Over at Slow Boring, Simon Bazelon and David Shor have a piece arguing that more sharply means-testing the CTC would make the program more popular without creating any administrative problems that the current means-tested CTC doesn’t already have. This is incorrect. A sharper means test would massively increase overpayment problems, which would then make the program less popular.

The CTC means test has two main administrative problems:

  1. It makes applying for the program a lot harder especially for the poor.
  2. It causes some families to receive overpayments that are then clawed back by surprise tax bills in the subsequent year.

Bazelon and Shor’s mistake is that they think the CTC means test only has problem number one. And since making the means-test sharper doesn’t make the program application any harder than it already is, they conclude that there is no downside to means-testing it even more.

But actually the CTC means test has problem number two as well. And making the means-test sharper does make that problem worse and making that problem worse does generate backlash that makes the program less popular.

The Overpayment Problem

The current CTC means test looks like this.

As a family’s modified adjusted gross income exceeds these various thresholds in the graph, the amount of the CTC benefit declines until it reaches zero.

This seems simple enough except that you don’t actually know what a family’s income is going to be until the end of the year. So to make payments to them during the year, you have to base those payments on their estimated income. Currently, the IRS does this by assuming that a family’s income in the current year will be the same as their income in the prior year.

But family income changes year to year, meaning that a family’s income could make them eligible for the full benefit last year but then increase so much that they are no longer eligible for it this year. Because the IRS bases their monthly payments on their income from last year while basing their eligibility on their income from this year, they end up with CTC overpayments, which are clawed back via a surprise tax bill in the subsequent year.

Right now, the means test thresholds are set so high that the overpayment problem will only affect a small percentage of people and those people will be high income and thus more able to manage a surprise tax bill. If you lower those thresholds, the overpayment problem will affect a much larger percentage of people and those people will be middle income and much less able to manage a surprise tax bill.

Dumping surprise tax bills on middle-income families is absolutely not a way to make a program more popular. We know this because the United Kingdom and Australia both tried this kind of means-testing when designing their child benefits and it generated enormous backlash that forced the governments to change the benefit designs.

From the United Kingdom experience:

From 2003, there were substantial and very visible problems with overpayments, or more precisely visible problems in terms of the impact on claimants of having to repay overpayments. The number of overpayments, and their costs, were much higher than predicted. In the first year (2003/04), about one-third of all tax credits awards paid – nearly 1.9 million awards – were overpaid, at a cost of nearly £2 billion (House of Commons Treasury Select Committee, 2006). Media coverage reported confusion, hardship and debt as the government sought to recover overpayments.

From the Australia experience:

Controversy about the new system started towards the end of the 2000–2001 financial year, when the time first approached for reconciling entitlements. In 1999–2000, there had been about 51,800 families with debts at the end of the year. In contrast, after the introduction of the new system, an estimated 670,000 families were overpaid – more than 12 times as many as the previous year.

Beware of Pollsters

In addition to being wrong in this specific case, Bazelon and Shor have provided a useful lesson in the limitations of polling gurus. Here they mix a literal half-understanding of the policy problem with polling to reach exactly the wrong conclusion about how to design a popular policy. It’s not that they think that surprise tax bills for middle-income families are popular. It’s that they don’t even know that this problem exists and certainly are not aware of what happened in other countries when they tried to do this.

Pollsters who are clueless about the administrative problems of a policy design surveying random people who are also clueless about the administrative problems of a policy design is one way to do politics. But I wouldn’t recommend it.