18 March 2010

Limits of the EITC

The Earned Income Tax Credit (EITC) reduces the tax burden of working class people with earned income, a benefit which as phased out as taxpayers approach middle class incomes.

It has received considerable criticism, however, and rightly so.

The provision is a complex one targeted at people who rarely have access to professional tax advisors. Indeed, it is the single most complex part of the tax code that applies to working class and middle class taxpayers who are not self-employed. As a result, the EITC is the single largest source of audits in the tax code. EITC mistakes are so common that working class taxpayers are audited more often than wealthier taxpayers much more likely to be intentionally violating the law.

Equally problematic, the effective marginal tax rates that result from the phase out of the EITC are the highest faced by taxpayers of any income. Also, these same people often lose many means tested benefits at the same time that they are facing an EITC benefit. As a result, taxpayers in the EITC phaseout income range often face effective marginal tax rates approaching 100% when means tested benefit phase outs are considered. High marginal tax rates for working class people is hardly a good way to help them escape poverty.

A new law review article further explores the problems with the EITC:

[T]he EITC reduces poverty only modestly, and even the maximum credit falls short of closing the gap between low-wage earnings and poverty. At the same time, gaps in other social welfare programs leave low-income workers vulnerable to the job disruptions that characterize low-wage work - when the EITC provides no assistance at all.


Getting The Working Poor Out Of The Tax System

Tax policy directed at the working poor needs to be minimally bureaucratic and provide good incentives for participants. As a general rule, those in poverty should owe no federal income or payroll taxes of any kind, and should not need to file a return to claim this benefit.

The resources of the IRS are not best spent chasing after working poor people who have trouble accurately calculating their entitlement to the EITC, waiters who don't fully report their tip income, and gray market part-time hairdressers and baby sitters. Its resources are better spent pursuing big businesses that take elaborate efforts to twist federal tax law to permit tax breaks not contemplated by the law, and affluent self-employed people and property owners who fail to report significant taxable income.

The jobs bill signed by President Obama today, which creates a payroll tax holiday for certain unemployed workers hired by businesses, is a step in the right direction.

Categorical Programs Are Often Preferrable To Means Testing

Social welfare programs, when possible, should have a categorical eligiblity test, like Medicare and public K-12 education, rather than a means test, like Medicaid and cash welfare programs.

Health Care

Once you are committed to the idea of universal health care, as our nation looks likely to be in a matter of weeks, the money spent administering means tested programs is no longer a matter of cost control; it simply a matter of shifting cost from one financing approach to another. This greatly reduces the value involved in high administrative costs.

School Breakfasts and Lunches

For example, the are many schools where the vast majority of students are poor. And, a large share of all poor students attend schools where the vast majority of students are poor, because neighorhoods tend to be made up of people with similar incomes. In schools where most students are poor, it may make more sense to make school breakfasts and lunch available for free to all students, than to try to keep track of free and reduced school nutrition program eligiblity and to collect lunch money from students who haven't applied to the program or are affluent enough not to qualify. The administrative cost savings involved in making entire schools eligible may offset the costs involved in providing benefits to more students. And, many of the new beneficiaries in these schools would have been eligible for a means tested program anyway, if they had applied, but have parents who never got their act together bureacratically well enough to receive those benefits.

Public Defenders and Bankruptcy

Programs that primarily benefit the working poor, even if they marginally benefit others, like the public defender system, and the bankruptcy system, should likewise be structured to avoid elaborate financial paperwork. Rather than basing access to these services on means, the nature of the benefit itself could be structured to favor the intended beneficiary.

For example, a universal public defender system, in which anyone can have a lawyer appointed for them in a criminal case regardless of means, is less likely to be abused if people who are convicted have a duty to reimburse the state for their defense if they have the means to do so. Perhaps 80% or more of people in the system already receive public defenders because they are indigent already, and many of those who don't are still far from affluent. Also, providing financial assistance to people who are charged with crimes and then acquitted would not be very expensive (since full acquittals at trial are quite rare as a percentage of all criminal cases filed, making about around 1% of the total) and have ample moral justification of a categorical benefit.

Similarly, a bankruptcy system that imposes a three year garnishment after the case is otherwise closed, for the benefit of creditors whose debts were discharged, would discourage high income people from filing for bankrputcy without using means as a gatekeeper to bankruptcy court relief.

Higher Education

Means tests may be unavoidable at times. But, they should be used sparingly and with an eye towards the incentives that they create.

For example, means testing may be a worthwhile part of the solution for higher education, where a large share of students are from affluent families, most of the rest of the students are from bureacratically competent middle class families (and most students are bureacratically competent), the benefit is relatively large in absolute dollar value, and the impact that financial assistance has on college attendance of students with low means is great. Yet, even in higher education, categorical screens based on factors like first generation college student status and academic merit are also very useful.

Rather than indiscriminantly funding all in state students, or all students who attend public institutions of higher education, it may make more sense to bring regular tuition more in line with the costs associated with providing higher education, and then to be more selective about who receives limited public assistance in getting a college education.

On one hand, the system does neither beneficiaries nor the public purse, few favors by providing financial assistance to students entering academic programs that they are likely to fail because they are ill prepared.

On the other hand, even a quite generous means test can end a subsidy for highly affluent upper middle class students, and in the process make more of the scarce funds available to fund higher education available to those who would likely forego higher education without financial assistance.


Our unemployment system is another major safety net that is deeply broken.

While not strictly speaking a means tested program, it devotes considerable resources to denying benefits to applicants, because they are unemployed for the wrong reasons, because they set their standards too high in their job hunt or don't devote enough effort to a job hunt, or because they get some marginal work that leaves them merely underemployed rather than unemployed.

For the long term umemployed, and the unemployed who didn't manage to build up a substantial emergency fund before losing work, credit cards are often a more meaningful form of unemployment and underemployment assistance than unemployment insurance.

The COBRA subsidies we've seen in response to the financial crisis are one sensible, catagorical response. Another policy option used in many countries is a system of severance payments that are due on termination of employment in the vast majority of cases that decouple the likely financial burden associated with a discharge from employment from the process of finding a new job.

No comments: