Teaching the Tax Code

A two-to-three-minute explanation [of tax incentives] can have substantial effects on labor supply over the subsequent year.

In Teaching the Tax Code: Earnings Responses to an Experiment with EITC Recipients(NBER Working Paper No. 14836), authors Raj Chetty and Emmanuel Saez test whether providing information about the Earned Income Tax Credit (EITC) affects EITC recipients' labor supply and earnings decisions. The EITC is the largest cash transfer program for low-income families in the United States. One of its major goals is to increase labor supply and earnings among low-income working households. To achieve that goal, however, taxpayers must understand how the EITC program applies to them. Yet surveys indicate that very few eligible individuals know whether working more would increase or reduce their EITC amount. In addition, they get little feedback about how their behavior affects their EITC refund, because the refund typically is received months after they make their decisions about labor supply.

Chetty and Saez conducted a randomized experiment at 119 H&R Block offices in Chicago in 2007 which involved 43,000 EITC claimants. Half of the H&R Block clients were randomly selected to receive a two-minute explanation by their tax professional about how the EITC works. By tracking the subsequent earnings of the 43,000 clients, the authors studied how information on the structure of the EITC affected total earnings in the year after the intervention.

The authors examined how tax preparers differed in the impact of their advice on taxpayers. One group, accounting for about half of all tax preparers, induced their clients to increase their EITC refunds by choosing an earnings level closer to the peak of the EITC schedule. The other group appears to have encouraged their clients to work more, but not to seek the highest possible EITC benefit. The taxpayers who met with tax preparers from this group had insignificant changes in EITC amounts, but they did raise their earnings on average toward the EITC phase-out region. The effects were larger for the self-employed, but were also substantial among wage earners, suggesting that providing information led to real labor supply responses.

When compared with other policy instruments, providing information has large effects, the authors find. The tax preparers who encouraged taxpayers to maximize the value of their EITC benefits generated the same labor supply response as a 33 percent expansion of the EITC program, while the other preparers induced the same response as a 5 percent tax rate cut.

It may be surprising that a two-to-three-minute explanation can have substantial effects on labor supply over the subsequent year. The authors suggest that is the case because the session combines simple information with advice from an expert at precisely a time when individuals are thinking about taxes. Chetty and Saez show that providing information can be a powerful policy tool because perceptions can be modified at low cost.

The main limitation of this study is that the authors were not able to characterize the mechanisms through which information and advice affect behavior. The decentralized implementation of their experiment made it difficult to precisely define the "treatment" that was provided by each of the tax professionals. Moreover, they were unable to determine how the treatment affected each client's perceptions about the structure of the EITC.

-- Lester Picker