Openness, protectionism and Britain’s productivity performance over the long run STePH eN B ROADB e R RY A ND NIC HO lA S CRAFTS
The sectoral patterns of comparative productivity performance are quite varied. Here the nine-sector analysis provided in Broadberry (1998) has been simplified onto a three-sector basis, distinguishing between agriculture, industry and services, as in Broadberry and Ghosal (2002) and Broadberry (2005). Industry includes mineral extraction, manufacturing, construction and the utilities, while services includes transport and communications, distribution, finance, professional and personal services and government. Both Germany and the United States caught up with and overtook Britain in terms of aggregate labour productivity largely by shifting resources out of agriculture and improving their comparative productivity performance in services rather than by improving their comparative productivity performance in industry (Broadberry, 1998). Broadberry (1993) first established that comparative labour productivity in manufacturing has remained stationary in both the US/UK and the Germany/UK cases since the late nineteenth century, and Table 11.1 shows that this result
generalises to industry as a whole. By contrast, in both cases the aggregate labour productivity ratio moves broadly in line with the labour productivity ratio for services. Although both Germany and the United States have improved their labour productivity performance relative to Britain in agriculture, there has also been a dramatic decline in the importance of agriculture, which can be seen in Table 11.2. Whereas in 1870 agriculture accounted for about half of all
employment in Germany and the United States, by 1999 this had fallen to under 3 per cent. The shift out of agriculture has nevertheless had an important impact on comparative productivity performance at the aggregate level. This is because in the late nineteenth century Britain already had a much smaller share of the labour force in agriculture, which has had a substantially lower value added per employee than in industry or services. Hence the large share of resources tied up in agriculture in the United States exercised a significant negative influence on the aggregate US productivity performance relative to Britain in the late nineteenth and early twentieth centuries, and as the importance of agriculture declined this negative influence was removed. Similarly, the relatively large share of resources in German agriculture had a negative effect on Germany’s aggregate productivity performance relative to Britain until after the Second World War. Note that Germany in 1950 had a bigger share of the labour force in agriculture than Britain in 1871.1 The labour productivity differences in Table 11.1 may be explained in part by differences in capital intensity. So before we turn to measures of openness, it will be useful to provide estimates of comparative levels of total factor productivity (TFP), where TFP measures the productivity of labour and physical capital, weighted by their respective shares in income.2 Comparing Table 11.3 with Table 11.1, we see that although capital explains a part of the labour productivity differences between the three countries, it is not sufficient to eliminate differences in TFP, some of which may be explained by openness. For the US/UK case, trends in comparative TFP and labour productivity at the aggregate level are similar, but with TFP differences generally smaller than labour productivity differences. One point worth noting here is that whereas the United States overtook Britain before the First World War in terms of labour productivity, it was only between the wars that the United States gained a TFP advantage. This would be consistent with the emphasis of Abramovitz and David (1973, 1996) on the importance of capital rather than TFP in American economic growth during the nineteenth century. It is also consistent with McCloskey’s (1970) claim that Victorian Britain did not fail, in the sense that the United States was still catching up in terms of aggregate TFP levels. In services, too, note that the US overtaking of Britain also occurred later in terms of TFP than in terms of labour productivity. For the Germany/UK case, again comparing Tables 11.1 and 11.3 we see that trends are very similar for comparative TFP and labour productivity at the aggregate level, with differences in TFP generally smaller than differences in labour productivity. Note that in industry, Germany had caught up with Britain in terms of TFP as well as labour productivity before the First World War.