Deflation: is it about growth or distribution?
If there is one topic that most mainstream economists can agree on, it is that deflation is fundamentally bad. As such, it should be avoided by all means and central banks should ensure that “it doesn’t happen here” as ex-Federal Reserve Chief Ben Bernanke once said in a famous speech.
The argument about why deflation is bad usually runs along the line that economic growth is harmed through several channels. Most notably, in the case of deflation, aggregate demand tends to fall because people, expecting lower prices in the future, will postpone some of their outlays. This reduced demand might put further downward pressure on prices, reinforcing the vicious deflationary spiral.
Economists also acknowledge that deflation has some re-distributional effects on wealth. However, most of them stress the negative impact of this redistribution on growth, without giving much thought to the cause itself. Indeed, in case of deflation, the purchasing power value of outstanding debt will increase, making it more and more difficult for debtors to service their debt at the risk of ultimately defaulting. This in turn is seen as reinforcing the downward pressures on growth, eventually leading to depression.
Interestingly enough, while most economists are in agreement about the negative impact of deflation on economic growth, empirical evidence is far less conclusive. A recent study by the Bank of International Settlements (see: Claudio Borio, et al. (2015): “The costs of deflations: a historical perspective”, BIS Quarterly Revue, March 2015) concludes that the link between output growth and deflation “is weak and derives largely from the Great Depression.”
To achieve this result, a sample covering up to 140 years, going as far back as 1870 and including up to 38 economies was analyzed. The study showed that the growth rate of economies in deflation was only half as large as that of economies in inflation (a statistically significant difference). However, dividing the data into three sub-samples (before World War I, interwar period and post-World War II) showed that the result was indeed only driven by the Great Depression experience. In the period before World War I, there was no noticeable output growth difference between the inflation and deflation periods; the same holds true for the post-WW II period.
But what about Japan? Hasn’t this country experienced a prolonged period of deflationary stagnation and wasn’t it one of the reasons why Bernanke gave his speech in 2002, laying the theoretical foundations of what would later be labeled “quantitative easing”? The BIS study shows that if instead of taking real GDP, one uses real GDP per capita or real GDP per worker, then – at least over the last 13 years – the Japanese growth performance can be roughly comparable to the US.
With these findings, one needs to ask why central banks are fighting deflation, increasing in the process their balance sheets multiple times. Three answers come to my mind. The first one is that they don’t know about or disagree with the BIS study.
The second one is that they parallel the 2007/8 financial crisis and its aftermath with what happened during the Great Depression, wanting to avoid at any cost the policy mistakes made back then. However, given that seven years after the latest crisis most economies seem finally to have recovered, there is a third explanation: they want to avoid the negative redistribution effects of deflation.
As mentioned above, deflation can really hurt debtors while, at first, benefiting creditors. I say “at first” because once a debtor defaults, its creditors take the loss. Governments are among the largest debtors. This is one reason why they have a natural penchant for inflation. And this is also the reason why central banks are “independent” from governments.
Then again, central banks are part of the government and battling deflation under the pretext that it harms growth has the positive side-effect of keeping indebted governments out of trouble. Did someone just say “conflict of interest”?