Thursday, February 17, 2011

Supply or Demand? Beyond the labels.

Philadelphia Federal Reserve President Charles Plosser has questioned the usefulness of using monetary policy as a stimulus tool by saying “monetary policy cannot retrain people”. This statement is the result of his belief that the current problem with unemployment in the US is mainly structural. The only way to reduce it is by retraining workers from industries in decline (construction workers) to industries that are supposed to grow faster (nurses).

This is not a new debate and others have already replied to Plosser’s comments by showing strong evidence that the high unemployment rate in the US economy is present in practically every sector, every group or every region, so it is difficult to see how the data supports Plosser’s statements (see for example Paul Krugman making this point).

The debate about what causes business cycles: changes in demand or changes in supply is certainly not new to academics and it is normally seen as a differentiating point between groups of economists (Keynesians versus neoclassical). From the perspective of non-academics the debate is not always clear partly because of the way economists use the words demand and supply at the aggregate level and how it relates to the concepts of demand and supply in the market for a single good.

When you look at one market, demand and supply are always equal. One can explain fluctuations in prices or quantities by referring to movements in supply or demand but we normally do not think about periods where demand is too low or supply is too high (unless the market is not clearing).

When it comes to aggregate output, recessions are understood as periods where there is slack in the economy: unemployment is high, capacity utilization low. Some immediately associate these periods with the notion that demand is low. We have enough supply (workers, factories) to produce more but companies do not have enough demand to operate at full capacity. If we ask someone who has not studied economics before it is likely that they will find this story intuitive (and this is probably the reason why most people tend to have a “Keynesian” view of the world event if they have not been taught economics before). But I could also refer to recessions as a period where employment and output are low and this is about production, not just demand. So can we talk about recessions as periods where aggregate supply is also low?

Let’s ignore the supply and demand labels for the moment and separate the issue in several questions and see when there is agreement and when there is not.

Q1. Are there periods where the economy is producing below potential as measured by full utilization of all our resources? I am sure everyone agrees that this is the case. High unemployment represents a use of our labor resources below potential.

Q2. Why aren’t we using all of our productive resources during recessions? While most will agree that recessions are temporary, they disagree on what causes the friction that keeps employment temporarily below the level that ensures “full employment”. Charles Plosser's statement indicates that the friction (today) is mostly about a mismatch of skills (construction workers need to train themselves to become nurses). Keynesians believe that a low level of “aggregate demand” is keeping output low. But what type of friction explains the low level of aggregate demand? A textbook Keynesian model will talk about price rigidity as the main mechanism for such a friction. Price rigidity leads to changes in some key relative prices (such as the real wage) that generate a low level of production. Price rigidities are not that intuitive for those who have not studied economics. Does this really mean that if we let the price level drop we will go back to full employment? How fast?

What the traditional Keynesian model is probably missing is a more realistic friction that explains why “aggregate demand” remains low for such a long period of time. This mechanism cannot be a friction a la Plosser (mismatch of skills or sectors or regions) otherwise the Keynesian prescriptions do not work. If you ask someone who has not studied economics before they will probably tell you a story of companies getting stuck (at least temporarily) in a low output equilibrium. If I do not hire more workers then there is not demand for your goods and the reason why I do not hire more workers is that you do not hire them either. This is what economist refer to as models with multiple equilbria, which have been proposed by academics often but they tend not to be too popular because they are almost impossible to test and some see as an ad-hoc explanation of cycles. Notice that, strictly speaking, in those stories or models the problem is one of demand as much as supply (the coordination failure is as much at the production level as it is at the demand level).

Q3. Should fiscal policy and monetary policy be used to stimulate an economy during a recession? The answer depends on how we answered the previous question. If the friction is a mismatch of skills in the labor market, it is difficult to imagine how monetary or fiscal policy can speed up the adjustment. But if we believe that:

a) Companies are ready to hire workers who happen to be unemployed and waiting to be hired.

b) They do not do so because they do not see enough demand for their products.

Then there is room for monetary fiscal policy to increase the demand and the output (supply) of the economy. Whether one believes that the friction is just price rigidity or simply a failure to coordinate to a quick return to full employment, traditional “aggregate demand” policies such as monetary and fiscal policy will do the job (even if one could call them aggregate supply policies as well).

Antonio Fatás

on February 17, 2011 |   Edit