Experience Stern | Faculty & Research

Overview: In the paper, “Policy by Committee,” NYU Stern Professor Toomas Laarits and co-authors Kaushik Vasudevan (Boston College), Ben Matthies (Notre Dame), and Will Yang (Yale) explore the dynamics of group decision-making by analyzing transcript data and outcomes from Federal Open Market Committee (FOMC) meetings. 

Why Study This Now: There is little academic research studying group decision-making, despite the fact that consequential decisions are often made in groups. One possible reason for the lack of research might be that assembling data from group meetings can be difficult. However, the advent of Large Language Models (LLMs) has made this kind of data easier to gather and synthesize - a task that only a few years ago would have been prohibitively expensive. Using an LLM, the researchers systematized over 40 years of FOMC meeting transcript data (over 16 million words) to analyze group decision-making in the high-stakes setting of US monetary policy.

What the Researchers Found: Members of the FOMC use distinct economic models to interpret incoming economic data, which in turn leads to differences in policy preferences. The views of members whose models better explain the data had more influence in the subsequent policy decisions. As a result, monetary policy decisions reflect the economic models of members whose models do a better job rationalizing recent data, rather than a fixed policy rule.

What Does This Change: A diversity of thought and perspectives can be a strategic asset in group decision-making processes. These findings are relevant to any managers and executives looking to use the power of groups and committees to inform better business decisions. 

Key insight: “Committees may improve decision-making,” noted Laarits. “However, member selection is important, as potential benefits require members with a diversity of models who are sufficiently willing to entertain one another’s views.”