Growth Accounting In Times of Turbulence And Death: Efficiency, Technology, Capital Accumulation and Human Capital 1929-1950

Growth accounting has long been an important tool to disentangle the proximate sources of economic growth (Solow: 1957, Griliches and Jorgenson: 1967). However, the method requires several assumptions about perfect competition in markets, the functional form of the production technology in use, Hicks-neutral technological change and constant factor shares in income. The majority of growth accounting studies has assumed that output is produced according to a two-input Cobb-Douglas aggregate production function. This reliance was questioned by Duffy and Papageorgiou (2000), who found that they could reject the Cobb-Douglas specification using a panel of 82 countries over a 28 year period. Furthermore, crosscountry evidence suggests large variances in labour shares of countries at various stages of development (Gollin: 2000). In addition, growth accounting results tend to be biased in the presence of inefficiency in the production process (Grosskopf: 1993).

Given some of the above-mentioned draw-backs, Kumar and Russell (2002) argue in favor of using a non-parametrical approach to growth accounting that neither requires assumptions about absence of market imperfection nor about the specific form of the production technology. In this study, we follow their reasoning and argue in addition that the notion of relative efficiency, or inefficiency, should be taken into account when performing historical growth accounting exercises, in particular during times of disintegration and war. In such turbulent situations institutional frameworks, access to world markets, war participation and the general economic environment are more likely to influence how efficient a country can transform the use of inputs into outputs.

The interwar and immediate post-war periods are interesting to study from this perspective since this period saw limited technological diffusion as countries became increasingly closed from world markets. In addition, the period was one of very modest capital growth in a majority of countries. In technological leading nations as USA and UK capital per worker hardly accumulated at all. In general, the period 1929-1950 saw the least capital growth per worker since 1890 (see Maddison: 1995). Nevertheless, there were increases in labour productivity during this period. Countries like Sweden, South Africa and Portugal nearly doubled in labour productivity between 1929 and 1950 whereas increases in Canada, Finland, Norway, Brazil, USA and New Zealand were substantial as well (around 50 percent). Foreman-Peck (1995, pp. 182-83) notes the paradox in that declining international trade, unemployment and general instability was accompanied by considerable real income increases in many countries. Foreman-Peck argues that prosperity of the time period was based upon the adoption of the revolutionary technologies from the late 19^th century, especially the applications of electricity, the internal combustion engine and the factory mode of production that offered scopes for catching-up.

However, the scope for catching-up could be dependent on the general “social capability” of a nation, as argued by Abramowitz (1986). The general social capability of a nation can be seen as a component of human capital which can abridge the process of technology adoption for late-comers. At the start of the 20^th century the rich nations had converged in terms of elementary-school enrolments and many of the poorer countries had started to expand mass primary schooling (Goldin: 2001, p.265). These educated cohorts started to enter into the workforce during the 1920’s and 1930’s and could have affected the general propensity for catching-up in certain nations, despite the world disintegration.

Download journal of accounting and economic: ziddu