Research Overview - Bill Russell

My research addresses five main areas.

1.      The theoretical basis of the negative long-run relationship between inflation and the markup

I develop the theoretical basis for a negative long-run relationship between inflation and the markup. Firms face coordination problems when changing prices in an inflationary environment. To avoid costly coordination failures they adjust prices cautiously and with delay, which reduces markups (profitability) while prices are being changed. I argue that the uncertainty firms face when changing prices does not disappear simply because inflation becomes stable; consequently, permanently higher inflation can permanently lower firm profitability. This theoretical work also examines the policy implications of the inflation–markup link (see Russell 2006 for an early overview).

2.      Estimating the Long-Run Relationship between Inflation and the Markup

I estimate the relationship between inflation and the markup using two distinct data assumptions. Early empirical work, much of it with Anindya Banerjee, treated inflation as an integrated process of order one (I(1)) and the price level as I(2). Using data for many countries, frequencies and aggregation levels, that work found a clear long-run negative relationship between inflation and the markup in the Engle–Granger sense.

However, apparent integration can result from structural breaks in the means of the series; the true data-generating processes are likely stationary around shifting means. The later approach therefore treats inflation as stationary but with a frequently shifting mean. Modern Phillips-curve theories imply inflation varies around a long-run rate that can change when monetary policy or inflation expectations change. If those mean shifts are not accounted for, tests commonly detect a unit root in inflation even though the unit root has no behavioural relevance. Using about fifty years of U.S. quarterly data, Russell (2011), Russell et al. (2011) and Russell and Chowdhury (2013) show that ignoring shifts in mean inflation reproduces standard results from the modern Phillips-curve literature; once mean shifts are allowed for, there is no significant empirical support for New Keynesian, hybrid or Friedman–Phelps specifications, nor for the role of the modelled expected inflation term in NK/hybrid theories. Put differently, the finding that dynamic inflation terms sum to one largely reflects unaccounted shifts in mean inflation; under stationarity around a shifting mean the relevant parameters should lie within the bounds imposed by stationarity.

3.      General empirical macroeconomics

The third strand of my work is broader empirical macroeconomics. This includes studies of price and wage inflation, employment dynamics, the impact of share prices on the Australian business cycle, and the role of exports in transmitting foreign business cycles between countries.

4.      Modelling coffee prices

The fourth strand concerns modelling coffee prices. Changes in government policies over time have altered the ratio of the producer price to the terminal price of coffee. Proper cointegration or error‑correction models must account for these shifts in the coffee price ratio; failure to do so produces biased estimates and incorrect inference. By modelling these shifts explicitly it is possible to identify the long-run producer share of the terminal price and to quantify producer losses attributable to past policy changes.

5.      Methodology: Is Macroeconomics a Science?

Finally, I engage with methodological questions about whether macroeconomics is a science. Presenting Russell and Chowdhury (2013) revealed a common response: many economists accept the empirical evidence that contradicts standard modern Phillips-curve theories, yet continue to prefer those theories because they provide representative-agent micro-foundations and rational-expectations structures. I argue this selective acceptance is unscientific. Russell (2013) therefore asks what makes a discipline scientific and proposes four conditions for macroeconomics to meet that standard. Applying these conditions to Friedman–Phelps and New Keynesian treatments of the Phillips curve shows that, while these theories are falsifiable in a Popperian sense, their empirical assumptions and predictions are often compromised. In short, while macroeconomics meets some criteria of a science, it routinely fails others, and so has further progress to make before it can be regarded as a “pure” science.

Friedman, M. (1953).  The Methodology of Positive Economics.  In M. Friedman (ed.) Essays in Positive Economics, pp. 3-43.

Phelps, E.S. (1967).  Phillips curves, expectations of inflation, and optimal unemployment over time, Economica, 34, 3 (August), pp. 254-81.

Phillips, A.W.H. (1958).  The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957, Economica, 25, pp 1-17.

Russell, B. (2013) Macroeconomics: Science of Faith Based Discipline?,  Dundee Discussion Papers, Department of Economic Studies, University of Dundee, September, No. 276.

Russell, B. (2006).  Non-Stationary Inflation and the Markup: an Overview of the Research and some Implications for Policy, Dundee Discussion Papers, Department of Economic Studies, University of Dundee, August, No. 191.

Russell, B., A. Banerjee, I. Malki and N. Ponomareva (2011). A Multiple Break Panel Approach to Estimating United States Phillips Curves, Dundee Discussion Papers, Economic Studies, University of Dundee, June, No. 252.

Russell, B. and R.A. Chowdhury (2013). Estimating United States Phillips Curves with Expectations Consistent with the Statistical Process of Inflation, Journal of Macroeconomics, vol. 35, pp. 24-38.