On May 21 and May 22, the Atlanta Fed hosted a virtual workshop on monetary and financial history. The workshop was organized by Michael Bordo (Rutgers University), William Roberds (the Federal Reserve Bank of Atlanta), and Warren Weber (formerly of the Federal Reserve Bank of Minneapolis and currently a visiting scholar at the Federal Reserve Bank of Atlanta). The workshop featured presentations on the history of money, banking, finance, and central banking. Six papers were presented along with two invited lectures. Both days of the workshop concluded with a panel discussion of contemporary policy issues from a historical perspective.
In this post, I'll discuss events from the workshop's first day, and in another post tomorrow I'll discuss day two of the workshop.
The workshop opened with a paper presentation by Ryland Thomas of the Bank of England and discussion by Clemens Jobst of the University of Vienna. The paper was titled "What You Owe or Who You Know? The Recipients of Central Bank Liquidity during the English Crisis of 1847," and its authors were Mike Anson (Bank of England), David Bholat (Bank of England), and Kilian Rieder (the Austrian National Bank). It focused on the Bank of England's responses to an 1847 financial panic. This episode is of interest to central bank historians because it was the first panic in which the Bank of England was subject to legislation (the Bank Charter Act of 1844, also known as Peel's Act) that limited its ability to extend credit in crisis situations. The paper analyzes the actions of the Bank during the panic, using a data set of credit actions (discounts and advances) hand-collected from the Bank's archives. The presentation emphasized that the constrained bank was able to provide emergency liquidity by rationing credit along several dimensions. In particular, the bank's credit actions discriminated against parties such as bill brokers, firms dealing in agricultural commodities, and firms located outside London, while favoring banks, London firms, and firms associated with bank directors. Ultimately these actions proved insufficient to stem the panic, however, and constraints on the Bank had to be eased by a government decree ("Chancellor's letter").
In his discussion, Jobst noted that the paper's results go against a traditional view of the 19th-century Bank of England as an arms-length ("frosted glass") lender. He also suggested that the paper's data set could uncover to what extent the structure of the London bill market shifted during the 1847 panic, with traditional dealers in bills ("bill brokers") apparently becoming disintermediated during the panic. Finally, Jobst noted that the 19th-century Bank used its interactions with the bill market in a quasi-macroprudential fashion, to maintain a sense of risk preference in the market as well as to exert control over the market. The paper's data set could thus be used to reveal the information available to the bank in its policy decisions as well as its resulting policy actions.
The second paper of the workshop was presented by Marc Flandreau of the University of Pennsylvania and discussed by William Goetzmann of Yale University. The paper ("How Vulture Investors Draft Constitutions: North and Weingast 30 Years Later") focuses on sovereign debt negotiations following a default by Portugal in 1828. The paper describes how a contender for the Portuguese throne was then able to obtain new loans from Portugal's London creditors during 1830–33, despite Portugal's recent default. In his presentation, Flandreau showed that a key aspect of the loan renegotiation was an 1827 British law that allowed a creditors' committee to control a defaulter's access to the London market, strengthening the creditors' bargaining position. The presentation also described how such control then allowed "vulture" investors in Portugal's defaulted debt to later earn large returns on their investment, and how the repayment of the renegotiated debt imposed high costs on Portuguese taxpayers.
In the discussion, Goetzmann observed that the 1827 law effectively inserted collective action clauses into the original Portuguese debt issue, even though the debt did not contain such clauses. Goetzmann also displayed a copy of a later (1855) lending agreement between the Portuguese crown and London creditors, which contained similar clauses to the contracts described in the paper, indicating that patterns documented by Flandreau persisted. Goetzmann then noted that cooperation between sovereign creditors was not unique to London, and he described how early (1790s) loans from Amsterdam creditors to the United States were subject to similar types of collective agreements. He further noted that many sovereign debt issues previously negotiated in Amsterdam were defaulted on during the Napoleonic period, creating a debt overhang problem and increasing the attractiveness of London as an alternative market for sovereign loans. International competition between sovereign lenders must therefore be considered as a constraining factor in the structure of sovereign loans.
Gary Gorton of Yale University delivered the workshop's first invited lecture, in which he described his research project with Ping He of Tsinghua University ("Economic Growth and the Invention of the Term Loan"). The focus of this research is the advent of term bank loans in the United States during the 1930s. Before that time, most business loans tended to be very short term—three months maturity or less—but could be rolled over at the discretion of the lending bank. Gorton noted that the 1930s witnessed rapid productivity growth in manufacturing concomitant with the expansion of term lending. He then presented a general equilibrium model that incorporates the growth effects of these two developments. In the model, an increase in bank term loans can result from an increase in financial efficiency (an improvement in banks' ability to predict borrowers' performance). Fitting this model to U.S. banking and macro data suggests that as much as one-third of economic growth during the 1930s resulted from an increase in financial efficiency experienced early in the decade.
Audience discussion of this result focused on regulatory changes as possibly contributing to the strength of the estimated financial efficiency effect. Among the changes mentioned were the debut of the Securities and Exchange Commission (which imposed new regulatory requirements on bond issues, raising the attractiveness of bank financing) and the Federal Deposit Insurance Corporation (which lowered the chances of an insured bank being caught in a run, making term lending less risky for banks).
The final paper of the first day of the workshop was presented by Marco Del Angel of California State University, Los Angeles, and was discussed by Kim Oosterlinck of the Université Libre de Bruxelles. The paper ("Do Global Pandemics Matter for Stock Prices? Lessons from the 1918 Spanish Flu," coauthored with Caroline Fohlin (Emory University) and Marc Weidenmier (Chapman University), focuses on the effects of the 1918 influenza pandemic on U.S. stock prices. The paper uses a new weekly stock price data set compiled by the authors, as well as data on death rates during the influenza pandemic period (1918–20). Del Angel presented results from vector autoregressions that show a persistent negative response of stock prices to upticks in pandemic deaths, even when the autoregressions incorporate series on war news obtained by filtering contemporary press reports. This same pattern holds when stock prices are disaggregated to the sectoral level. Negative pandemic developments thus exerted a strong negative effect on stock prices.
Several people in the audience argued that the paper's results might be affected by changes in Federal Reserve policies during this period. In the discussion, Oosterlinck suggested Fed policy could be incorporated through the inclusion of an interest rate series in the autoregressions. Oosterlinck also noted that the persistent effects of pandemic shocks on stock prices indicate an apparent inefficiency in markets' ability to process the economic impact of pandemic developments, a surprising pattern that merits some further discussion in the paper. He also recommended an event-study approach to analyzing stock price movements, which could exploit regional variation in influenza outbreaks, as a complement to the paper's existing econometric methodology.
The first day concluded with a panel discussion of the COVID-19 pandemic and ensuing policy responses. The first panelist was Alan M. Taylor of the University of California, Davis. Taylor's remarks drew on his recent paper on historical pandemics ("Longer-Run Economic Consequences of Pandemics," coauthored with Sanjay Singh of the University of California, Davis, and Òscar Jordà of the Federal Reserve Bank of San Francisco). Taylor and his coauthors survey a number of pandemics starting with the Black Death (1331–50) and conclude that their effects are different from the other major category of demographic shock: wars. Pandemics tend to make labor scarce relative to capital (wars can do the opposite), exerting downward pressure on returns to capital and interest rates. Moreover, these effects tend to persist. Taylor argued that these patterns are consistent over centuries of data and that we should expect similarly persistent effects from the COVID-19 pandemic.
The second participant in the panel was Ellis Tallman of the Federal Reserve Bank of Cleveland. Tallman noted that the pandemic has resulted in a historic U.S. debt expansion comparable to that seen during World War II. Tallman also noted that following the WWII expansion, U.S. fiscal and monetary policy turned more conservative, with the Korean War largely financed through tax increases. He observed that prospects for renormalization of policy were more uncertain under current circumstances and that the extraordinary magnitude of recent policy responses raised the issue of capacity to respond to possible future crises. Commenting on Taylor's work, Tallman observed that the rapid deployment of vaccines meant that fundamentals for a strong recovery were more favorable now than in many historical pandemic events.
The last panelist in the session was Gary Richardson of the University of California, Irvine. Richardson agreed with Tallman's comment that the availability of modern medical technology suggests that a stronger recovery might be expected now than what followed past pandemics. Another unusual feature of the COVID-19 pandemic has been that its death toll has been concentrated in older individuals, many of whom had left the labor force, again suggesting less impact from labor scarcity than experienced in earlier pandemics. Richardson also contrasted U.S. policy paths following the two world wars, arguing that policy after World War I was more "hands off," whereas policy following World War II was more consciously aimed toward a transition to the restructured peacetime economy. In his view, the current situation is more reminiscent of the post-WWI scenario.
In the subsequent audience discussion, Michael Bordo suggested that the post-WWII scenario is more like current circumstances, particularly with respect to the inflation outlook. Owen Humpage (Federal Reserve Bank of Cleveland) pointed out that post-WWII Fed policy was more accommodative than is commonly recognized due to the presence of the "even keel" policy of stabilizing prices on newly auctioned Treasury debt, initiated after the 1951 Fed-Treasury accord. Taylor agreed with the other panelists that modern medical technology has improved the fundamentals for recovery as compared to past pandemics, but he suggested that the traumatic experience of the COVID pandemic could have persistent behavioral effects—on, for example, patterns of consumption. Responding to Bordo, Richardson argued that while the Fed has the tools to control inflation, its independence is likely to come under pressure from both political parties, complicating the inflation outlook. Robert Hetzel (Federal Reserve Bank of Richmond) suggested that recent increases in the money stock also point to an increase in underlying inflation, as occurred immediately after both world wars, but Richardson argued that this would depend on the extent of earlier patterns of consumption behavior returning after the pandemic. Taylor mentioned that forecasts of inflation based on money growth have not done well in recent decades. Another issue discussed was whether the COVID-19 pandemic was more comparable to the 1957–58 influenza pandemic, which had relatively mild economic effects, than earlier pandemics, when modern medical technology was not available. Taylor suggested that because there have been relatively few modern pandemics, it is not possible to answer this question with high statistical confidence.
In tomorrow's post, I'll cover the presentations and discussions in the workshop's second day.