Abstract |
- The main objective of the study was an attempt to obtain a
general explanation for the geographical pattern of agricultural income.
This was undertaken in the belief that certain possible determinants
of agricultural income, specifically those related to natural
resource characteristics, have not been satisfactorily considered in
investigations up to the present time. Consequently, various determinants
which other researchers have found important, e. g., location,
education, age, etc. , were combined in one model, together
with a number of variables designed to measure various characteristics
of natural resources.
As far as the natural resource characteristics were concerned,
considerable emphasis in this study was placed upon an empirical
investigation of the range of choice hypothesis. This hypothesizes that
the narrower the range of choice permitted by the natural resource
complex, the greater will be the incentive toward specialization and hence higher incomes. Conversely, it was believed that the wider
range of choice of enterprises permitted, the greater will be the
tendency toward highly diversified, self-sufficient and low income
farming units. One of the complications involved in testing the range
of choice hypothesis was the desirability of measuring the opportunity
for pursuing a number of enterprises rather than the result of producing
them. It was believed, however, that one of the more important
determinants of the range of choice was the long-run climatic
conditions. In order to arrive at climatic measures suitable for the
purposes of this study, various modifications were made on those
developed by Thornthwaite and Mather.
Regression techniques were used extensively for the ensuing
empirical investigation, which relied heavily on data from secondary
sources. County data for Kansas and Oregon in 1959 and 1960 provided
the basis for the bulk of the empirical analysis although some
consideration was also directed towards these states for the years
1939 and 1949.
Three models, differing only with respect to the climatic
measures employed, were constructed consisting basically of two
equations. The first equation utilized the median earnings of male
farmers and farm managers as the dependent variable together with
measures of age, education, race, land capability, irrigation, range
of choice and off-farm employment as independent variables. The
second equation reflecting earnings from off-farm employment
consisted of the percentage of commercial farm operators who work
off the farm as the dependent variable and measures of location,
time available for off-farm employment, age, race and education as
independent variables.
The empirical results which were obtained for Kansas and
Oregon in 1959 differed in many respects. Most notable was the fact
that the range of choice hypothesis was far more significant in Kansas
than in Oregon. In addition, the expected decrease in importance
of this hypothesis through time did not materialize in either Oregon
or Kansas. Possible reasons for this are due to the shortness of the
time period studied and that there has been a differential impact of
technology. In the first equation, age and education also proved to
be significant in both Oregon and Kansas, while in addition in the
former state off-farm employment was important, as was irrigation
in the latter state. The latter two were also found to have increased
in importance over the last 30 years.
Empirical results obtained from the second equation revealed
that time availability and location were the most important determinants
in Kansas and Oregon in 1959. In Oregon, however, evidence
was obtained concerning the desirability of modifying the
conventional location matrix theory because the unique employment
opportunities offered by the lumber industry mean that off-farm
employment opportunities are not synonymous with the presence of
urban concentrations.
|