Rational Expectations and Aggregate Supply

Now, let's apply this reasoning to Aggregate Supply.

Remember, Short Run Aggregate Supply (SAS) is defined by the Surprise Principle -- that is, the SAS curve tells us how production changes when producers are surprised by the change in the price level, or, alternatively, how prices change when producers are surprised by the growth or shrinkage of demand. In other words, the SAS tells us how prices and quantities change when producers' expectations are inaccurate. If their expectations were accurate, they wouldn't be surprised.

We will find that this tells us something about the relationship between the short and long run Aggregate Supply, when we put it together with Rational Expectations. Take a look at the next figure:

Figure 2: Long And Short Run Aggregate Supply with Irrational Expectations

The figure shows the long and short run Aggregate Supply. The dots labeled Y1, ... Y5 represent production and the price level (relative to the expected price level) in five successive years, Y1, Y2, ... Y5. If production and prices are often as they are shown here -- bunched together away from the long run aggregate supply curve, "biased" like the shots fired by candidate B in the target example -- then something is wrong. If people have rational expectations, they would soon learn to anticipate patterns like this, and would correct their expectations for the bias. So they would not be surprised by the changes in price levels and production, and that in turn means they would have to be back on the LAS -- these could not be movements along the SAS curve.

If people have rational expectations then movements along the SAS curve have to look more like Figure 3, below:

Figure 3: Long And Short Run Aggregate Supply with Rational Expectations

In Figure 3, the observations of production and the price level are approximately "unbiased" with respect to the LAS curve. That means that people can have rational expectations and still be surprised by these price and production changes -- so they can be movements along the SAS curve.

This is something we ought to be able to check. Is it really true that expectations are unbiased? It is a little tricky to get information about what peoples' expectations are, but not impossible, and economists have made a number of studies on this. The evidence points mostly in one direction. On the balance of the evidence, it seems likely that people's expectations about most economic events are indeed unbiased.

Let's see what that implies about government policy.

Macroeconomic Policy