Can supply and demand explain inflation?

Yes, but not in the way that most people assume.

Brian Albrecht has a very good post criticizing ad hoc theories of inflation, such as those that point to a mysterious rise in “collusion”.  He favors the traditional supply and demand approach to prices:

The important part is that while supply and demand is overly flexible, the issue is worse with other models. Any story that relies on implicit agreements, folk theorems, and coordination failure arguments is likely unfalsifiable. There is no way to tell when firms move prices together due to collusion or due to normal supply and demand. The two outcomes are “observationally equivalent.”

Why did collusion start when it did? There was a supply shock. But do you know what other theory predicts supply shocks lead to higher prices? Supply and demand.

These models do not show us how a market switches from no collusion (low prices) to collusion (high prices). The folk theorem offers no means for the prediction of future prices. 

In my view, there are actually two problems here.  Collusion is generally a weak explanation for the rise in the price of a particular product, and changes in individual prices do not explain a change in the overall price level.

Consider the following hypothetical.  Suppose there is a 10% fall in the relative price of housing.  How would that impact the overall cost of living?  At first glance, you might assume that this would reduce the overall price level.  After all, housing has a weight of nearly 30% in the CPI. But this is to confuse changes in relative prices with changes in the overall price level.  The supply and demand model only explains relative prices (with one exception, which we’ll consider momentarily.)

Recall the maxim, “Never reason from a price change”.  Why did housing prices fall?  Let’s suppose that the decrease was caused by a decline in population growth.  With fewer people and a fixed quantity of land, assume the price of housing declines by 5%.  But fewer people also represents a negative supply shock, reducing the supply of labor.  That sort of shock tends to be inflationary.  Suppose non-housing prices rise by 5%.  In that case, the relative price of housing will decline by 10%, but the overall CPI will rise (as housing comprises less than 50% of the CPI.)   You cannot explain movements in the overall CPI by looking at changes in a single component of prices.  You need to consider why prices changed.  

So if the relative price of individual goods and services does not explain absolute changes in the overall price level, why do I claim that supply and demand can help us model inflation?  It turns out that there is one very important relative price—the price of money.  By definition, the price level is the inverse of the relative price of money.  Thus a rise in the price level is nothing more than an equivalent fall in the relative price (purchasing power) of money.  And S&D theory can explain changes in the value of money.  If you double the money supply and the demand for money doesn’t change, then the value of money falls in half.  Prices double.

Now let’s use S&D theory to consider the impact of a decline in population.  Start with an automatic monetary regime, such as a pure silver standard.  In the long run, a lower population would likely lead to a higher price level.  With fewer people there would be less demand for coins.  Prices rose after Europe was hit by the Black Death around 1348.  To be sure, fewer people might also lead to fewer silver discoveries, but the direct effect of a lower population reducing money demand would probably be more significant.

The short run effect of slower population growth is trickier.  It might initially lead to lower nominal interest rates.  Because the interest rate is the opportunity cost of holding silver coins, this might temporarily increase the demand for silver—pushing the price of goods and services lower.  During the gold standard period, prices tended to be lower during periods of low interest rates.

With a fiat money regime, one also needs to account for the response of the central bank.  Under NGDP targeting, a lower population would clearly lead to higher prices (as RGDP would fall.)  In contrast, if the central bank is targeting interest rates at a fixed level, then a falling population might accidentally trigger tighter money by pushing the equilibrium interest rate to a level below the policy rate.  This could at least partly explain what happened to Japan after the early 1990s. 

There are no easy answers here. But if we evaluate inflation using a model of the supply and demand for money, we can at least think about the problem in a coherent fashion.  In contrast, looking at the supply and demand for individual goods is of no help at all when trying to explain changes in the overall price level.  The S&D model can only explain relative prices.  And the price level reflects the relative price of only one good—money.