The Economics of Recession: A Survey
Part 3/10, September 2022 [1]
Arturo Estrella

3. What causes recessions?

          Regardless of how a recession is precisely defined, its occurrence is clearly an undesirable event for most economic participants. Activity in certain industries can be severely impaired and at the aggregate level we see declining output, employment, and profits as well as rising unemployment rates. So, why do recessions happen? As may be expected, some economists blame the government and its institutions for recessions and others blame the private sector.

          Monetary policy has often been featured as the villain of the story. We have seen how already in 1837 Lord Overtone implicated the Bank of England as contributing to the adverse effects of recessions if not to recessions themselves. In the United States, a frequent target has likewise been the nation’s central bank, the Federal Reserve System. A caveat to this line of reasoning is that the United States experienced recessions long before the Fed was founded in 1913. Nonetheless, we can conceive of the possibility that monetary policy has always played a role if we accept that a decentralized system of money creation is just as much a form of monetary policy as a central bank with a specific structure and functions defined by law. Empirically, the articles referenced here sidestep this issue by considering U.S. data for the period since World War II, when U.S. monetary policy was firmly in the grasp of the Federal Reserve.

          Christina and David Romer [R11] take a novel approach to inference about the stance of monetary policy. Instead of relying exclusively on quantitative data, the authors employ a narrative approach whereby they examine documents released by the Federal Open Market Committee (FOMC), in charge of monetary policy decisions at the Fed, to “look for times when concern about the current level of inflation led the Federal Reserve to attempt to induce a recession (or at least a ‘growth recession’).” The authors find altogether six times at which the Fed made such decisions during their period of analysis and also find statistical evidence that the Fed’s actions were indeed followed by slowdowns in the U.S. economy that were more severe than would be expected absent the Fed’s actions. In short, they find both intent and consequences.

          A few years later, Bernanke, Gertler, and Watson [R12] consider both monetary policy and oil price movements as causes of recessions. Three major recessions in the United States followed soon after oil price shocks that occurred in the 1970s, suggesting a possible causal connection. Using a structural statistical model, the authors attempt to sort out the direct effects of monetary and oil price shocks on the aggregate economy. Their nuanced conclusion points mainly to monetary policy as a cause of the recessions. Specifically, they find that oil shocks per se had no direct causal effect but that the effects of monetary policy reactions to the oil shocks helped generate the recessions. In the causal chain of the story, oil price increases led to higher inflation, which caused the Fed to tighten, which in turn caused slowdowns in the real economy.

          An article by Sims and Zha [R13] again poses the question “Does monetary policy generate recessions?” Like the previous reference, this article employs an identified statistical model and comes to a nuanced conclusion. It assumes that monetary policy may either react to ongoing economic conditions or may introduce unexpected shocks, which may reflect a change in approach or simply random decisions unrelated to past or current economic data. The principal conclusion of the article is that most variation in monetary policy is of the reactive type rather than the unpredictable. If there is bilateral causation between policy and the economy, it may be difficult to disentangle cause from effect, but that does not necessarily negate the effectiveness of policy to affect real economic activity.

          Drawing on extensive experience as economic advisor at the Federal Reserve Bank of Richmond, Marvin Goodfriend [R14] recounts “How the world achieved consensus on monetary policy.” Goodfriend is not shy about assigning responsibility to the Fed for generating recessions, for example, in “The Volcker Fed brought the inflation rate down to 4 percent by 1984, although it precipitated recessions in 1980 and 1981–82 to do so.” More generally, he examines how concerns about inflation and the Fed’s reactions have led to economic downturns of varying severity. However, he proposes that the greater macroeconomic stability of the post-Volcker era has made it possible for the Fed to engineer mild recessions to stave off inflation and in some cases to avoid recession altogether. It would be intriguing to extend the analysis to the recession that followed shortly after the article was published.

          Adrian and Estrella [R15] take a data-driven approach to the question of whether Fed tightening causes recessions. Looking at changes in the federal funds rate as an indicator of Fed monetary policy, they infer the end dates of 13 monetary policy tightening cycles during the period from 1955 to 2005, each following a large sustained increase in the federal funds rate. They then show that 9 of those episodes were followed within 18 months by an NBER-dated recession and that 10 (including the 9) were followed by an increase in the unemployment rate. What distinguishes the remaining three episodes? They were the only ones in which the Treasury 10-year rate was above the 3-month rate, as it is most of the time, when the Fed stopped tightening. We return to this result later when we discuss forecasting recessions.

          All in all, the evidence seems clear that monetary policy has the potential to cause recessions and that it has been a significant contributing factor in the United States since the 1950s. One important reason may be a short-run policy tradeoff between inflation control and real economic growth. In fact, several of the foregoing articles suggest that the Fed may have intentionally triggered recessions in order to rein in inflation.

          How about fiscal policy? Here too there may be a tradeoff between fiscal responsibility and macroeconomic stimulus, which was the focus of considerable debate during the recession of 2008-09. Allsopp and Vines [R16] consider the proper role of fiscal policy when monetary policy follows a consensus approach that strives for medium-term price stability and short-term real stabilization, not unlike the approach that Goodfriend describes. The article argues that fiscal policy, contrary to what the consensus view may suggest, should not be relegated to a passive role in those circumstances but may be employed in the service of economic stabilization and long-run interest rate determination. The article provides an example of how fiscal policy could help trigger a recession in a country within the European Monetary Union if the constraints of the Stability and Growth Pact were binding on national fiscal policy.

          The real business cycle (RBC) theory of macroeconomic fluctuations dispenses with the need to blame government policies for the occurrence of recessions. In this view, aggregate fluctuations are driven by real shocks, that is, by unexpected changes in the structure of the real economy such as technological innovations. These shocks are then transmitted through the economy over time as a result of dynamic interactions across the various sectors. According to the RBC approach, the economy could undergo something that looks like the observed business cycle even if monetary and fiscal policy had no real effects.

          Hansen and Prescott [R17] apply RBC theory to U.S. data to investigate the possibility that technology shocks caused the 1990-91 recession. In contrast to econometric analysis such as that of [R12] and [R13], which is based on statistical estimates of empirical models, RBC proponents typically start with a theoretical model containing parameters whose values are calibrated based on related empirical analysis. The goal is to simulate the calibrated model and produce numerical time-series data whose properties mimic selected characteristics of the actual observed data as closely as possible. In [R17], the authors develop a three-sector model of the U.S. economy and use it to simulate real GNP from 1984 to 1993. A time-series plot of the results follows a pattern very similar to actual real GNP over the period, including a recession at around the same time as in the actual data.

          Caggiano, Castelnuovo, and Groshenny [R18] examine the recessionary effects of a different type of shock. In this case, the shock is to the level of uncertainty present in the economy at a macroeconomic level, which may be modeled empirically in alternative ways using financial market data. The results of this article suggest that an uncertainty shock has measurable properties similar to those of a negative demand shock, which could lead to a recession. Moreover, the response of the unemployment rate to uncertainty is found to be larger in recessions than in economic expansions.

          Most of the literature on the causes of recessions is based on data for a single country, implicitly assuming that domestic factors are sufficient to explain domestic recessions. Christiansen [R19] takes a broader geographical view of recessions by introducing the concept of “severe simultaneous recessions,” defined as recessions that occur simultaneously in at least half of the six developed countries included in the empirical sample of the article. Conceptually, we could surmise that these multi-country recessions occur as a result of common driving factors or of contagion across countries. The article finds that these types of recessions tend to be more serious than those that are limited to one country. The main focus of this reference is on prediction, which is discussed later in the survey, but it addresses an international dimension of recessions that is lacking from most of the research in the field.

Readings referenced from book The Economics of Recession

R11    Christina D. Romer and David H. Romer (1989), ‘Does Monetary Policy Matter? A New Test in the Spirit of Friedman and Schwartz’, in Olivier Jean Blanchard and Stanley Fischer (eds), NBER Macroeconomics Annual 1989, Cambridge, MA, USA and London, UK: MIT Press, 121–70

R12    Ben S. Bernanke, Mark Gertler and Mark Watson (1997), ‘Systematic Monetary Policy and the Effects of Oil Price Shocks’, Brookings Papers on Economic Activity, 1997 (1), 91–157

R13    Christopher A. Sims and Tao Zha (2006), ‘Does Monetary Policy Generate Recessions?’, Macroeconomic Dynamics, 10 (2), April, 231–72

R14    Marvin Goodfriend (2007), ‘How the World Achieved Consensus on Monetary Policy’, Journal of Economic Perspectives, 21 (4), Fall, 47–68

R15    Tobias Adrian and Arturo Estrella (2008), ‘Monetary Tightening Cycles and the Predictability of Economic Activity’, Economics Letters, 99 (2), May, 260–64

R16    Christopher Allsopp and David Vines (2005), ‘The Macroeconomic Role of Fiscal Policy’, Oxford Review of Economic Policy, 21 (4), Winter, 485–508

R17    Gary D. Hansen and Edward C. Prescott (1993), ‘Did Technology Shocks Cause the 1990-1991 Recession?’, American Economic Review: Papers and Proceedings, 83 (2), May, 280–86

R18    Giovanni Caggiano, Efrem Castelnuovo and Nicolas Groshenny (2014), ‘Uncertainty Shocks and Unemployment Dynamics in U.S. Recessions’, Journal of Monetary Economics, 67, October, 78–92

R19    Charlotte Christiansen (2013), ‘Predicting Severe Simultaneous Recessions Using Yield Spreads as Leading Indicators’, Journal of International Money and Finance, 32, February, 1032–43

[1] The original version of the survey was published in The Economics of Recession, Edward Elgar Publishing, 2017.