Any residually measured total factor productivity index confounds the productivity gains achieved from better inputs and greater organizational efficiency - movement toward the production function - with the output from process-related technological change. Of course, what may be classified as organizational or management efficiency in one firm or one industry could result from a technological breakthrough or an entrepreneurial insight in another firm or another industry. This relationship suggests the importance of technology diffusion on the measured productivity growth of any one firm or industry. Some do not view this as an issue at the aggregate level, except in the case of international technology transfer. However, it is an issue in the sense that the diffusion of technology supports new growth theory by emphasizing purposive actions in the identification, adoption, and implementation of others’ technology, that is, technological change is endogenous.
Beginning with Denison (1962), efforts were made to identify empirically factors underlying changes in a production-function shift parameter or index such as A(t). That is, economists were not simply interested in measuring the productivity residual, but also explaining it. Once the key explanatory factors could be identified, policies could be formulated to enhance economic growth. The first order of business was to make sure that the researcher had actually measured technological change correctly.
In analytical terms, growth accounting is an attempt to remove from the A(t) residual all factors except a pure technological-change component. The idea being, of course, that by breaking down observable output growth into components associated with measurable factor inputs, a purer residual will result, and that residual will represent or measure technological change. After Denison’s pioneering research, he and others pursued to decompose total factor productivity growth into a number of categories. This early research of Denison (1967, 1972, 1974, 1979, 1984), Kendrick (1973), and Jorgenson and Griliches (1967), and others, is summarized in terms of the more recent work of Kendrick and Grossman (1980) as reported in Table 5.1. The average annual growth rate of total factor productivity in the US domestic economy between 1966 and 1976 was 1.4 percent. In that time period, advances in knowledge - for example, from R&D, learning by