by Terence Mills

The interwar period of the 1930s and austerity measures had important economic impacts. Nick was a major contributor to debates on productivity in Britain, Europe and the United States.

I first encountered Nick in 1974 when I was a postgraduate and he was in his first spell in the Department of Economics at Warwick, but our friendship and research collaboration only began in 1987 when Nick became Professor of Economic History in the School of Economic Studies at Leeds University, where I was lecturer in econometrics.

Nick had been appointed to replace the recently retired Maurice Beresford and there must rarely have been such a contrast between two professors of economic history. I was familiar with all the pubs close to the university and the pair of us soon gravitated to them on a lunchtime, where we got to know each other’s research interests and shared our passion for cricket, particularly of the county championship during the 1960s.

I had just come across the work of Charles Feinstein and others on the timing of the ‘climacteric’, a downturn in economic growth in Victorian Britain, and, as a time series analyst, was intrigued as to how trend growth had been modelled in this literature. Nick’s interest was piqued after I introduced him to the then novel methods of trend extraction based on time series containing unit roots, which seemed to be particularly suited to historical output data.

We got together with Steve Leybourne, now Professor of Econometrics at Nottingham University but then a doctoral student at Leeds, who was developing a suite of computer programs for applying the Kalman filter, a recursive algorithm for computing optimal forecasts of the components of state space models, to economic time series data. These models provided a statistical framework for extracting optimal estimates of the ‘smooth’ trend component of the British output series, from which we could more easily establish the timing and extent of any climacteric in the data.

Four papers were quickly published, the first on the climacteric in Britain and France, the second examining trends and cycles in British industrial production from 1700 to 1913, and the remaining two extending this analysis to a wider set of European countries.

These laid the foundation for a research program that has lasted for a further thirty years in which Nick introduced concepts from endogenous growth theory. These included models of endogenous innovation which provided a theoretical economic framework, and I played around with various statistical approaches to modelling evolving trends. This enabled us to provide characterisations of the evolution of output, real wages and industrial production through the pre-industrial age, the breakout from the Malthusian epoch to the first and second industrial revolutions, and the presence and timing of the climacteric.

An offshoot of this research program were two papers which focused on the growth process in European countries after 1918 and during and after the ‘Golden Age’ of economic growth from 1946 to 1973. Here the framework was a model for evolving growth that allowed the ‘Janossy’ hypothesis to be tested, this being the idea that, after all shocks caused by war and recovery had played out, trend growth would revert to its pre-World War I level, for which, perhaps unsurprisingly, we found little evidence to support.

As well as being a world-renowned expert on the Industrial Revolution, Nick was keenly interested in the economic history of the interwar period, particularly the 1930s. I collaborated with him on two studies of the most important economic aspects of these years: an analysis of the impact of the austerity measures in place during the early part of the decade and an empirical examination of the role played by the subsequent expansionary rearmament policies during the run up to the outbreak of war in 1939.

Nick was also a major contributor to debates on productivity in Britain, Europe and the United States. We published historical studies on productivity in British railways and the economy in general during the Victorian era, in post-war British and German manufacturing and on UK price-cost markups over the long run. In recent years, we examined productivity trends in both the UK and the US, confirming that the slowdown in productivity since the financial crisis was indeed unprecedented in UK economic history.

In all our research together, my enduring memory of Nick was of him asking me if I could obtain a statistical measure of some key economic concept and of me replying that it was not immediately obvious how this might be done but that I would go away and consider how it could be. This invariably led to a non-standard statistical approach that often required the construction of new estimators, the setting up of simulation experiments, and the writing of programs to calculate them. These were some of the most challenging research problems that I have encountered but were immensely fulfilling to solve. When receiving the results, Nick would respond by saying that they were of great interest and could be interpreted in ways that provided evidence to support or even reject the issue that we were investigating.

Before his sad and untimely death, Nick and I had embarked on two further studies of the economic growth process. The first was on whether the growth process was, as is typically assumed, always exponential in nature or whether there have been any economic epochs in which it could be characterised as additive. The second was on testing Oded Galor’s unified growth theory: that countries will always break out of the Malthusian stagnation epoch of zero output per capita growth, but they would do so at different times and at different speeds. The preliminary empirical work for both has been completed but Nick’s crucial and imaginative economic and historical insights will no longer be available to enhance and interpret the statistical results: as a consequence, the projects currently remain in limbo.

About the author

Terence Mills is Emeritus Professor of Applied Statistics and Econometrics at Loughborough University.