Meet Joel Flynn: The Macroeconomist Asking Whether a Negative Mood Can Tank the Economy

Recently, discussion about a ‘vibe-cession’, or an episode of negative sentiment that might cause poor economic performance, has gained prominence in the financial press and drawn serious attention from policymakers struggling to read contradictory macroeconomic signals.

In recent research, new faculty member Joel Flynn together with Karthik Sastry (Princeton) make a first attempt to understand the macroeconomic consequences of narratives. They introduce new tools for measuring and quantifying economic narratives and use these tools to assess narratives’ importance for the US business cycle.

“The idea that sentiment matters for business cycles is an old one, going back at least to Keynes and Pigou," said Professor Flynn. "We had read [Nobel Prize-winning Yale economist Robert] Shiller’s book on Narrative Economics and it really inspired us to test and quantify some of the hypotheses he put forward. These ideas seemed very reasonable to us and coming up with a framework within which we could combine theory and data to test the ideas seemed like a very fun challenge.”

To measure narratives, they use resources not available to Keynes: large textual databases of what economic decision makers are saying and natural language-processing tools that can translate this text into hard data. Specifically, they study the text of US public firms’ SEC Form 10-K, a regulatory filing in which managers share “perspectives on [their] business results and what is driving them,” and their earnings report conference calls. To process these data, they used sentiment analysis, textual similarity analysis, and a fully algorithmic ‘latent dirichlet allocation’ model that looks for any repeating patterns in firms’ language. They then obtained quantitative proxies for things like firms’ general optimism about the future, their excitement about artificial intelligence trends, or their adoption of new digital marketing techniques.

They found that firms with more optimistic narratives tend to accelerate hiring and capital investment. This effect is above and beyond what would be predicted by firms’ productivity or recent financial success. Strikingly, firms with optimistic narratives do not see higher stock returns or profitability in the future and also make overoptimistic forecasts to investors. That is, firms’ optimistic and pessimistic narratives bear the hallmarks of Keynes’ ‘animal spirits’: forces that compel managers to expand or contract their business but do not predict future fundamentals.

They estimate that narratives explain about 20% of the US business cycle since 1995. In particular, narratives explain about 32% of the early 2000s recession and 18% of the Great Recession. This is consistent with the idea that contagious stories of technological optimism fuelled the 1990s Dot-Com Bubble and mid-2000s Housing Bubble, while contagious stories of collapse and despair led to the corresponding crashes.

“I would say the most important takeaway from our research is that contagious sentiment could, and much more research on this topic is needed, be one of the primary drivers of boom-bust behavior in the economy.”

How does this study fit into your broader research portfolio?

My broader research across macroeconomics and theory is often concerned with how frictions in decision making and strategic interaction at the microeconomic level generate emergent phenomena both at the level of the markets and the wider economy. The approach that we took in this paper is one example of that broader approach of focusing on a friction (here, it was misspecified beliefs that spread contagiously) and seeing how that can matter for important questions (here, why business cycles happen).