Files
Abstract
The aggregate approach to the evaluation of
production oriented expenditure in the United
States has consistently shown high rates of
return to agricultural research investments.
These efforts, while showing the value of agricultural research, are limited in the information they can provide policy and budget decisionmakers.
More recently, efforts such as the analysis
of four major commodity groups by Bredahl and
Peterson and cross sectional studies by Evenson,
White, and Havlicek have begun refining the level
of analysis to individual regions and states and
for specific commodities (Evenson, Bredahl and
Peterson, White and Havlicek).
The objective of this study is to further disaggregate these commodity groupings of Bredahl
and Peterson into individual commodities and to
begin investigating the impact of interregional
research "spillovers." Case studies of corn,
wheat, and sorghum are made, using individual
states as observation units over the time period
for which research data on individual commodities
are available. The empirical results presented
in this paper are the results of some first
attempts at estimation. Further work is being
done to improve specification of variables
measuring weather, cash inputs, risk, and other
factors influencing yield response of grain commodities.
A special cross sectional-time series
algorithm is used in parameter estimation.
Theoretical Framework, Model, and Data
The conceptual framework of this study is
particularly constrained by the availability of
only 11 years of research expenditure data on
individual commodities (1967-1977). The short
number of years prohibits use of a 12-, to 13 year
polynominal Almon distributed lag which has
been used in previous studies to investigate the
research lag structure (White, et al., Quance
and Lu). The nonavailability of production
input data for specific commodities for other
than farm census years also limits utilizing the
traditional aggregate production function
approach to the analysis of agriculture research
expenditure.
Faced with these data limitation problems,
the framework of a supply response analysis is
developed as an alternative to investigating the
impact on research expenditures on productivity
of individual grain commodities. The supply
response model which is derivable from the production function, expresses the quantity of a
commodity offered for sale as a function of input
and output prices, technical parameters, and a
variety of shifters, such as weather and technological change. Supply analysis studies have
typically represented the effects of technological
change as a linear trend variable. While the research evaluation studies have not focused
on the processes of technological adoption and
diffusion, their analyses have shown that public
investments in agricultural research activities
have contributed to increases in productivity.
On this basis, lagged research expenditures on
individual commodities will be used as the technological change shifter in the commodity response functions.