Guest Post by Alex Rodrigue: The Fed and Lehman

The following is a guest contribution by Alex Rodrigue, a math and economics major at Haverford College and my fantastic summer research assistant. This post, like many others I have written, discusses an NBER working paper, this one by Laurence Ball. Some controversy arose out of the media coverage of Roland Fryer's recent NBER working paper on racial differences in police use of force, which I also covered on my blog, since the working paper has not yet undergone peer review. I feel comfortable discussing working papers since I am not a professional journalist and am capable of discussing methodological and other limitations of research. The working paper Alex will discuss was, like the Fryer paper, covered in the New York Times. I don't think there's a clear-cut criteria for whether a newspaper should report on a working paper or no--certainly the criteria should be more stringent for the NYT than for a blog--but in the case of the Ball paper, there is no question that the coverage was merited.

In his recently released NBER working paper, The Fed and Lehman Brothers: Introduction and Summary, Professor Laurence Ball of Johns Hopkins University summarizes his longer work concerning the actions taken by the Federal Reserve when Lehman Brothers’ experienced financial difficulties in 2008. The primary questions Professor Ball seeks to answer are why the Federal Reserve let Lehman Brothers fail, and whether explanations for this decision given by Federal Reserve officials, specifically those provided by Chairman Ben Bernanke, hold up to scrutiny. I was fortunate enough to speak with Professor Ball about this research, along with a number of other Haverford students and economics professors, including the author of this blog, Professor Carola Binder.

Professor Ball’s commitment to unearthing the truth about the Lehman Brothers’ bankruptcy and the Fed’s response is evidenced by the thoroughness of his research, including his analysis of the convoluted balance sheets of Lehman Brothers and his investigation of all statements and testimonies of Fed officials and Lehman Brothers executives. Professor Ball even filed a Freedom of Information Act lawsuit against the Board of Governors of the Federal Reserve in an attempt to acquire all available documents related to his work. Although the suit was unsuccessful, his commitment to exhaustive research allowed for a comprehensive, compelling argument to reject the justification of the Federal Reserve’s in the wake of Lehman Brothers’ financial distress.

Among other investigations into the circumstances of Lehman Brothers’ failure, Ball analyzes the legitimacy of claims that Lehman Brothers lacked sufficient collateral for a legal loan from the Federal Reserve. By studying the balance sheets of Lehman Brothers from the period prior to their bankruptcy, Ball finds “Lehman’s available collateral exceeds its maximum liquidity needs by $115 billion, or about 25%”, meaning that the Fed could have offered the firm a legal, secured loan. This finding directly contradicts Chairman Ben Bernanke’s explanations for the Fed’s decision, calling into question the legitimacy of the Fed’s treatment of the firm.

If the given explanation for the Fed’s refusal to help Lehman Brothers is invalid, then what explanation is correct? Ball suggests Secretary Treasurer Henry Paulson’s involvement in negotiations with the institution at the Federal Reserve Bank of New York, and his hesitance to be known as “Mr. Bailout,” as a possible reason for the Fed’s behavior. Paulson’s involvement in the case seems unusual to Professor Ball, especially because his position as a Secretary Treasurer gave him “no legal authority over the Fed’s lending decisions.” He also cites the failure of Paulson and Fed officials to anticipate the destructive effects of Lehman’s failure as another explanation for the Fed’s actions.

When asked about the future of Lehman Brothers had the Fed offered the loans necessary for its survival, Ball claims that the firm may have survived a bit longer, or at least for long enough to have wound down in a less destructive manner. He believes the Fed’s treatment of Lehman had less to do with the specific financial circumstances of the firm, and more with the timing of the its collapse. In fact, Professor Ball finds that “in lending to Bear Stearns and AIG, the Fed took on more risk than it would have if it rescued Lehman.” Around the time Lehman Brothers reached out for assistance, Paulson “had been stung by criticism of the Bear Stearns rescue and the government takeovers of Fannie Mae and Freddie Mac.” If Lehman had failed before Fannie Mae and Freddie Mac or AIG, then maybe the firm would have received the loans it needed to survive.

The failure of Lehman Brothers’ was not without consequence. In discussion, Professor Ball cited a recent NYT article about his work, specifically mentioning his agreement with its assertion that the Fed’s allowance of the failure of the Lehman Brothers worsened the Great Recession, contributed to public disillusionment with the government’s involvement in the financial sector, and potentially led to the rise of “Trumpism” today. 

Inflation Uncertainty Update and Rise in Below-Target Inflation Expectations

In my working paper "Measuring Uncertainty Based on Rounding: New Method and Application to Inflation Expectations," I develop a new measure of consumers' uncertainty about future inflation. The measure is based on a well-documented tendency of people to use round numbers to convey uncertainty or imprecision across a wide variety of contexts. As I detail in the paper, a strikingly large share of respondents on the Michigan Survey of Consumers report inflation expectations that are a multiple of 5%. I exploits variation over time in the distribution of survey responses (in particular, the amount of "response heaping" around multiples of 5) to create inflation uncertainty indices for the one-year and five-to-ten-year horizons.

As new Michigan Survey data becomes available, I have been updating the indices and posting them here. I previously blogged about the update through November 2015. Now that a few more months of data are publicly available, I have updated the indices through June 2016. Figure 1, below, shows the updated indices. Figure 2 zooms in on more recent years and smooths with a moving average filter. You can see that short-horizon uncertainty has been falling since its historical high point in the Great Recession, and long-horizon uncertainty has been at an historical low.

Figure 1: Consumer inflation uncertainty index developed in Binder (2015) using data from the University of Michigan Survey of Consumers. To download updated data, visit https://sites.google.com/site/inflationuncertainty/. 

Figure 2: Consumer inflation uncertainty index (centered 3-month moving average) developed in Binder (2015) using data from the University of Michigan Survey of Consumers. To download updated data, visit https://sites.google.com/site/inflationuncertainty/. 

The change in response patterns from 2015 to 2016 is quite interesting. Figure 3 shows histograms of the short-horizon inflation expectation responses given in 2015 and in the first half of 2016. The brown bars show the share of respondents in 2015 who gave each response, and the black lines show the share in 2016. For both years, heaping at multiples of 5 is apparent when you observe the spikes at 5 (but not 4 or 6) and at 10 (but not 9 or 11). However, it is less sharp than in other years when the uncertainty index was higher. But also notice that in 2016, the share of 0% and 1% responses rose and the share of 2, 3, 4, 5, and 10% responses fell relative to 2015.

Some respondents take the survey twice with a 6-month gap, so we can see how people switch their responses. Of the respondents who chose a 2% forecast in the second half of 2015 (those who were possible aware of the 2% target), 18% switched to a 0% forecast and 24% switched to a 1% forecast when they took the survey again in 2016. The rise in 1% responses seems most noteworthy to me-- are people finally starting to notice slightly-below-target inflation and incorporate it into their expectations? I think it's too early to say, but worth tracking.

Figure 3: Created by Binder with data from University of Michigan Survey of Consumers

Racial Differences in Police Use of Force

In an NBER working paper released today, Roland Fryer, Jr. uses the NYPD Stop, Question and Frisk database, the Public Police Contact Survey,  to conduct "An Empirical Analysis of Racial Differences in Police Use of Force." The paper also uses data collected by Fryer and students coded from police reports in Houstin, Austin, Dallas, Los Angeles, and several parts of Florida. The paper is worth reading in its entirety, and is also the subject of a New York Times article, which summarizes the main findings more thoroughly than I will do here.

Fryer estimates odds ratios to measure racial disparities in various types of outcomes. An odds ratio of 1 would mean that whites and blacks faced the same odds, while an odds ratio of greater than 1 for blacks would mean that blacks were more likely than whites to receive that outcome. These odds ratios can be estimated with or without controlling for other variables. One outcome of interest is whether the police used any amount of force at the time of interaction.. Panel A of the figure below shows the odds ratio by hour of the day. The point estimate is always above 1, and the 95% confidence interval is almost always above 1, meaning blacks are more likely to have force used against them than whites (and so are Hispanics). This disparity increases during daytime hours, with point estimates nearing 1.4 around 10 a.m.

Panel B shows that the average use of force against both blacks and whites peaks at around 4 a.m. and is lowest around 8 a.m. The racial gap is present at all hours, but largest in the morning and early afternoon.

Fryer builds a model to help interpret whether the disparities evident in the data represent "statistical" or "taste-based" discrimination. Statistical discrimination would result if police used race as a signal for likelihood of compliance of likelihood of having a weapon, whereas taste-based discrimination would be ingrained in officers' preferences. The data are inconsistent with solely statistical discrimination: "the marginal returns to compliant behavior are the same for blacks and whites, but the average return to compliance is lower for blacks – suggestive of a taste-based, rather than statistical, discrimination."

Fryer notes that his paper enters "treacherous terrain" including, but not limited, to data reliability. The oversimplifications and cold calculations that necessarily accompany economic models  never tell the whole story, but can nonetheless promote useful debate. For example, since Fryer finds racial disparities in police use of violence but not shootings, he notes that "To date, very few police departments across the country either collect data on lower level uses of force or explicitly punish officers for misuse of these tactics...Many arguments about police reform fall victim to the 'my life versus theirs, us versus them' mantra. Holding officers accountable for the misuse of hands or pushing individuals to the ground is not likely a life or death situation and, as such, may be more amenable to policy change."

Estimation of Historical Inflation Expectations

The final version of my paper "Estimation of Historical Inflation Expectations" is now available online in the journal Explorations in Economic History. (Ungated version here.)

Abstract: Expected inflation is a central variable in economic theory. Economic historians have estimated historical inflation expectations for a variety of purposes, including studies of the Fisher effect, the debt deflation hypothesis, central bank credibility, and expectations formation. I survey the statistical, narrative, and market-based approaches that have been used to estimate inflation expectations in historical eras, including the classical gold standard era, the hyperinflations of the 1920s, and the Great Depression, highlighting key methodological considerations and identifying areas that warrant further research. A meta-analysis of inflation expectations at the onset of the Great Depression reveals that the deflation of the early 1930s was mostly unanticipated, supporting the debt deflation hypothesis, and shows how these results are sensitive to estimation methodology.

This paper is part of a new "Surveys and Speculations" feature in Explorations in Economic History. Recent volumes of the journal open with a Surveys and Speculations article, where "The idea is to combine the style of JEL [Journal of Economic Literature] articles with the more speculative ideas that one might put in a book – producing surveys that can help to guide future research. The emphasis can either be on the survey or the speculation part." Other examples include "What we can learn from the early history of sovereign debt" by David Stasavage, "Urbanization without growth in historical perspective" by Remi Jedwab and Dietrich Vollrath, and "Surnames: A new source for the history of social mobility" by Gregory Clark, Neil Cummins, Yu Hao, and Dan Diaz Vidal. The referee and editorial reports were extremely helpful, so I really recommend this if you're looking for an outlet for a JEL-style paper with economic history relevance.

My paper grew out of a chapter in my dissertation. I was interested in inflation expectations in the Great Depression after serving as a discussant for a paper by Andy Jalil and Gisela Rua on "Inflation Expectations and Recovery from the Depression in 1933:Evidence from the Narrative Record." I also remember being struck by  Christina Romer and David Romer's, (2013, p. 68) remark that a whole “cottage industry” of research in the 1990s was devoted to the question of whether the deflation of 1930-32 was anticipated.
I found it interesting to think about why different papers came to different estimates of inflation expectations in the Great Depression by examining the methodological issues around estimating expectations when direct survey or market measures are not available. I later broadened the paper to consider the range of estimates of inflation expectations in the classical gold standard era and the hyperinflations of the 1920s.

A lot of my research focuses on contemporary inflation expectations, mostly using survey-based measures. Some of the issues that arise in characterizing historical expectations are still relevant even when survey or market-based measures of inflation expectations are readily available--issues of noise, heterogeneity, uncertainty, time-varying risk premia, etc. I hope this piece will also be useful to people interested in current inflation expectations in parts of the world where survey data is unreliable or nonexistent, or where markets in inflation-linked assets are underdeveloped.

What I enjoyed most about writing this paper was trying to determine and formalize the assumptions that various authors used to form their estimates, even when these assumptions weren't laid out explicitly. I also enjoyed conducting my first meta-analysis (thanks to the recommendation of the referee and editor.) I found T. D. Stanley's JEL article on meta-analysis to be a useful guide.