RDP 2003-10: Productivity and Inflation 1. Introduction

Tim Bulman and John Simon

September 2003

That inflation has costs is widely accepted. What is less clear is the path by which inflation generates these costs – there are many alternative theories. The interaction of inflation with the tax system, the reduction in the value of the price mechanism, the diversion of resources from productive activities to managing inflation, or even the cost of adjusting prices on menus have all been posited as costs of high inflation.[1] However, quantifying these channels empirically is much harder than describing them theoretically. Regardless, whatever the channel of effect, they must all ultimately reduce output. And inflation's negative effect on output is most likely to be reflected in lower productivity growth.[2] Consequently, in considering the costs of inflation, the relationship between inflation and productivity is key. This paper investigates this relationship without attempting to isolate the strength of any particular channel. Notwithstanding this, our results suggest something about the characteristics of the channel and we discuss these in some detail later in the paper.

At the simplest level, broad historical correlations suggest a negative relationship between productivity and inflation (Table 1). Most OECD countries had low productivity growth and high inflation in the 1970s and, to a lesser extent, the 1980s. Productivity growth then generally increased through the 1990s at the same time as inflation generally fell.

Table 1: OECD Productivity and Inflation Experiences
Average annual percentage change in consumer price index and GDP per employed person
  1970–1973 1973–1979 1979–1989 1989–1999
Australia CPI 6.3 11.4 7.9 3.0
  Productivity 1.8 1.9 1.0 2.2
Canada CPI 4.6 9.0 6.8 2.4
  Productivity 2.6 0.6 0.9 1.1
Denmark CPI 7.1 10.5 7.1 2.4
  Productivity 2.7 1.2 0.7 1.8
France CPI 6.2 10.2 7.7 2.1
  Productivity 4.0 2.4 2.2 1.3
Germany CPI 5.3 5.0 3.0 –
  Productivity 3.4 2.7 1.5 –
Italy CPI 6.6 15.6 11.7 4.4
  Productivity 4.1 2.6 2.0 1.7
Japan CPI 7.6 10.3 2.7 1.3
  Productivity 5.8 2.8 2.6 1.1
Korea CPI 11.1 15.8 9.3 5.7
  Productivity 4.0 4.7 4.8 4.6
Netherlands CPI 6.8 7.4 3.0 2.2
  Productivity 4.2 2.1 −0.3 0.6
New Zealand CPI 8.0 13.0 12.0 2.0
  Productivity 3.6 −1.6 0.5 0.6
Norway CPI 7.9 8.5 6.0 2.6
  Productivity 1.8 2.6 1.9 2.3
Sweden CPI 6.8 9.3 7.9 3.8
  Productivity 2.2 0.5 1.4 2.5
Switzerland CPI 6.4 4.7 3.3 2.4
  Productivity 2.0 0.6 0.3 0.5
UK CPI 8.1 14.4 7.6 4.0
  Productivity 3.8 1.3 1.9 1.7
US CPI 4.9 8.2 6.1 3.2
  Productivity 2.3 0.6 1.2 1.7

Sources: CPI inflation – IMF; real GDP per person employed – OECD

The duration of these OECD fluctuations suggests a long-term, low-frequency relationship between inflation and productivity, however, this makes it difficult to establish statistical causation. To address this problem, this paper focuses on industry-level data. Industry-level data offer a number of potential benefits over aggregate data. Because each industry is different there is much more variation in the data, which in turns brings greater statistical power. Also, we suspect that industry-level characteristics may affect the nature of the relationship. Thus, the industry-level results are interesting in their own right.

Footnotes

See Blanchard and Fisher (1989) for a discussion of how nominal rigidities affect various economic models. [1]

Productivity is that component of output unrelated to changes in the amount of capital and labour inputs. While inflation may affect the accumulation of labour and capital it is most likely that its major effect will be to impede the efficiency of their organisation – hence lowering productivity. We address the possibility that inflation affects the accumulation of capital by looking at the difference between labour productivity and multifactor productivity: see especially Section 7.1. [2]