Victorian Growth: A Rejoinder

Abstract

It should be noted that the pattern, allowing for the differences in dates chosen for comparison, is similar to that in my original calculation and in Aldcroft's calculations. We are not disagreeing about the evidence but about its interpretation. The issue can be put as follows: if for purposes of description one wishes to distinguish two periods within each of which a single rate of growth in total productivity is supposed to obtain, what year should be chosen to separate the two periods ? That is, between which periods is the difference between the average rates of growth most significant, in the statistical sense of significant ? Aldcroft believes that it is almost certain that the periods should break in I 890, whereas I believe, in the words of my article, that given the uncertainties of the data ... the most precise defensible statement is that there was little cause for alarm in the 1 The source for these figures, as for Aldcroft's Table 2, is app. Table 2o in C. H. Feinstein, National Income, Expenditure and Output in the United Kingdom, I 855-1965 (Cambridge, I 9772). I have subtracted the annual rate of change of fixed reproducible capital .per man, multiplied by capital's share in domestic product (assumed to be o 4 throughout on the basis of a rough adjustment for the capital component of Feinstein's estimates of income from self-employment and income from land rents), from the annual rate of change of gross domestic product per man. The reported statistics are averages of these annual figures. The choice of peaks in the business cycle is W. W. Rostow's (his major peaks, in which conditions of virtually full employment were reached) in his British Economy of the Nineteenth Century (Oxford, I 948), P. 33.