AJ. Cataldo II VIRGINIA POLYTECHNIC INSTITUTE AND STATE UNIVERSITY

THE EARNED INCOME CREDIT HISTORICAL PREDECESSORS AND CONTEMPORARY EVOLUTION

Abstract: The Revenue Reconciliation Act of 1993 (RRA93) significantly expanded the earned income credit (EIC), which was changed to include lowincome taxpayers without dependents. Evolving, most directly, from the “workfare” plan (1972) proposed by Senate Finance Committee Chairman, Russel B. Long, and in response to President Nixon’s Family Assistance Program (FAP), the post1974 EIC was not the first of its kind. It had two predecessors.

The EIC of 1923 through 1931 benefitted taxpayers with or without dependents and excluded any “workfare” feature. A second EIC, in name only, was in effect for the 1934 through 1943 lax years.

This paper develops a historical framework for study of the post1974 EIC. This framework necessarily precedes any investigation of contemporary issues relating to the twentyyear history of the post1974 EIC which, unlike its first predecessor, appears destined to continue as a permanent, expanding mechanism for the delivery of basic subsistence to the “working poor.” The resolution of these contemporary issues will determine whether the post1974 EIC is destined to replace or continue to coexist with a (presumably) more costly welfare delivery system.

INTRODUCTION

Initially designed to partially offset the adverse and rapidly growing impact of increasing Social Security taxes on the working poor, the EIC has undergone several expansionary stages since first introduced (in its current form) for the 1975 tax year.1 The Revenue Reconciliation Act of 1993 (RRA93) eliminated the separate health insurance and newborn child components of the earned income credit (EIC) and provided for a revised, basic

I would like to thank Fritz Scheuren, Director of the Statistics of Income Division individuals of the Internal Revenue Service, for providing the 1989 sample data used for portions of this paper. Also, two anonymous reviewers for their insightful comments and suggestions.

‘Table 3 summarizes EIC phasein, flat, and phaseout ranges from 1975 through 1996. Exhibits 1, 2, and 3 provide graphical representations of the same.
Received October 1994 Revised January 1995 Accepted August 1995

EIC to include lowincome taxpayers without dependents. This, everevolving, post1974 EIC is not the first of its kind.
The first EIC was available for the 1923 through 1931 tax years. Unlike the post1974 EIC, the first EIC was never dependency exemptionbased, did not provide for a “refundable” credit, and did not, in its final form, seek to maintain any form of “workfare” or work incentive feature.

A new, revised EIC became available for the 1934 through 1943 tax years. This second EIC was a “credit” in name only. It was what today might be referred to as a “deduction” and eventually evolved into the current “standard deduction,” available to nonitemizer taxpayers.

This paper describes and distinguishes between the first (1923 through 1931), second (1934 through 1943), and current (post1974) EICs. It summarizes events leading to the post1974 EIC and provides a basic structural framework for analysis of the post1974 EIC period. It reviews many of the concerns raised by contemporary policymakers and researchers that remain unresolved (or are exacerbated) after RRA93.

The lack of widespread knowledge and acceptance of EIC advanced payment options and the failure of the RRA93 expansion of the EIC to provide for any form of wealthbased means test represent problems unlike those previously encountered by our tax collection system. However, the incompatibility of the “workfare” provisions of the contemporary EIC with existing welfare systems suggests that certain historical lessons have gone unnoticed by contemporary policymakers.

THE FORM OF THE EIC

The post1974 EIC contains a “work incentive” feature, increasing its political palatability and distinguishing it from a “negative income tax” or some other form of a “guaranteed minimum income.” Ammer and Aramer (1977, p. 284) define a “negative income tax” as

(a) form of welfare payment whereby all lowincome individuals and families receive a direct cash subsidy from the government that is sufficient to raise them to
2 A tax “credit” results in a dollar for dollar reduction in the taxpayer’s tax liability. A tax “deduction” is multiplied by a taxpayerspecific marginal tax rate to determine the taxpayer’s tax liability reduction. A tax credit, therefore, is more valuable (to any taxpayer) than a tax deduction for an equal amount.

subsistence level. The subsidy itself is the negative tax. Supporters . .. argue that it could replace all other welfare programs, along with the bureaucracy and alleged waste they engender. Critics, however, believe it would remove incentives to work … A version of this idea was put into practice in 1975, when all U.S. social security recipients received a supplementary check for $50, regardless of their income.

A negative income tax is a welfare program, providing a basic grant to individuals with no income. (The Aid to Families with Dependent Children Foster Care (AFDCFC) program is an example of a welfare program). As Exhibit 1 illustrates, the negative income tax is phasedout as income increases, until it eventually reaches zero and the individual begins to pay tax. The general form of the post1974 EIC is also illustrated by Exhibit I. Note that the EIC provides for a variation of a negative income tax, but differs from the true form in two respects: (1) the EIC increases as the taxpayer’s earned income (El) increases and (2) a taxpayer with no El receives no credit.

A true negative income tax would typically provide for a larger (smaller) subsidy as income decreases (increases). Therefore, the El requirements of the post1974 EIC, unlike other welfare programs, maintains a politically popular “worklare” element.

The new El credit addresses abilitytopay issues by reducing the impact of rising, regressive Social Security taxes on the working poor. This objective is achieved while simultaneously maintaining Social Security contributions characterized by progressive benefit structures (and based on the family unit).

The lowincome taxpayers intended to benefit from the post1974 EIC are typically not subject to progressive income tax rates. Therefore, EIC provisions are consistent with a family assistance philosophy while leaving intact the existing Social Security tax (and benefit) system.

Beginning on July 1, 1979, eligible taxpayers anticipating an EICbased refund had the option of receiving an advanced EIC payment (AEIC), reported as a reduction of the EIC on the taxpayer’s federal income tax return, and limited to the amount available for one qualifying child. In this respect, the EIC possesses a cash or nearcash feature and behaves like a negative income tax or welfare benefit to lowincome taxpayers.

The successful public acceptance of this alternative to welfare programs such as AFDCFC, etc., must necessarily precede as income increases, the negative income tax (subsidy) decreasesAmount of Income Subject to Tax the elimination of alternative, presumably more costly,3 basic subsistence delivery systems. However, as Yin and Forman [1993, p. 951] point out,”. . . almost none of the recipients obtains the benefit incrementally during the course of the year” and Holt [1994, p. 759] indicates that “… fewer than 1 in 200 EIC recipients takes advantage …” of the AEIC.

THE POST1993 EIC REINFORCES CERTAIN OPERATIONAL DEFINITIONS

RRA93 provides for the beneficial inclusion of lowincome taxpayers without dependents for post1993 tax years. Prior to RRA93, there was some disagreement with respect to the appropriate operational definition of the marriage tax penalty (MTP). Should the MTP include or exclude its largest component, the EIC [Rosen, 1987 and 1988, and Mclntyre, 1988]?
The ElCbased component of MTPs for the 1974 through 1993 tax years might have been eliminated from consideration as true MTPs, under the assumption that such amounts were attributable to the decision to have children, as distinguished from the decision to marry. Though this point was not addressed in the literature,4 the post1993 EIC, with its beneficial inclusion of lowincome taxpayers without dependents, eliminates the potential for such distinctions for post1993 tax years. Researchers interested in the historical relevance and magnitude of post1974 ElCdriven MTPs may find it useful to note this distinction between pre1994 and post1993 periods to the extent that it presents the need for research design modification.
Though the post1993 EIC will undoubtedly result in greater MTPs [Lipman and Williamson, 1994, and Polinsky, 1993], it has evolved from a qualified dependency exemptionbased poverty reduction measure, to one more inclusive of the general population. Though still “workfare”dependent, the inclusion of

‘ For example, the popular press has recently drawn attention to discussions of a return to the use of orphanages to replace AFDCFC group homes, etc. The administrative costs of such programs include overpayments to providers, which frequently go uncollccted, since the overriding concern of such agencies is the shortage of placement facilities (see, for example, Report by the Auditor General of California, 1986).

Costs of the postRRA93 EIC are projected to approximate about $25 billion in fiscal year 1998, roughly 150 percent of the federal share of the AFDCFC program [Yin and Forman, 1993, p. 951].

4 This may be the result of early emphasis on supplementing and/or eventually replacing the AFDCFC programs with the EIC. See Hoffman and Seidman [1990] and Campbell and Peirce [1980].

lowincome taxpayers without dependents as potential beneficiaries links the EIC more closely with general welfare programs and reduces its similarity to the dependentbased AFDCFC program.

HISTORICAL FRAMEWORK

Three EICs have emerged over discontinuous periods and in very different forms: the first EIC existed for a nine year period (1923 through 1931) and the postWorld War I benefits were primarily shortterm, economic stimulusmotivated, and available to all taxpayers; the second EIC lasted for a ten year period (1934 through 1943) over which this “credit” evolved to take the form of our current “standard deduction;” and the post1974 EIC has evolved and expanded during this twentyone years and continues changing with the passage of RRA93 (1974 through 1994).

THE FIRST EIC (1923 THROUGH 1931)

The Revenue Act of 1921, enacted November 23, 1921, and amended March 4, 1923 (see Revenue Act of 1924, H.R. 6715, Public Law No. 176, p. 264), sought to stimulate an economy recovering from World War I. Beginning with the 1923 (and extending through the 1931) tax year(s), a nonrefundable EIC was established and maintained. As described by Pechman [1987, pp. 109],
. . . the earned income allowance was granted in the form of a deduction that ranged from 10 to 25 percent of earned net income. In some years the deduction was allowed for normal tax purposes only; in others it was allowed for both normal tax and surtax. In all years a certain minimum amount of income ($3,000 or $5,000) was presumed to be earned whether it actually was or not, and the deduction was limited to a maximum ranging from $10,000 to $30,000. The tax value of the deduction was . . . never worth more than $495 for a family of two (in 192831)…

This first, nonrefundable EIC was originally formulated under the proposition that a distinction should be made favoring (disfavoring) earned (unearned) income. The preference of a system taxing lightly income earned relative to that resulting from investments was supported under the “ability to pay” principle, but was not without difficulty in administration. The discussion of this inequity and the administrative difficulties were addressed (see the Revenue Act of 1924, H.R. 6715, Public Law No. 176, p. 264):

The taxpayer who receives salaries, wages, and other earned income must each year save and set aside a portion of his income in order to protect him in case of sickness and in his old age, and in order to provide for his family upon his death. On the other hand, the person whose income is derived from investments already has his capital and is relieved of the necessity of saving to establish it.

The difficulty comes when an attempt is made to divide that income which is in part earned and in part unearned into the two classes. Such a segregation would involve either (1) treating as unearned that part of the taxpayer’s income which represents a reasonable return upon the capital invested and considering the remainder as earned or (2) treating as earned income an amount representing a reasonable allowance as, in connection with the administration of the excessprofits tax, salary for the personal services actually rendered by the taxpayer.

Initial proposals (see Ways and Means Committee, 68th Congress, 1st Session, House Report 179, p. 78) of a 25 percent tax reduction for taxpayers whose incomes were earned would have benefitted salaried and professional people, but would have excluded farmers (currently, Schedule F income) and selfemployed or small business persons (currently, Schedule C and certain Kl passthroughs for closely held S corporation income).

Because of the administrative difficulties in arriving at an equitable solution to the segregation of earned and unearned components of certain classes of income, the “earned” income credit became a misnomer. This first EIC was extended to all individuals subject to the normal tax.

Furthermore, it imposed no requirements with respect to the existence of a dependent or qualifying dependency exemption. In fact, upon review of the early part of this period, this first EIC would appear to have gone so far as to favor (disfavor) single (married) taxpayers by providing for minimum and maximum EICs of $20 ($12.50) and $75 ($555), respectively.
5 This credit could increase to as much as $90 when the entire personal exemption amount was taken by one spouse (i.e., married, filing separately) and in the event that no dependency exemptions were available [KixMiller & Baar, 1924, pp. 1920].

It is important to note, however, that these provisions were designed prior to/in the absence of our post1970 (and current) system of separate tax rate schedules for married and single taxpayers [Brozovsky and Cataldo, 1994, pp. 179180]. During this period, twoearner, married taxpayers were effectively permitted the option of using the same progressive tax rate table twice. Therefore, the “rate” component of the marriage tax bonus (MTB) associated with the pre1971 period tax rate schedules might, today, be perceived as more than adequately compensatory for the failure to extend additional preferential treatment to married taxpayers in the form of a larger EIC.

TABLE 1

The First EARNED INCOME CREDIT (1923 through 1931)

Total Returns Taxable Returns
Tax Total Average Average Taxable Average Average
Year Returns Wage EIC Returns EIC Tax Rate
1923 7,698,321 $1,844 $28.65 4,270,121 $51.65 2.67%
1924 7,369,788 $1,848 $ 4.16 4,489,698 $ 6.82 2.74%
1925 4,171,051 $2,336 $ 5.89 2,501,166 $ 9.82 3.35%
1926 4,138,092 $2,415 $ 5.96 2,470,990 $ 9.97 3.33%
1927 4,101,547 $2,491 $ 6.07 2,440,941 $10.21 3.68%
1928 4,070,851 $2,668 $ 8.55 2,523,063 $13.79 4.62%
1929 4,044,327 $2,769 $ 5.46 2,458,049 $ 8.98 4.04%
1930 3,707,509 $2,676 $ 6.71 2,037,645 $12.21 2.63%
1931 3,225,924 $2,581 S 5.42 1,525,546 $11.47 1.81%
192331 Averages $ 6.03 $10.41 3.275%

A summary of the average EIC for total and taxable returns for the 1923 through 1931 tax years (Internal Revenue Code Section (IRC §) 1200(a) of the Revenue Act of 1924) is provided in Table 1 [SOI, 1931, pp. 3743]. Though retained for the 1924 through 1931 tax years/the shortterm, postWorld War I stimulus nature of this first EIC is apparent when comparing the 1923 average EIC of $52 per taxable return to the significantly lower average EICs for the post1923 tax years.

THE SECOND EIC (1934 THROUGH 1943)

A new, revised EIC was made available for the 1934 through 1943 tax years. This EIC bore little resemblance to its predecessor. This “credit” resulted in a reduction of taxable amounts subject to the normal tax (as opposed to the surtax). Brozovsky & Cataldo [1994, pp. 173174] state that

(u)nlike the preceding EIC, this “credit” on earned income was comparable to what is today referred to as a “deduction”. This “credit” on earned income was restricted in amount to 10% of the first $14,000 of “net income” for a maximum deduction of $1,400 for single or married taxpayers.

This EIC evolved into a variable standard deduction (1944 through 1963), a semivariable standard deduction (1964 through 1976), and currently takes the form of an inflationindexed, fixed standard deduction (1977 through present), subject to phaseout and eventual elimination for the post1990 tax years [Brozovsky & Cataldo, 1994, pp. 168169].
Events Immediately Preceding the Post1974 EIC

What later came to be known as the New Frontier, the War on Poverty, and the Great Society, represented the culmination of considerable discussion regarding the possibility of a negative income tax or a guaranteed minimum income or (basic subsistence) allowance as a means of perfecting or completing Roosevelt’s New Deal. As described by Hildebrand [1967, pp. 12],

(w)ithin the first six months of 1966 alone, three important official documents . . . appeared, all of which commend the goal of universal guaranteed minimum income . ..
War on Poverty (1964) efforts originated and extended through the Kennedy (19611963), Johnson (19631969), and Nixon (19691975) administrations. These efforts coincided with/were overshadowed by the assassinations of President John F. Kennedy (1963), Robert F. Kennedy (1968), and Martin Luther King (1968), passage of the Civil Rights Act (1964), the “Americanization” (1965) of the Vietnam War (19571975) [Zarefsky, 1986], and Nixon’s Family Assistance Program (FAP) [Burke and Burke, 1974].
Aaron (1978), in his forward,

. . . argues that the Great society did not fall of its own weight, but rather was eclipsed by external events — the war in Vietnam, the dissolution of the civil rights

66 The Accounting Historians Journal, June 1995
coalition, and the political defalcations of the Nixon Administration.
The initial, post1974 EIC closely resembled the $400 credit or “workfare” plan (1972) proposed by, then, Chairman of the Senate Finance Committee, Russel B. Long (Democrat Louisiana). Senator Long’s father, Senator Huey Long (Democrat Louisiana), had been a proponent of the “Share the Wealth Society” (1934) prior to his assassination in September, 1935 [Burke and Burke, 1974, p. 151].
Russel Long had been an outspoken critic of Nixon’s FAP, for its failure to provide adequate work incentives. The FAP was rejected by the Senate Finance Committee (1972) in favor of Chairman Long’s “workfare” plan.

THE POST1974 EIC (1975 FORWARD)

The Tax Reduction Act of 1975 provided the foundation upon which our current EIC is based. This post1974 EIC was a delayed outgrowth of the extended period of political liberalism throughout the 1960s and 1970s.
The post1974 EIC differed significantly from the first EIC (1923 through 1931). These differences are restated and summarized in Table 2.

TABLE 2
A Comparison of the First EIC (1923 through 1931) and the Post1974 EIC

First EIC

Never refundable.
Never restricted or based only on “earned” income.
Always available to single taxpayers or taxpayers without dependents.
Introduced at a time when the same rate schedules applied to single and married taxpayers.
Predated the Social Security system.

Post1974 EIC

Refundable, and beginning July 1, 1979, as a payroll advance (AE1C).
Alwavs based onlv on “earned”
First available to lowincome taxpayers without dependents for post1993 tax years (RRA93).
Introduced after separate, post1970, rate schedules were developed for single and married taxpayers.
Postdated the Social Security svstem.

The evolution of the contemporary EIC is divided into three distinct phases: Phase I initial implementation (1975 through 1990), Phase II expansion into additional, separable components for health care costs and newborn children6 (1991 through 1993), and Phase III merger of previously developed separable ElCbased components and the inclusion of lowincome taxpayers without dependents (1994 forward).

Eligibility for the EIC, as modified throughout Phases I and II, included criterion for dependent (1) relationship, (2) residency, and (3) age limitations. The similarities of these early requirements to the dependencybased AFDCFC welfare program has been mentioned previously and is apparent. The Phase III period resulted in the elimination of qualified dependency exemption requirements.

Phase I (1975 Through 1990)

The EIC, in its recent historical form, was first made available for the 1975 tax year and only to lowincome workers who maintained a household for dependent children, for whom they were able to claim an exemption. Designed to reduce the impact of Social Security taxes and encourage the pursuit of employment by lowincome individuals, El includes wages/salaries (and related compensation) and net earnings (losses) from selfemployment (Schedules C, F, and, potentially, Schedule E income). This EIC was the first refundable credit. It was treated as a tax payment and, therefore, subject to refund.

The Tax Reform Act of 1976 provided for a continuation of the EIC for the 1976 and 1977 tax years, while liberalizing the requirements for claimants. The Tax Reduction and Simplification Act of 1977 extended the ETC through the 1978 tax year. Finally, the Revenue Act of 1978 increased the amount of the EIC and made it a permanent component of the tax law.
The Tax Reform Act of 1986 provided for inflationindexation of phaseout amounts, maximum adjusted gross income (AGI) limitations, and the maximum available EIC amount. (These amounts are adjusted, annually, according to changes in the Consumer Price Indexbased measures between August 31 of the preceding year and August 31 of the current year). For the 1987 tax year, the EIC phaseout at the 10% rate began at $6,925 ($9,000 as indexed for post1987 tax years) of the greater of El or AGI (see Table III). No EIC was available for taxpayers with

6Taken in lieu of the child and dependent care credit (Form 2441).

AGIs or El of $15,432 (or greater) for the 1987 tax year ($17,000 as indexed for post1987 tax years).
Indexation from 1987 base amounts resulted in increases of EIC phaseout amounts and ElCbased maximum AGIs of approximately 9%, 4%, 5%, 5%, 5%, and 3% (all rounded) for the 1988, 1989, 1990, 1991, 1992, and 1993 tax years, respectively. The relationship of EIC phaseout amounts to ElCbased maximum AGIs remained relatively stable throughout this post1987 through pre1994 period, with the EIC phaseout amounts approximating 53% of the ElCbased maximum AGIs.

EXHBIT 2
Contemporary EARNED INCOME CREDIT — Phase I 1975 through 1990 (in nominal dollars)

s?;
88’\89’v90’
$1,000 $900 $800 $700 $600 $500 $400 $300 $200 $100

Amount of EARNED INCOME CREDIT
ZZAV

$0 J2.5K $5K $7.5K $iOK $12.5K

$15K $17.5K $20.264K

TABLE 3
Evolution of the Post1974
EARNED INCOME CREDIT (EIC)
(1975 through 1994 and 1995 & 1996
before InflationAdjustment)

Phase Max Avg
Tax Qualify Credit Out “Flat” Range El or Max Rfnd
Year(s) Dpndnts Rate Rate Begin End AGI EIC EIC
Phase I: Initial Implementation of the Post1974 EIC
197578 >0 10.00% 10.000% $4,000 $ 4,000 $ 8,000 $400 $203
197984 >0 10.00% 12.500% $5,000 $ 6,000 $10,000 $500 $270
198586 >0 11.00% 12.222% $5,000 $ 6,500 $11,000 $550 $317
1987 >0 14.00% 10.000% $6,080 $ 6,920 $15,432 $851 $452
1988 >0 14.00% 10.000% $6,240 $ 9,840 $18,576 $874 $540
1989 >0 14.00% 10.000% $6,500 $10,240 $19,340 $910 $560
1990 >0 14.00% 10.000% $6,810 $10,734 $20,264 $953 $605
Phase II: Supplemental Health Care and Newborn Components Added
$21,250 $1,192 A
$1,235 A
” $428 A
$357 A
$22,370 $1,324
$1,384
$451
$376
$23,050 $1,434
$1,511
” $465
$388
$7,520 $11,840
$7,750 $12,200
$7,140 $11,250
1991 1 16.70% 11.930%
2 17.30% 12, ,360%
Health 6.00% 4. .285%
Newborn 5.00% 3, ,570%
1992 1 17.60% 12. .570%
2 18.40% 13. ,140%
Health 6.00% 4. .285%
Newborn 5.00% 3. ,570%
1993 1 18.50% 13. .210%
2 19.50% 13, ,930%
Health 6.00% 4. .285%
Newborn 5.00% 3.570%

Phase III: Supplemental Components Combined with the Basic Credit & Inclusion of LowIncome Taxpayers Without Dependents

1994 0 7.65% 7.650% $4,000 $5,000 $9,000 $306
1 26.30% 15.980% $7,750 $11,000 $23,753 $2,038
2 30.00% 17.680% $8,425 $25,300 $2,528
1995 0 7.65% 7.650% $4,168 $5,210 $9,378 $319 B
1 34.00% 15.980% $6,252 $11,462 $24,764 $2,126 B
2 36.00% 20.222% $8,779 $27,090 $3,160 B
1996 0 7.65% 7.650% $4,343 $5,429 59,772 $332 B
1 34.00% 15.980% $6,515 $11,943 $25,804 $2,215 B
2 40.00% 21.060% $9,148 S29,318 S.3,659 B
Note A: $793. Not specified as a separate measure for each EIC component. Note B: Projected at an annual inflation rate of 4.2% each for 1995 & 1996.

Exhibit 2 was developed from the information contained in Table 3 [SOI, 19771993] and provides a graphic representation of the contemporary EIC during the 1975 through 1990 (Phase I) tax years [SOI, 19771992], under the simplifying assumption that the taxpayer’s El is equivalent to their AGI. For post1986 tax years, the slope of the phasein (i.e., credit percentage) of the EIC increased, and, with the exception of the 1987 tax year, the flat range, over which the maximum EIC was available, increased in breadth.
However, it is generally acknowledged that during this period, the “working poor” lost ground with respect to the inflationindexed value of exemptions. As pointed out by Sommerfeld, et al., p. 56,

. .. (I)n the mid1970s a family of four began to owe an income tax only after it earned about 1820% over (emphasis added) the poverty line. By the mid1980s, the same household was required to pay an income tax after earning an income level significantly below (emphasis added) the poverty level. Indexation for inflation and increases in the earned income credit, the standard deduction, and the personal exemption allowed the working poor to be exempt from income tax until their income exceeded the poverty level. These changes are expected to keep taxpayers in the same position relative to the poverty line thereafter by indexing the standard deduction and exemption amount.

Table 3 summarizes key components of the recent historical EIC for the entire 19 year, post1974 period (i.e., 1975 through 1993) under review. The following formulas, as they relate to the data summarized in Table 3, provide the basis for the calculation of the EIC for all three phases (including the separable/ supplemental health care and newbornbased EICs, not graphically depicted, but available during the Phase II period of 1991 through 1993), where the EIC is the lesser of (la) or (lb).

(la) Credit Rate X Min { El, Flat RangeBEG1N }
(lb) Maximum EIC
[ PhaseOut Rate X ( Max { AGI, El )
Flat RangeEND ) ]

The flat range is that range of El over which the EIC is maximized. The beginning of the flat range represents that El level where the maximum EIC is first achieved. The end of the flat range represents the highest possible El level where the maximum EIC can be generated. Beyond this ceiling, the EIC is reduced at the phaseout rate (see Table 3).
Phase II (1991 Through 1993)

For the 1991 [SOI, 1993] through 1993 tax years, additional, separable EIC components were made available for supplemental health care/insurance expenses and newborn/young child dependents. Married, surviving spouse, or head of household filing status (i.e., at least one qualifying child in the household) remained a requirement for qualification for the (1) basic, (2) supplemental health insurance, and (3) supplemental young child EICs. A new Form EIC was developed and used during this period. However, the schedules, rules, and tables associated with these separate EIC components were very difficult for the average taxpayer to understand and were eliminated/merged back to a single basic EIC form for the post1993 period (i.e., Phase III).

Throughout the history of the Federal Insurance Contributions Act/SelfEmployed Contributions Act (FICA/SECA) based Social Security and (later) Medicare tax, beginning with the 1937/1951 tax years, the amount to which an employee/employer or selfemployed taxpayer was subject had been limited by a ceiling or “wage base.” This wage base was adjusted through intermittent statutory or (currently) automatic inflationindexed increases, but remained regressive. The EIC continued to partially or fully offset the regressive effects of Social Security and Medicare and promote tax progressivity.

Separate, higher wage bases were established for FICA and SECA Medicare components at 1.45% (each for employer and employee) and 2.9% (for selfemployed taxpayers), respectively, during the 1991 through 1993 tax years. These increased wage bases for the medicare components of FICA/SECA coincided with the establishment and maintenance of separate supplemental health care and newborn EIC components.

Phase III (1994 Forward)

For post1993 tax years, congruent with the first EIC (1923 through 1931), lowincome taxpayers without dependents are included as potential beneficiaries of the EIC.7 With the post
7 The EIC was originally denied to persons without children to avoid benefits to (1) students, (2) retired persons (to avoid duplication of the benefits already

1993 move toward a broader definition of the “working poor,” initially established EIC phasein percentages equal related phaseout percentages for lowincome taxpayers without dependents, and are equivalent to the rates used for FICA employer and employee contributions at 7.65% each (see Table 3).

EXHIBIT 3

Contemporary EARNED INCOME CREDIT —
Phases II & III 1991 through 1994 and 1995 & 1996 Projected*
Credit Rates (PhaseIn Percentages) & Flat Ranges* Earned Income Level

911 912 921 922 931 932 940 941 942 950 951 952 960 961 962
Tax Year — Min. Req’d No. of Qualified Dependents *1995 & 1996 Projections at 4.2% Annual Inflation Rates.
provided by the credit for the elderly (Schedule R)), and (3) parttime workers w:ith small amounts of El.

Exhibit 3 illustrates the recent historical trend for the credit rates and flat ranges associated with post1990 (i.e., Phases II and III) EICs. The generally upward “creeping” (or inflationindexed) movement of these flat ranges is apparent throughout the period.

The higher wage bases established for the Medicare component (i.e., 1.45% and 2.9% for employer/employee and selfemployed taxpayers, respectively) for the 1991 through 1993 (Phase II) period have been completely eliminated for post1993 tax years, creating a “flat” or proportional tax for this (previously regressive) component of FICA/SECA. The elimination of this ceiling coincides with the establishment of the no dependent EIC inclusion for post1993 tax years (see Exhibit 3).

CONTEMPORARY ISSUES

The difficulties associated with the distinction between earned and unearned income in the establishment of the first EIC (1923 through 1931) were considered, addressed, and circumvented by making the credit available Lo ail taxpayers and based on both earned and unearned income. These complications were not avoided under the more contemporary, post1974 EIC, which attempts to create a distinction between earned and unearned income. As a result, problems, other than those encountered in the early 1920s, have evolved.

First, in providing for the separate treatment of earned and unearned income, the post1974 EIC has included both wages and selfemployment income. Such amounts are often subject to manipulation, and, as O’Neil and Nelsestuen (1994) have determined, a significant portion of EIC benefits may be providing assistance to middleclass or even very wealthy taxpayers.8 Similar conclusions were drawn by Cataldo (1994) in his review of the projected trends and increased phasein credit percentages associated with the broader, post1993 ElCbased flat ranges.

For example, wealthy taxpayers, already involved in taxminimizing strategies and in control of closely held corporations, might legitimately provide/manipulate the cost of their own services (i.e., salaries) to El levels falling within their respective flat ranges, in efforts to maximize potentially refund

s O’Neil and Nelsestuen [1994J proposed implementation of a “cliffbased” wealth restriction measure, where the taxpayer’s taxable and nontaxable interest income and taxable dividend income might be used, in aggregate form, as a proxy for wealth and a means test for EIC benefits elimination.

able EICs. Similarly, middle (and high) income, selfemployed taxpayers in the process of startup or expansion, might manipulate income/expense items and even make postyearend elec

TABLE 4

Selected Descriptive Statistics from 1989 SOI Returns Taxpayers Benefitting from the EIC

EIC
Count (NT) for Amounts > $910
AGI Class N Int Div’s Sch E Sch C/F PALs AMT Mean Std Dev
AGIs Below One Dollar ($1):
AGI<$1 140 77 38 19 75 54 37 $496 $293 PCTTTLS 2% 23% 36% 13% 8% 50% 77% For All Returns (N= 140; N=280 for Sch C/F) in AGI<$ 1 Class Min $0 $0 $(7)M $(421)K Mean $90K $7K $(369)K $(6)K Max $1.9M $334K $507K $102K Std Dev $261K S35K $1M $43K AGIs Above Zero ($0): AGI<$1K 75 0 0 0 1 0 0 $113 $123 AGI<$2K 131 0 0 1 20 2 1 $227 $115 AGI<$3K 159 5 0 1 18 2 1 $341 $88 AGT<$4K 207 5 1 4 28 2 0 $467 $97 AGI<$5K 225 6 0 2 42 1 1 $604 $105 AGI<$6K 249 9 4 5 32 0 1 $725 $130 AGI<$7K 303 10 2 6 57 3 0 $839 $147 AGI<$8K 323 12 3 6 52 6 0 $860 $142 AGI $19,340, the EIC was unavailable (see “Max El or AGI” column of Table 3). M = Millions. K = Thousands.

tions (like the IRC 179 expense election) to maximize refundable EICs.

Table 4 summarizes selected descriptive statistics for all returns qualifying for the EIC for the 1989 tax year. It was developed from the 1989 SOI Public Use File [SOI, 1992]. The maximum available EIC for the 1989 tax year was $910 (see Table 3), therefore, this amount was selected as the appropriate breakpoint for developing the number of returns with taxable interest income (Int), dividends (Div’s), Schedule E rent and royalty income (Sch E), and selfemployment earnings from Schedules C and F (Sch CIV). Also provided are the frequencies of taxpayers with passive activity losses (PALs) and the alternative minimum tax (AMT), both of which are typically associated with middleto highincome taxpayers. These amounts are summarized separately, by AGI class, and illustrate the underlying rationale for the EIC wealth restrictionbased limitations advanced by O’Neil and Nelsestuen (1994) and the everincreasing tax planning opportunities for ElCmaximization, in the absence of such wealth restrictions, described by Cataldo (1994).

Of particular interest are the mean EICs available to highincome taxpayers in the zero or negative AGI class. Generally, these are highincome taxpayers. These taxpayers are receiving EICs in average amounts in excess of those available to taxpayers below (above) the $4,000 ($14,000) AGI classes.

Many of these highincome taxpayers, no doubt engaged in nonEICrelated tax planning efforts, are, by chance or design, earning additional aftertax, ElCbased returns of between 9% and 14% (i.e., $910 divided by the $6,500 through $10,240 flat range (1989), respectively) on potentially manipulated El amounts.

Second, participation ratesbased studies indicate that many (i.e., 14% to 25%) taxpayers, not otherwise required to file an income tax return, but qualifying for the EIC, may not be av/are of the ElCbased benefits foregone through their failure to file an income tax return [Scholz, 1994]. Furthermore, “…almost none of the recipients obtain the benefits incrementally during the course of the year.” The result is the “… almost total ineffectiveness of the advance payment option…” [Yin and Forman, 1993, pp. 951 & 953].
The failure to achieve very high AEIC participation rates, through public awareness of its availability, defers any progress (and administrative cost reductions) associated with the elimination of alternative welfare delivery systems. Though the Internal Revenue Service has expanded efforts to publicize the EIC program, they tend not to promote the AEIC option to minimize noncompliance [Yin and Forman, pp. 954956].

Third, “(s)elfemployed individuals might declare work when none had taken place in order to receive a higher value of credit” [Steuerle, 1993]. Generally, this argument is consistent with the overall trend of increasing EIC credit rate(s) (see Table 3), and their potential to exceed the combined federal and (net) selfemployment tax rate. RRA93 aggravates this problem.

For example, taxpayers, with two dependent children for the 1996 tax year, might generate an EIC of 40%, while subject to a marginal federal income tax rate of only 15% (or even 0%) plus the selfemployment tax rate of less than 15.3% (after adjustments), for a net “profit” of 10% (i.e., 40% minus approximately 30%) (see Table 3).
Furthermore, separate and apart from selfemployment earnings overreporting, “… nearly one in three of those receiving the credit in 1990 was ineligible” [Kirchheimer, 1993]. Again, the incentive for false reporting evolves from the fact that the EIC benefits or phasein credit rate frequently exceeds the taxpayer’s marginal tax rate (see Table 3). Additional overpayments arise from income variability, multiple employers, and those cases where both married taxpayers work and elect the AEIC option [Holt, 1994, pp. 760762].
Finally, the politically popular “workfare” or work incentive feature of the EIC is questionable. Moore [1993, p. 106] reminds us that

(f)or people already working, the EITC will increase their overall income level in each and every range. Conventional analysis and a wide range of studies indicate the “income effect” alone tends to discourage work, since a family can attain any particular level of income with less work than in the absence of the EITC payment.
(T)he “substitution” effect of the EITC varies with each of the ranges. In the phasein range, it encourages

‘Yin and Forman [1993, p. 954], noting the findings of a U.S. General Accounting Office report [GAO/GGD92132, 1992], explain that the IRS, in an effort to increase full compliance with the EIC, awarded the credit to 600,000 taxpayers for the 1991 tax year. These taxpayers failed to claim, but appeared to qualify for the credit. The IRS subsequently found that 270,000 (45%) of the awards were incorrect.

work, because the reward for an additional hour of work has increased. In the (flat) range, the substitution effect does not come into play because the credit remains at the maximum level as income is increased. Finally, in the phaseout range, the substitution effect provides a disincentive (emphasis added) to work, since the worker now finds that the EITC is reduced as income increases. Thus, the worker’s effective wage has been reduced.

In the (flat) range, only the income effect applies, and tends to discourage work.

The everincreasing flat range (see Exhibits 2 and 3) results only in the “income effect,” and work efforts may be discouraged once achieved/surpassed. The phaseout range provides for a work disincentive. The family is richer as a result of the EIC, and additional work is less rewarding [Moore, 1993, p. 106]. These effects exacerbate the effective marginal tax rates of EIC recipients to a minimum of 34.5% (i.e., the phaseout rate of 19.5% plus the marginal tax rate 15% for the 1993 tax year).

SUMMARY

The EIC phasein rates have increased from 10%, for the 1975 through 1978 period, to a planned rate of as much as 40% for the 1996 tax year. Like the first EIC, established in the early 1920s, the postRRA93 EIC is now available to taxpayers without dependents. Unlike the first EIC, expansion of the post1974 EIC has taken the form of a “workfare”based variation of guaranteed income or negative income tax. Increases in coverage and phasein credit percentages have taken place prior to the resolution of problems with compliance and delivery, preventing the EIC from replacing alternative welfare systems.

Without a wealthbased restriction, the EIC cannot progress to replace traditional welfare systems possessing such features. The higher, postRRA93 EIC benefits and/or phasein credit rates may provide unintended beneficiaries with everincreasing economic incentives for the “positioning” of their earned income within broadening flat ranges. Retaining the politically popular “work incentive” feature of the EIC, which is incompatible with existing welfare systems, is the very vehicle through which such manipulation becomes possible.
Those responsible for the first EIC (19231931) were concerned with the infeasibility of successfully separating earned and unearned income components. They anticipated and completely avoided this latter issue, appearing to have left us with a historical lesson unlearned.10 And this politically popular “work incentive” feature may not even be supported by empirical evidence when substitution and income effects are considered.

If these issues remain unresolved, the next logical step in the evolution of the EIC may remain the elimination of the “work incentive” component. However, while elimination of the need for any distinction between earned and unearned income would resolve problems associated with ElCmaximizing tax planning and provide for the elimination of concurrent transfer payment systems, a wealthbased meanstest would still be necessary.

REFERENCES

Aaron, Henry J., Politics and the Professors: The Great Society in Perspective,
Washington, D.C.: The Brookings Institution (1978). Ammer, Christine and Dean S., Dictionary of Business and Economics, New
York, N.Y.: The Free Press (1977). Brozovsky, John and Cataldo, A.J., “A Historical Analysis of the ‘Marriage Tax
Penalty’,” Accounting Historians Journal, Vol. 21, No. 1 (June 1994): 168
193. Burke, Vincent J. and Vee, Nixon’s Good Deed: Welfare Reform, New York, N.Y.:
Columbia University Press (1974).
Campbell, Colin D. and Peirce, William L., The Earned Income Credit, Washington, D.C.: American Enterprise Institute (1980). Cataldo, A.J., “Targeting Earned Income Credit “Flat Ranges”: SubsidyBased
Tax Planning for the Working Wealthy and Middle Class,” Unpublished
Working Paper (1994). Hildebrand, George H., Poverty, Income Maintenance, and. the Negative Income
Tax, Ithaca, N.Y.: New York State School of Industrial and Labor Relations
(1967). Hoffman, Saul D. and Seidman, Laurence S., The Earned Income Tax Credit:
Antipoverty Effectiveness and Labor Market Effects, Kalamazoo, Mi: W.E.
Upjohn Institute for Employment Research, (1990). Holt, Stephen D., “Effect of 1993 Budget Act on the Advance Payment Option of
the Earned Income Credit,” Tax Notes, Vol. 62, No. 6 (February 7, 1994):
759763.
10 Some of these issues have, very recently, been addressed and/or are being addressed in the form of formal executive and legislative proposals. For example, the U.S. Treasury has proposed EIC antifraud measures, targeted at selfemployed taxpayers (see Westlaw, 1995 DTR 116 dll, June 16, 1995, and Westlaw, 1995 DTR 115 d38, June 15, 1995).
Also, Public Law 1047 (H.R. 831) places restrictions on EIC eligibility for post1995 tax years. Taxpayers with “disqualified income” in excess of $2,350 will no longer qualify for the EIC. Disqualified income includes taxable and taxexempt interest, dividends, and net rent and royalty income.
Kirchheimer, Barbara, “EITC Gets High Marks, But Flaws Need Correcting,”
Tax Notes, Vol. 61, No. 5 (November 1993): 521 523. KixMiller & Baar, 1925 United States Income and War Tax Guide, Chicago, IL:
Commerce Clearing House (1924). I.ipman, Francine J. and Williamson, James E., “Recent Proposals to Redesign
the EITC: A Reply to an Economist’s Response,” Tax Notes, Vol. 62, No. 9
(February 28, 1994): 11751180. Mclntyre, Michael J., “Rosen’s Marriage Tax Compulations: What Do They
Mean?”, National Tax Journal, Vol. XLI, No. 2 (June 1988): 257258. Moore, Robert L., “Recent Proposals To Redesign The EITC: An Economist’s
Response,” Tax Notes, Vol. 60, No. 1 (July 1993): 105108. O’Neil, Chcrie J. and Nelsestuen, Linda B., “The Earned Income Credit: The
Need for a Wealth Restriction for Eligibility Determination,” Tax Notes, Vol.
63, No. 9 (May 1994): 1189 1201. Pechman, Joseph A., Federal Tax Policy, Fifth Edition, Washington, D.C.: The
Brookings Institution (1987). Polinsky, Alexander, “Tax Bill’s Stiff Wedlock Penalty Tests Limits of ‘Family
Values’,” Tax Notes, Vol. 60, No. 1 (July 5, 1993): 1112. Report by the Auditor General of California, The Department Of Social Services
Does Not Use Information From. Its Audits To Adjust Rates For Group Homes
That Receive Funds From The Aid To Families With Dependent Children
Foster Care Program, P540, Sacramento, CA (August 1986). Rosen, Harvey S., “Thinking About the Tax Consequences of Marriage,” National
Tax Journal, Vol. XLI, No. 2 (June 1988): 259260.
, “The Marriage Tax Is Down But Not Out,” National Tax Journal,
Vol XL, No. 4 (December 1987): 567575. Scholz, John K., “The Earned Income Tax Credit: Participation, Compliance,
and Anlipoverty Effectiveness,” National Tax Journal, Vol. XLVJI, No. 1
(March 1994): 6387. Sommerfeld, Ray M., Madeo, Silvia A., Anderson, Kenneth E., and Jackson,
Betty R., Concepts of Taxation, 1994 Edition, Orlando, FL: Harcourl Brace &
Company (1993). Statistics of Income (SOI) for 1931 and 1975 through 1991 (1933 & 19771993,
respectively). United States Treasury Department, Bureau of the Internal
Revenue (Washington, D.C.): 1933 & 19771993, respectively. Statistics of Income (SOI) 1989 Public Use Sample, United States Treasury Department, Bureau of the Internal Revenue (Washington, D.C.): 1992. Sieucrle, Gene, “Economic Perspective: Measuring the Impact of Job Subsidies:
Is There a Double Standard?” Tax Notes, Vol. 61, No. 7 (November 15,
1993): 865866. Yin, George K. and Forman, Jonathan B., “Redesigning the Earned Income Tax
Credit Program Lo Provide More Effective Assistance for the Working Poor,”
Tax Notes, Vol. 59, No. 7 (May 17, 1993): 951960. Zarefsky, David, President Johnson’s War on Poverty, University, Alabama: The
University of Alabama Press (1986).