Should One Worry about the Prospect of Deflation?

It seems that there is now virtual unanimity in regarding deflation as a threat to be avoided at all costs, comparable in severity to the problem of inflation in the 1970s. This is despite the fact that my suggestion of many years ago to define deflation as referring to declining output was not adopted: it would be neater for all if one spoke of price changes in terms of inflation versus disinflation, and of output changes in terms of reflation versus deflation, instead of using the terms “deflation” as the opposite of inflation. But there can be no doubt that what people are concerned about is the possibility that prices will fall. 

How did such a concern come to dominate economists, despite the fact that (gently, and/or temporarily) falling prices were as common as (gently, and/or temporarily) rising prices over many centuries prior to 1914 in a period that people do not generally equate with disaster in the Western economies? As with so many of the propositions in economics, it seems to have originated with no less than John Maynard Keynes. In chaps. 19 to 21 of the General Theory, Keynes was mainly concerned to refute the notion that falling prices provide a simple antidote to unemployment, pointing out that the then-traditional analysis neglected the impact of lower prices on aggregate demand. Once this is done, the net effect is equal to that which operates by increasing the real value of the money stock; and this could be achieved more readily by monetary expansion than by price deflation. Ergo, declining prices had no role to play in rational demand management policy.

One can agree with this proposition without going overboard, as most contemporary economists do, regarding mild price declines as a social evil to be avoided at all costs. How one moved from the consensual position that it makes no sense to seek reflation via price declines to the dominant contemporary view that price declines represent a threat to the possibility of managing demand is far from evident, but it seems that this is what happened.

It has been customary to buttress this view with a reference to Japan. In fact, Japan has not suffered anything like a catastrophic decline during the supposed lost decades; living standards continued to improve (as measured by the rate of growth of real GDP per employed person) at a rate of about 0.8 percent per year. Admittedly this is distinctly less than in the previous 20, 30, or 40 years, but this is mainly because the opportunities of rapid growth were distinctly less, because by 1990 Japan had more or less caught up with the world technological frontier. It is also less than the 1.6 percent per year that the US enjoyed, but the income distribution deteriorated far less and the labor force participation rate improved by almost 5 percent, as opposed to the decline of over 4 percent in the US. (A “cheap” way of improving measured productivity is to reduce employment per capita.) There are difficulties in making a comparison with the Euro Area, because of its changing composition and the earlier absorption of East Germany by West Germany and its consequent entry into the Euro Area, but it seems that the rate of productivity increase there was broadly comparable to that in Japan.

The important thing to appreciate is that it is the real growth rate per capita that is the relevant determinant of the improvement in living standards, and not the growth rate of the total economy in real terms (still less is it the growth rate in nominal terms). The major determinant of the faster growth in the US has been faster population growth (of about 27 percent rather than 2 percent), which has almost zero welfare significance.[1] Admittedly the US enjoyed faster per capita real growth, but since – as is well-known – the benefits of this went almost exclusively to “the 1 percent”, even this is of minimal welfare significance.

Let us ask analytically what is the effect of a negative rate of inflation. It is a problem that is caused by the zero lower bound to the nominal interest rate. If this constraint is in operation, this makes it impossible to use monetary policy to reduce the real rate of interest when desired, which has an effect in reducing aggregate demand. But note (1) That under many circumstances it is possible to offset this undesired effect through a more expansionary fiscal policy; (2) That when this is not possible (because, for example, of constraints on fiscal policy) it may still be possible to use “unorthodox” monetary policy (like quantitative easing) and (3) That in the worst analysis the effect is confined to the interest elastic component of aggregate demand. Even when it cannot be offset, the effect on demand is thus likely to be modest. It is hardly comparable to the problem of stopping inflation in the early 1970s.

Since there is no empirical evidence that disinflation is likely to be disastrous, nor convincing theoretical reasons for fearing modest and/or temporary price declines, one is bound to conclude that the present fevered attempts to prevent “deflation” are misguided.

[1]  Until recently I would have said zero welfare significance, but then I read Thomas Piketty’s book Capital in the Twenty-first Century. This shows that the concentration of capital tends to approach s/g, where s is the propensity to save and g is the growth rate of population and growth due to invention. It follows that high population growth tends to retard the concentration of capital, which some of us would regard as a benefit. This is the only benefit I can see in fast population growth.

This entry was posted in Inflation. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s