Every Monday, the National Bureau of Economic Research, a nonprofit organization made up of some of North America’s most respected economists, releases its latest batch of working papers. The papers aren’t peer-reviewed, so their conclusions are preliminary (and occasionally flat-out wrong). But they offer an early peek into some of the research that will shape economic thinking in the years ahead. Here are a few of this week’s most interesting papers:
Title: “Tax Cuts For Whom? Heterogeneous Effects of Income Tax Changes on Growth and Employment”
Authors: Owen M. Zidar
What he found: Tax cuts for lower-income groups provide a bigger economic boost than cuts that target the rich.
Why it matters: Measuring the economic impact of tax policies isn’t easy. Lawmakers take economic conditions into account when they pass tax cuts or increases, which can make it hard to distinguish between cause and effect. (If the economy improves after a tax cut, is that the result of the tax cut, or did Congress cut taxes because they expected the economy to improve?) Moreover, Congress often passes multiple reforms at once, so it’s hard to separate the impact of individual changes. In this paper, the author tries to overcome those challenges by looking at regional variations in who is actually affected by tax changes. Connecticut, for example, has a disproportionate number of high earners, so it is more affected than other states by tax cuts or increases that target the wealthy. He finds that tax cuts that affect the bottom 90 percent of earners provide a measurable economic stimulus, while tax cuts that affect the richest 10 percent have a “weak to negligible” impact.
Key quote: “These effects hold at both the state and federal levels, and are not confounded by changes in progressive spending, state trends, or prior economic conditions. The effects are larger in states with high unemployment rates and seem to come from increased durable consumption, investment, and labor force participation.”
Data he used: The National Bureau of Economic Research’s “tax simulator,” which is based on a sample of actual tax returns.
Title: “An Extrapolative Model of House Price Dynamics”
Authors: Edward L. Glaeser, Charles G. Nathanson
What they found: Volatility in home prices can be partly explained by buyers’ incorrect assumption that past price increases accurately reflect housing demand in their area.
Why it matters: In the mid-2000s, U.S. home prices soared too unprecedented levels, then crashed spectacularly, bringing the economy down with them. That bubble-like behavior is hard to explain with traditional economic models, which expect buyers and sellers to base their decisions on rational estimates of supply and demand. In this paper, the authors develop a model that helps explain the volatility in home prices. Buyers and sellers do try to make their decisions based on market fundamentals, but since they have no direct way of measuring supply and demand, they approximate it by extrapolating from recent sales. Taken in isolation, that approach is reasonably sound, but what buyers don’t take into account is that past buyers all made the same calculation. Repeated over and over again, these largely rational assumptions lead to irrational behavior in prices: Prices rise, so buyers infer demand is increasing, so they are willing to pay more, which leads the next buyer to think demand is rising even faster, and so on. Eventually the market becomes so far detached from fundamentals that prices are forced to adjust, leading to a bust.
Key quote: “One of the more interesting findings is that the bubble-like features of markets disappear when information is either too good or too bad. If buyers have highly accurate direct signals about the state of demand, then momentum, mean reversion and excess volatility disappear. But these features also disappear if buyers have access to relatively limited data on the number of past housing transactions. The most extreme fluctuations occur when buyers have relatively good data about past prices, but limited data on the underlying fundamentals.”
Data they used: Federal Housing Finance Administration house-price indices and median home price data from the 2000 Census for large U.S. metropolitan areas.
Title: “Consumer Bankruptcy and Financial Health”
Authors: Will Dobbie, Paul Goldsmith-Pinkham, Crystal Yang
What they found: Filing for bankruptcy protection helps financially struggling borrowers hold onto their assets and raise their credit scores and makes them more likely to own their homes.
Why it matters: Chapter 13 of the U.S. bankruptcy code allows individuals to reduce their debts while retaining many of their assets. (A separate provision, Chapter Seven, wipes out more debt but forces borrowers to give up many of their assets.) In theory, bankruptcy is meant to give borrowers a chance to get back on their feet while giving lenders a chance to recoup at least part of what they’re owed. But measuring the impact of bankruptcy on borrowers is difficult because, by definition, people who file for bankruptcy are in financial trouble; even in a best-case scenario, they’re likely to end up worse off than people who never got into trouble in the first place. In this paper, the authors take advantage of the fact that bankruptcy cases are assigned to judges at random. Since some judges are more lenient than others, the authors are able to study the impact of bankruptcy protection on people who are in similar financial circumstances. They find that in the five years after a bankruptcy filing, people who are granted protection have a 13.2 percentage point greater probability of being homeowners and have a 14.9-point higher credit score, on average, than those not allowed to enter bankruptcy. The amount of debt they have in collection fell by $1,315 on average.
Key quote: “Our results suggest that the ex-post benefits of consumer bankruptcy on important outcomes, such as credit access and debt repayment, are significantly larger than previously assumed by this literature. Moreover, we find that consumer bankruptcy also impacts a number of outcomes previously assumed to fixed, such as asset holdings and labor supply.”
Data they used: Records from individual bankruptcy filings from 72 federal bankruptcy courts, matched to credit bureau records from TransUnion.