Medi-Cal), AFDC/TANF programs, as well as the Food Stamp, SSI,
and California’s General Assistance (GA) programs. MEDS data are used by the Department of Health and Human
Services for quality control analyses and are thought to be high quality for the welfare and Medicaid populations.
15The file generally includes individuals paid cash wages of more than $100 in a calendar quarter, and
domestic workers paid cash wages more than $750 in a calendar quarter.

---