In our current research, we investigate the expectations and decisions of agents who lack a comprehensive understanding of their economic environment. These agents construct “mental models” of the economy derived from their experiences or mental simulations. These models are employed to formulate expectations regarding both exogenous and endogenous variables and their mutual relationships. Such expectations serve as the foundation for forward-looking decisions. As these choices feed back into the economic system, interesting dynamics may emerge. Our studies aim, among other things, to determine whether these dynamics can account for fluctuations at the macroeconomic level, or within specific economic segments. Our approach integrates both theoretical and empirical methods.
Recent papers of this research program are the following:
A growing body of evidence shows that, when forming expectations, households, firms, and even experts often deviate from rational expectations, adhering to intuitive models about macroeconomic relationships that conflict with actual experience. The “stagflationary” intuitive model – high output comes with low inflation – is a prominent example. Starting with a linear difference model in Blanchard-Kahn (1980) form, we develop a generic macroeconomic framework in which expectations emerge from an interplay of two mental systems. A rigorous thinking system forms expectations corresponding to standard rational expectations. An intuitive system forms expectations based on associative memory, perturbing rational expectations. As a result, households behave as if they were subject to cognitive discounting, and autonomous innovations in intuitive expectations are a source of macroeconomic fluctuations. We illustrate the tractability of the framework by applying it to “stagflationary” expectations and a New Keynesian model.
We examine intertemporal consumption-savings decisions under conditions where agents lack knowledge of the value function for future states. This ignorance may stem from several factors: the cognitive challenge of deriving the value function, limited opportunities for learning, or a belief in the instability of the data-generating process (DGP) that governs the dynamics of state variables. Our model posits that agents employ an availability heuristic, substituting the complex object of a value function with the “available” one of a utility function. The model assumes a finite planning horizon of arbitrary length. The decision-making framework consistently takes the form of a straightforward two-period problem. Simulations of this simple model effectively replicate key empirical phenomena, such as buffer stock saving and moderate consumption smoothing, thereby offering a compelling explanation for these observed behaviors.
Fluctuations in income are one of the most important sources of economic risk for households. Despite its crucial role for consumption and savings decisions there is very little research that focuses on households' expectations about their personal income. Based on the complete corpus of text data from the Wall Street Journal and the New York Times from 1980 to 2023, we estimate a topic model to investigate the role of media in expectation formation. We test the hypothesis that households pay attention to news topics that are most informative about their income prospects, in line with rational inattention.
Successful innovations are a key driver of long-run economic growth. In practice, potential innovations come with a great deal of uncertainty, i.e., a dearth of objective information on the likelihood of eventual success or failure. We present a tractable growth model in which entrepreneurs are characterized by their openness to novelty. Greater openness makes it more likely that, against the backdrop of uncertainty, potential innovations are checked out; but greater openness can also lead to exuberance, to negative signs being ignored – and then to misallocation and crisis-induced paralysis. We analyze this trade-off and show that it implies a hump-shaped relationship between openness and long-run growth. The calibrated model predicts that over a significant part of the range the negative effect of openness dominates, a result we show to be consistent with the empirical pattern. On the other hand, the calibrated model suggests that heterogeneity in entrepreneurial openness to novelty, leading to diversity in beliefs about an innovation, helps growth. Under uncertainty, belief diversity is an advantage and not an aberration.