====== FULL TEXT ======

Dear Sens. Moynihan and Packwood and
Reps. Rostenkowski and Archer:

In recent years, one of the most profound changes in the American approach to low-income assistance has been the attempt to subsidize earnings rather than leisure, to encourage work over welfare, and to help moderate-income workers with children -- the group with the largest tax increases in the post-World War II era. In different ways, you and I, as well as many others with widely varying philosophies, have been involved in this often bipartisan process for some time, especially through enactment, reform, and changes in the earned income tax credit (EITC) in 1986, 1990, and now most likely, in 1993 -- expansions that have taken place while welfare cash transfers per family have been allowed to erode.

I consider the attempt to reorient the social transfer system toward work and inclusion, as opposed to separation and exclusion, to be one of the most important challenges facing modern government. I also recognize that history is likely to judge this effort as being merely in its infancy. In the United States, however, the movement toward a more mature adolescence is threatened by a lack of attention to administrative issues. Bluntly stated, the IRS cannot enforce the EITC as it is currently designed, much less as it would be expanded in the Senate and House tax bills. Unless this enforcement problem is corrected, the attempt to subsidize work could be set back years as one abuse story after another begins to hit the press. I urge you, therefore, to give attention to this issue now before it is too late.

Here is the story already, as told to me by one IRS official. You can easily find confirmation by talking to officials in the field offices of the IRS. More and more, taxpayers are beginning to walk in the doors of the IRS, file returns, and claim credit for work that cannot be verified. IRS suspicions are aroused by a variety of factors. The credit recipient often seems to know how much income to declare to receive the maximum EITC. In some cases, the recipient fails to declare any deductions for earnings from self-employment. A person will declare $7,500 from odd painting jobs, for instance, but will not declare $2,000 for paint supplies because the deduction would effectively reduce the EITC due.

Since the EITC has been around since 1975, why is this only now an issue? It was in the 1990 budget agreement that the EITC rate of credit was allowed to rise above 15.3 percent. In 1991, the EITC rate of credit for a household with two children was increased to 16.7 percent, with further scheduled rises to 25 percent by 1994. Once the rate of EITC credit exceeded the 15.3-percent Social Security tax rate, more subsidy was owed to the taxpayer than the taxpayer usually owed the government in taxes.

The IRS has never been able to enforce the tax laws well for individuals who are self-employed and keep their own books. Statistics on the "subterranean" economy reveal that a very large percentage of all noncompliance lies with this group. Essentially, there are few checks that an IRS examiner can make -- no one else, such as an employer, keeps a separate set of books -- and the amount of tax to be collected is small relative to the cost of time to the IRS and, ultimately, the cost of the enforcement effort that must be paid for by other taxpayers.

When the rate of subsidy for work begins to exceed the tax rate, incentives are reversed. Rather than a "subterranean" economy where the hiding of income reduces taxes, society finds itself with a "superterranean economy" where the overreporting of income generates additional subsidies from the government. A couple of years ago, I began to warn that this problem will represent one of the major future compliance issues facing both the IRS and welfare agencies (see Tax Notes, May 6, 1991, p. 647, and May 13, 1991, p. 787). Now the IRS is finding that my warnings are coming true. This is only the beginning, moreover, as the incentive is positively related to the size of the net subsidy, or subsidy less tax. For 1992 returns with two children, the net rate of subsidy was still only about 3.1 percent of earnings, and the maximum net subsidy was $237. Assuming full implementation of either the Senate Finance Committee or House bills, the net subsidy rate for a household with two children would rise to about 24 percent and the maximum net subsidy to more than $2,000. Social Security benefits also would increase in the long run for some because of the additional Social Security taxes paid, although this is not likely to be an important factor.

Unfortunately, it is hard to get good figures on the extent of noncompliance with the EITC. Past IRS research numbers on noncompliance included large amounts attributed to a misunderstanding of "support" and "maintenance of household" tests -- tests that required that the taxpayer, not AFDC or the welfare system, be the principal source of support for dependent children. These requirements were changed recently and, in any case, have no relationship to the fairly new enforcement problem detailed here.

The issue, moreover, goes beyond illegal declaration of income never earned. Under current law, if one person decides to baby-sit for a second person's kids and the second person reciprocates for the first, then each can legitimately claim to have worked. If they pay each other $10,000, then both are entitled to EITCs.

Just so my rationale for making these objections is not called into question, I have worked for some time on trying to improve the tax status of low-income workers. I helped design some simplifications of the EITC back in the late 1970s, put the EITC back onto the table in recent years by initially proposing its indexing as part of my job as economic coordinator of Treasury's 1984-86 tax reform efforts, succeeded in getting legislative attention to the extraordinary decline in the value of the personal exemption (see, for instance, Sen. Patrick D. Moynihan, Family & Nation), correctly warned the White House (to no avail) that some of the EITC changes proposed in 1989 (and enacted in 1990) were going to be extraordinarily complex and unworkable, and, as an adviser, helped the National Commission on Children to reach bipartisan consensus on the merits of a child credit. The simple fact is that the EITC has become large enough that we must pay as much attention to its design and implementation as we do to other parts of the nation's complex, and sometimes unmanageable, transfer system.

There are no easy answers to this problem. I have suggested elsewhere limitations on earned income tax credits to the amount of taxes paid by a business, requirements for information reports, and exclusion of certain activities of self-employment from the credit. A recent article by George K. Yin and Jonathan Barry Forman (see, Tax Notes, May 17, 1993, p. 951) suggested limiting the EITC to the Social Security tax rate and supplementing it with child credits. We need also to give attention to the interaction of EITCs and jobs credits granted to employers. Given enough time to work out details, of course, such changes could be made revenue-neutral with regard to whatever amount eventually was to be spent. For instance, in any conference on the forthcoming tax bill, savings due to improved EITC compliance provisions could still be spent on the working poor. My strong advice, however, is not to try to dodge the problem: it is already growing, even under current law, and you soon could have a monster on your hands.

Gene Steuerle is a senior fellow at the Urban Institute and an economic consultant to Tax Notes.