The Inequality of Recession

The public perceives recessions simply as a sign that the economy is “bad.” However, many policymakers continue to define recessions as “two-quarters of negative gross domestic product.” This outdated definition implies that policy responses could take more than six months to even begin addressing a downturn. The National Bureau of Economic Research (NBER), the organization informally responsible for determining recessions, has repeatedly clarified that each recession is evaluated holistically and is subject to expert opinion. However, given the consequences of any lag in the policy response, economists focused on policy have worked tirelessly on early indicators. Most notably, economist Claudia Sahm, creator of the Sahm Rule, relies on monthly unemployment data instead of quarterly GDP data. These popular rule-of-thumb recession indicators can be improved by acknowledging a harsh reality about the American economy: recessions generally impact vulnerable groups first.

Predicting recessions is always difficult, and anticipating economic downturns is a rather noisy game. Within this noisy space, the Sahm Rule is an incredibly reliable measure, potentially one of the best early indicators. There is a simplicity to the underlying mechanics of the Sahm Rule; prioritizing unemployment data lends itself quite naturally to adjustment since there are many measures of employment. However, when substituting different measures of unemployment, disaggregated for race, gender, education, industry, U-measure (which has different categories of labor utilization, delineating between a person who was laid-off and someone else who might have just had hours cut at work), or any other metric, two simple truths emerge: First, compared within a Sahm Rule construction, vulnerable groups often exist in a near-constant state of recession, relative to standard unemployment. Second, many disaggregated measures of unemployment are often much faster, albeit sometimes less reliable. But that does not mean these measures are not useful.

Newer research by Pascal Michaillat and Emmanuel Saez attempts to employ a similar approach, using job vacancy rates to create a probability of recession. Additionally, forthcoming research by Mark G. Sheppard further modifies the Sahm Rule, applying different measures of unemployment to provide several new and quicker indicators of recession, most notably the U-6 Sahm Rule, which incorporates more comprehensive measures of labor utilization. While this latter research is still in its early stages, the preliminary findings appear promising, with some indicators outperforming the official declaration of a “recession” from the Business Cycle Dating Committee of NBER by more than six months, which would offer incredible advance notice for policymakers aiming to preempt a potential downturn. The research relies on a fundamental fact about recessions: they are not uniform experiences but rather affect various individuals, locations, and industries in different ways. The findings rest on widely recognized labor market inequalities. Simply put, some people feel the impact of recessions sooner than others.

Imagine recessions as the proverbial house fire of the economy, with traditional indicators like the Sahm Rule serving as a kind of built-in sprinkler system. The basic Sahm Rule model is quite easy to understand; the underlying data is uncontroversial, and the parameters are fixed, which makes describing such a sprinkler system fairly straightforward. The original design’s simplicity is optimized to provide nearly automatic relief, often manifested in the form of fiscal policy, though these indicators can be improved. The advantage of the previously mentioned research is that it focuses on the economically vulnerable as a key signal. This is similar to adding a smoke detector alongside the more rigid sprinkler system, where some false alarms may occur due to the sensitivity of the new signal. Still, early detection can also be incredibly helpful in preventative measures.

Note: If you want to learn more about the author’s research on the Modified Sahm Rule, please check the following website and the Modified Sahm Rule tool, which is currently in beta testing and complements this academic work.