Penitence after accusations of error,...

Transcription

1 Penitence after accusations of error,... Comments Martin Eichenbaum NBER, July 2013

2 Background Economists have long argued about the role that policy played in major macro episodes and the way policy institutions responded to those events. Obvious examples: the Great Depression, the Great Inflation of the 1970 s, the Great Recession. Traditional research aimed at examining large policy errors has revolved around two competing visions the people vision, the institution vision.

3 The People Vision The Great Inflation Mistaken estimates of the natural rate (Orphanides) Policy makers efforts to learn about structure (Sargent). Monetary policy didn t satisfy the Taylor principle (Clarida, Gertler and Gali). The Great Recession A speculative boom - bust in housing markets.

4 The People Vision The Great Inflation Mistaken estimates of the natural rate (Orphanides) Policy makers efforts to learn about structure (Sargent). Monetary policy didn t satisfy the Taylor principle (Clarida, Gertler and Gali). The Great Recession A speculative boom - bust in housing markets. The housing crisis, the financial bubble, low SAT scores and frizzy hair were all due to due to bad monetary policy (Taylor)

5 The Institutional Vision The Great Inflation Time consistency and discretion in setting policy (Kydland and Prescot, Barro and Gordon). Expectations traps (Chari, Christiano and Eichenbaum) Dominance of the fiscal authority over the monetary authority. Sargent and Wallace on the fiscal deficits. Cochrane, Sims and Woodford on the fiscal theory of the price level.

6 The Institutional Vision and the Great Recession Over-reliance on short-term funding of long-term assets. the financial crisis as a bank-run. Time inconsistency and the too-big-to-fail problem. Bad incentives of the rating agencies,... Regulatory failure.

7 Julio s Contribution Analyze a particular force that he argues helps accounts for four qualitative changes in Fed policy. Tendency of Fed to act as if it were penitent when critics successfully argue that bad outcomes were a product of Fed mistakes". Basic idea: Critics argue that Fed policy was responsible for a series of ex-post bad outcomes. Fed becomes averse to vilified pattern of behavior. Julio s labels the Fed s change in policy as penitence.

8 The Four Episodes During the Great Depression, Fed became averse to open market purchases that raised banks excess reserves. After the Great Depression, Fed became averse to basing decisions based on whether banks were extending credit for speculative purposes or not. After 1960, Fed became averse to creating recessions to fighting inflation, and After 1979, Fed became averse to smoothing interest rates.

9 What does penitence mean? An empirical regularity:... the avoidance of actions that have been criticized for having led to poor outcomes in the past is enough for me to use the word penitence. A behavioral pattern somehow embedded into institutional outcomes. Lots of references to like sinners, remorse, apologies,... Paper doesn t really look at the mechanism underlying either interpretation of penitence. Not clear whether either interpretation has testable restrictions. Ex post characterization of changes in policy, understanding the psychological motives of policymakers.

10 A positive interpretation of the paper Putting more realistic people into the people vision Paper implicitly argues for the importance of the types of disagreements and learning between decision makers that mainstream models don t capture. Standard learning models involve a representative agent who is learning about the details of the world, Parameter values, signal extraction problems about the underlying state... (see page 16 of Julio s paper). The episodes that Julio summarizes involve much deeper disagreements about the nature of the world.

11 What I think Julio is saying Over the past 100 years, there were on four occasions, serious ex-ante disagreements about basic features of monetary policy. The data intervened, the critics were vindicated and the Fed learned its lesson. The implicit message: We need to work on models where there s persistent disagreements among agents and decision makers. I agree and think these types of models will have fundamentally important implications for the way we think about macro data and policy.

12 Persistent Disagreement The best economists in the world persistently disagree about fundamental issues. Fiscal austerity, financial regulation, the merits of further quantitative easing, fixed vs floating exchange rate regimes... If we don t agree, why should the agents in our model? The problem: Most of the time the world changes in slow, sometimes hard to discern ways. So it s easy to sustain disagreement. When evaluated using normal macro data, many models are essentially observationally equivalent.

13 Disagreement: the housing crisis Transcript from FOMC meeting, June 29-30, The price-rent ratio is very high by historical standards, suggesting that housing might be overvalued by as much as 20 percent., Joshua Gallin (Board of Governors). both nominal and real long-term interest rates have declined substantially over the last decade. Productivity growth has been surprisingly strong since the mid-1990s, producing rapid real income growth primarily for those in the upper half of the income distribution. And the large baby-boom generation has entered its peak earning years and appears to have preferences for large homes loaded with amenities....home prices actually look somewhat low relative to median family income, particularly compared to the late 1970s. Richard Peach (FRB, New York)

14 Disagreement at the Fed Transcript from FOMC meeting, June 29-30, [... ] if we are in a world that is going to have much lower real rates of interest for some time to come, one would expect to see the price-to-rent ratio go up. Maybe this line in the chart has another 40 percent to go to get to equilibrium! William Poole (President, FRB, St. Louis) Mr. Chairman, I d like to propose that he buy my house in Washington. Richard Fisher (President, FRB, Dallas)

15 Cracking the observational equivalence problem Bringing more or unusual data to bear on analysis. Use micro data, including data on agents actual expectations, to evaluate macro models. On rare occasions, the tectnoic plates shift and we receive compelling new data: the Great Depression, the Great Inflation or the Great Recession. Unless you have spiked priors, these types of observations force you to re-evaluate our view of the world.

16 Responding to the data On October , in responding to Congressman Henry Waxman, Alan Greenspan said he was shocked that financial markets had failed to respond as anticipated. I made a mistake in presuming that the self-interests of organizations, specifically banks and others, were such that they were best capable of protecting their own shareholders and their equity in the firms. He conceded that he had found a flaw in his bedrock believe of 40 years or more that markets tend to self-regulate. I made a mistake. It has been painful to realize that I had put too much faith in the ability of the markets to self-correct.

17 Conclusion Julio s paper reminds us of the need to model persistent disagreements among agents. There s a nascent literature on this topic that revolves around the failure of the common prior assumption. See Burnside, Eichenbaum and Rebelo for application to housing. This literature seems particularly relevant for macro where the key debates are often about issues where there s few big data points that compellingly favor one side or the other. A little bit like religion. Agents and policy makers ought to spend some time at non-denominational congregations.