Debate: What Causes Recessions?

This article covers a number of mechanisms that can bring a business cycle to an end. I am not attempting to look at the academic literature in depth, rather it is mainly what I have seen in financial market research. It is written in a debate format; I list the various possibilities and discuss the pros and the cons (from my point of view). However, I am not indicating any preferred viewpoint. The idea is that this is an introduction to the topic, and I will refer back to this later (and possibly update to add links to more detailed articles).

The largest issue facing the U.S. economy (and the bond market) is: will the economy fall into recession before the labour market tightens substantially? If this was the case, the U.S. will be heading into recession with short rates at zero. The question then arises - what would trigger this recession? (The same situation holds true for many of the other developed economies, albeit with some differences in their situations.)

What Is A Recession? 

For the U.S., the National Bureau of Economic Research (NBER) has a Business Cycle Dating Committee that gives an official call as to which months the economy was in recession. The committee looks at a wide range of indicators, and takes considerable time to decide that the economy has fallen into recession (the call that the economy entered recession is often made after the recession has ended).

I am unaware of any other bodies making such determinations for other countries. As a result, a rule of thumb is typically used - the economy is in recession if real Gross Domestic Product (GDP) has contracted for at least two quarters in a row. (Real GDP means that GDP is adjusted for inflation. In a high inflation environment, nominal GDP may keep expanding even if real GDP is contracting.) This rule of thumb is often referred to in the U.S. as well, particularly if the NBER committee has not yet made a determination about a slowdown.

The advantage of the NBER Committee is that it eliminates one big problem: real GDP is often revised heavily, and so recessions appear and disappear if you use the "two quarter rule". Also, it reduces arguments over when recessions exactly occurred; otherwise partisan economists will tend to place recessions during the terms of Presidents that they do not approve of.

I will now turn to a variety of mechanisms that could cause recessions.

Malinvestment

Malinvestment is a term from Austrian economics. There are two versions of "malinvestment" that I refer to here:

1. The formal version, from Austrian Business Cycle Theory. I will again refer to the bibliographic post by "Lord Keynes" for a critique of this theory; for a pro-Austrian view, you can go the Mises Institute. The formal theory revolves around the idea that the rate of interest is not at the natural rate (although the Austrians use some long-winded variant of a natural rate) because it is set by the central bank at the wrong level. There is a convoluted mechanism relating the time preferences of investors to the time it takes for an investment to pay off.
2. The informal version, which is invoked by "Internet Austrians". This is usually a verbal description of stupid investments by the private sector, such as (a) the telecom boom that ended in 2000, and (b) the U.S. housing bubble. The central bank is usually blamed for this (because the government is always the problem if you are an Austrian), but the linkage to the formal Austrian theory is somewhat tenuous.

And in order to get Austrians really indignant, I'll lump Keynes' (and Kalecki's) observations on business investment in with this explanation. The Kalecki profit equation tells us that net business investment is an important source of profits for the business sector. If businesses stop investing, profits fall, reducing the incentive to invest further. This creates a downward cycle that leads to a recession.

Critique: The problem with this explanation is that a collapse in business investment is a defining characteristic of a recession. Saying that a collapse in investment lead to a recession is almost as uninformative as saying that people losing jobs causes the unemployment rate to rise. In order for this to be useful for predicting recessions, it is necessary to find a reason to see why investment will fall. Although this may be possible, it appears that the drivers for investment change from cycle to cycle.

However, the formal Austrian business cycle definition of malinvestment does appear to offer a hypothesis that you can test. My understanding is that it doesn't hold up to empirical analysis, but I have not looked at that in detail.

Inventory Cycle

Since inventory investment is a component of investment, this is similar to the "Malinvestment" theme. However, it is more closely tied to inventory data, and easier to characterise. It has become less dominant in recent decades due to (a) the rise of the service sector, and (b) outsourcing to low-cost countries. A good portion of inventories are now floating on the Pacific, and probably not picked up by national statistical agencies. That said, the auto inventory situation bears watching.

Monetary Policy Dominance

This mechanism is simple - when (real) interest rates get too high, a recession will ensue. The Volcker tightening episode is the classic example.

Other than that case, I have not seen too many people explicitly assert that high interest rates are the primary or only reason the economy goes into recession. However, my feeling that this view is embedded as a hidden assumption in a lot of economic analysis. For example, there is a wide consensus that interest rates are stimulating the economy, and the only real risk is that growth and inflation take off.

Monetarists (at least the old school Monetarists; not sure how the Market Monetarists would characterise things) would argue that the money supply is what matters, and not interest rates. And modern academics will insert a lot of mystical sounding discussions of "expectations". However, expectations about monetary policy are reflected in longer-term yields, so "expectations" are operationally equivalent to saying "bond yields are too high", and not just "the policy rate is too high".

Critique: This is part of my "Interest Rate Effectiveness" theme. Many Post-Keynesians (in particular, Modern Monetary Theory proponents) are skeptical about the effectiveness of interest rates.

Fiscal Dominance

As the economy expands, the amount of taxes paid increases, and social welfare transfers drop. This means that the fiscal deficit contracts. Since the government deficit is a net injection of money and income to the other sectors, this deficit contraction represents a drag on income for the other sectors of the economy. There are two channels of action:

1. the drag on incomes, as noted above; and
2. portfolio balance effects. Stock market valuations typically rise during an expansion, as does nominal incomes. The smaller deficit means that the weighting of government bonds in private portfolios drops. The demand to hold government bonds creates supply - in the form of a recession, which raises the deficit.

Critique:  One criticism is that you are supposed to look at "cyclically adjusted" deficits, which are not supposed to move across the cycle unless fiscal settings change. However, I do not see that cyclically adjusted deficits have any predictive power (even if the cyclical adjustment was done correctly, which does not appear to be the case). What I view as a more valid criticism is that this theory may not offer predictive power. We know that the fiscal deficit will always contract during the expansion (if fiscal policy settings are unchanged). The question is whether some means of telling whether fiscal policy has become "too tight", and thus cause a recession.

Financial Follies

This is another variant of "malinvestment", but focussing on the errors committed by financial market participants. During an expansion, financial market participants take on more and more risk, until they get overextended. The "de-risking" of financial market participants damages the real economy. (The last crisis was a classic example.) To my mind, Minsky developed the cleanest version of this theory, but he viewed his work as building closely upon Keynes.

Critique: Once again, the question of causality crops up. We know that financial market participants have an inability to remain disciplined for more than a couple of years, and so risk taking always increases during the cycle. The question, once again, is whether we can identify any threshold that will trigger a recession.

Energy Determinists

U.S. recessions after 1970 have often been associated with oil price spikes. This is no accident, according to the group I label the "Energy Determinists". These are people who are fixated on the price of energy to explain everything about the economy, based on the observation that energy inputs are a major component of a great deal of production. The economic effect of higher energy prices is straightforward: it is like  a tax on consumers of energy paid to energy producers. Those energy producers are often located in foreign countries, and in any event, generally save the extra profits caused by higher prices. This means that aggregate demand drops. 

Critique: I cannot argue with the physics, and the business cycle timing (for the U.S.) has been pretty good. That said, we still have a causality issue: energy prices tend to rise during an expansion, and fall during a recession. This implies that you should expect oil prices peak at the end of the expansion, even if they have no predictive power.

Random Shocks

This is a catch-all category; stuff happens and the economy is knocked into recession. This is used a lot in mainstream Dynamic Stochastic General Equilibrium (DSGE) models.

Critique: If you believe DSGE models, the economy is inherently stable, courtesy of the central bank. If the economy was stable, shocks are unlikely to be big enough to knock the economy into recession.

Real Business Cycle Theory

I almost forgot this, but since it has "business cycle" in its name, I guess I should include Real Business Cycle (RBC) theory in my list. This theory was the first wave of DSGE models, of which I am not a fan.

Believe it or not, there is not really a business cycle in RBC theory. The economy is assumed to be at "full employment" at all times. Although the unemployment rate moves up and down, that is solely because people decide that they would rather have leisure time than work. (The implication is that "The Grapes of Wrath" can be summarised as the story of some Okies on a real bummer of a California vacation.) Growth goes up and down solely because of "shocks" to productivity (they are "real shocks").

Critique: There are some ideas that are so absurd that only an academic could believe them. This is one of them. (With apologies to George Orwell.)

Final Remarks

There are a lot of things that build up during the cycle, and then unwind during the recession. Which factors are the causal mechanism remains unclear to me. As a result, I expect to refer back to this article over time as I look at this topic.

(c) Brian Romanchuk 2014