1958:419–431). In his 1968 presidential address to the American Economic Asso-
ciation, Kenneth E. Boulding (1970) speculated that the rate of return on the small
investment in economic research was several hundred percent per year over the period
from 1945 to 1965 (Boulding 1970:151).6
Despite these spectacular results, organized research and development (abbrevi-
ated as R&D) as a whole has had only a modest impact on the rate of economic
growth, as much of it generates no new knowledge. Edward F. Denison (1962) esti-
mates that only 7.5 percent of U.S. per capita growth from 1929 to 1957 can be
attributed to organized R&D. Today, over half of these expenses are for defense
and space programs, which have had only incidental benefits to civilian production.
Furthermore, many economists are skeptical that creativity flourishes in the institu-
tionalized R&D setting. Much technical progress results from on-the-job problem
solving and performance improvement rather than from work done in R&D depart-
ments. Technical progress in DCs’ firms, regardless of the department of origin, can
have substantial spillover effect in increasing LDC output through trade, the spread
of multinational corporations, and the purchase or borrowing of technology from
abroad (see Chapter 15). The impact of these spillovers may be greater than that of
the R&D of developing countries, which is usually a smaller percentage of GNP than
in DCs.
Yet, studies like Denison’s assume R&D spending is a flow cost used to produce
output in a given year rather than an asset that accumulates through time. Thus,
these studies, which assume current spending alone measures innovation, leave out
accumulated knowledge (Kamien and Schwartz 1982:51).
A firm’s size, monopoly power, and product diversification will determine how
much R&D it does. If it is large, monopolistic, and diverse, the enterprise is more
likely to capture the benefits from R&D (Kennedy and Thirlwall 1972:61–62).
However, in competitive product markets like grain, the individual producer can
appropriate only a small fraction of the benefits accruing from research. For exam-
ple, Griliches indicates that consumers received almost all of the social returns of
government-sponsored research on hybrid seed corn in the United States (Griliches
1958:419–437). Corn farmers, or even the hybrid seed corn industry, would prob-
ably not have undertaken the research, as private rates of return were far below
social rates. When such a divergence in these rates exists, the case for government
investment in research is strong.
In the 1950s and 1960s, U.S. and British technological leaders, with the high-
est ratio of R& D spending to GNP, had some of the lowest rates of productiv-
ity growth. How fast technology diffuses determines global inequality and LDCs’