Macroeconomic Theory
Jeffrey Parker, Reed College

Paper of the Week

Assigned paper

Coibion, Olivier, Yuriy Gorodnichenko, and Rupal Kamdar. 2018. "The Formation of Expectations, Inflation, and the Phillips Curve." Journal of Economic Literature 56 (4):1447-91.

Reading suggestions

• The Journal of Economic Literature is published by the American Economic Association as a journal of detailed surveys of emerging bodies of literature. As such, its articles tend to be long and boring, because their purpose is to provide a detailed summary of a large collection of papers on a topic. This particular paper is not nearly as long and detailed as some in the journal, but it is still a bit longer and more comprehensive than some of the papers of the week.
• In the 1970s, macroeconomists (starting with Robert Lucas) began to incorporate expectations explicitly into their models. This created an immediate problem of observability: the government did not collect data on economic agents' expectations about inflation, output, or other variables. Because they could not simply plug in a variable for expected inflation, they made assumptions about how expectations were formed as functions of observable variables and plugged those functions into their equations. This is the origin of the "FIRE" (full-information rational expectations) model of expectations, which attempted to improve on older, "myopic" expectation models in which people were assumed to expect the present conditions to continue into the future.
• Much of this paper is concerned with different measures of inflation expectations that have emerged since that time. It assesses their sources and their performance in the context of the Phillips curve, one of the most important applications of inflation expectations both in the 1970s and now.

Questions for analysis

1. What are the principal ways that the full-information rational-expectations hypothesis of expectations formation has been tested and what were the basic outcomes of the test?
2. Briefly describe some of the alternative expectation models that have been proposed and tested.
3. Briefly discuss the recent behavior of the Phillips curve when expectations are proxied by the measures shown in Figure 1. 
4. Briefly discuss the nature, advantages, and disadvantages of "market-based" and "survey-based" expectations measures. Among the surveys discussed in Table 7, which seem to be the most useful and which the most flawed, and why?
5. How do these various expectations measures perform when as components of the expectations-augmented Phillips curve?