Distributional Impacts of Inflation

Eakinomics: Distributional Impacts of Inflation Over the holiday season, University of Chicago professor and former Chairman of the Council of Economic Advisers Austan Goolsbee took to the pages of The New York Times to argue that the debate over inflation was missing a key element: the differential impacts across different income levels and groups in the population. Clearly, he has a point about the basics: If you drive a lot, a rise in gasoline prices will have a disproportionate impact on your well-being. But I’m not convinced that his key proposal – that the federal government should release more data on distributional impacts – would change policy much at all. The starting point is his argument that “Differences like that mean that the inflation rate a person faces depends on what that person buys and where he or she lives and shops. People who live in more rural states, for example, most likely drive significantly more miles per year — so fuel inflation would matter a great deal to them.” Fair enough. He then argues that “the federal government does not release data showing how rising prices affect Americans across different income brackets. Without it, we may have a distorted picture of the economy. And with data the government already collects, it wouldn’t be hard to do.” As a result: “The Biden administration could ask the agency to compile distribution tables for inflation similar to what one might see for unemployment or taxes.” Perhaps, but getting statistically reliable estimates at a variety of points in the income distribution might require substantially larger monthly surveys. But put that technical quibble aside. What, exactly, would this accomplish? After all, we already suspect that inflation hurts the poor. Goolsbee himself points out “From mid-2019 to early 2020, the Consumer Expenditure Survey, the government’s primary data source on how consumers spend money, showed that those in the highest 20 percent of earnings spent, on average, less than two-thirds of their annual income. Households among the lowest 40 percent of earners actually spent more than their annual income, meaning they are most likely dipping into savings or taking out loans (although inflation can reduce the burden of debt for borrowers). Even at the same inflation rate, rising prices pinch spenders more.” What difference would having a monthly estimate of the difference make? Second, there is a real question to be answered as to what breakdowns are desirable. When one thinks about it, Goolsbee’s point is not really about inflation at all. At its root, the issue is that because different people consume different mixes of goods and services – with different prices – a dollar does not mean exactly the same thing to any two people. There is literally an infinite number of potential sub-groups one could construct: across incomes (what break points?), age, (what break points?), gender, race, geography, and so on. What are, and are not, desirable distributional comparisons? Finally, will the politics support the analytics? Goolsbee notes that the Bureau of Labor Statistics “developed an experimental series on how inflation affects those age 62 and above.” Why? Because you can make the analytic case that Social Security benefits should be indexed not to the general rate of inflation, but to the inflation in goods and services that matter to seniors. Did that affect the indexing of benefits? No. It is the age of focusing on inequality in all aspects of life. But acknowledging “inflation inequality” is one thing. Building a data apparatus to document it each month strikes me as costing more than any benefits that might accrue in better policymaking.