Why has inflation stayed so subdued?

Numerous forecasters have claimed that inflation would at some stage pick up, either as a result of large fiscal deficits and high government debt or due to labour markets tightening.

Such forecasts have often been accompanied by predictions that the business cycle expansion, especially in the US, was approaching an end, recession was imminent, and that elevated asset prices would therefore be vulnerable to significant declines. Yet these predictions or expectations have largely been disappointed, posing a conundrum for the forecasting community. Across the US, the Eurozone, Japan, the UK, Canada and many other OECD economies, inflation has remained at or below 2% despite a prolonged business expansion and low levels of unemployment.

Where have the forecasters gone wrong? Why has inflation remained so subdued?

The fundamental error that forecasters are making is that they use inflation-forecasting models that rely on a “reduced form” analysis of inflation - that is, a proximate analysis of the causes of inflation. For example, large fiscal deficits by governments have in the past sometimes been associated with rising rates of nominal spending growth and inflation. In this case, forecasters would take a short cut - using a reduced form equation -perhaps forecasting inflation as directly related to the increase in the fiscal deficit after a period of several quarters.

Alternatively, tight labour markets and low unemployment were associated in the 1950s, 1960s and 1970s with wage increases and subsequently price increases. Here the reduced form analysis takes the form of the well-known “Phillips curve”, the idea that as unemployment falls wages (and prices) invariably start to rise. Inflation is modelled as inversely related to the unemployment rate.

The problem is that whereas these models may have worked in the past they are not working currently. The flaw in both these short-cut approaches is that forecasters are ignoring the true origins of inflation, namely excess growth of broad money¹ and its normal counterpart, bank credit. In the past decade there has been essentially no excess growth of broad money and credit in the developed economies and therefore no significant or sustained increase in inflation. However, because the pre-conditions for inflation in forecasters’ reduced form models - high fiscal deficits, or low unemployment - have emerged, the models and the modellers have continued to expect inflation sooner or later.

In short, provided that broad money growth across the developed world remains low and stable it should be entirely possible for the current business cycle expansion to continue for several more years with low inflation. Higher inflation, rising interest rates and a collapse of asset values is not inevitable – at any rate within the next two or three years.