208
The declining importance of maize relative to other crops in the commercial farms sector is
further confirmed by production quantity data. Table 7.4 presents these data in tabular form for all
recorded crops, while figure 7.3 illustrates changes in production for the three leading crops in the
commercial farms sector: maize, wheat, and soybeans. The graph indicates increases in wheat and
soybean production that are large in percentage terms but that do not match the magnitude of the
decrease in maize production for the same period. Note, however, that the most recent indicated
downturn in maize production (1989) is not substantially greater than previous downturns from which
production later recovered.
A more complete analysis of these data would attempt to reconcile fluctuations in production
with specific political and economic events such as droughts, changes in government, currency
devaluations, or price liberalizations. Such an analysis would require additional data not contained in
the CSO reports and is therefore beyond the scope of the present exercise. In the absence of specific
statistical controls for such events, a growth rate may nonetheless be calculated, though it matters a
great deal where one places the starting and finishing points for the calculation.
The growth rate "r" is derived as follows:
(1) Y = a(l+r)'
where Y is either area, production, or yield, depending on which growth rate is desired. The variable
T is a time trend.
Equation (1) can be linearized in logs as shown in equation (2):
(2) LnY = Ln(a) + Ln(1 +r)(T) = A + BT
Equation (2) then forms the basis for a linear regression of the log of either area, production, or yield
on a time trend. The growth rate estimate is then calculated from the ordinary least squares parameter
estimate for the trend variable:
(3) r = eB - 1
Area growth rates estimated from the data in table 7.3 are presented in table 7.5 for two time
periods: a fifteen-year period from 1975 to 1989 and a seven-year period from 1983 to 1989. The data
suggest that area under maize has declined at an average annual rate of 4.1 percent since 1974, with
the decline accelerating in more recent years. On the other hand, areas under wheat, soybeans, and
several other crops have increased at an accelerating rate. Given the volatile nature of the figures, care
must be exercised in interpretation.
The CSO data permit further disaggregation of commercial farm data by region, but the way
data have been disaggregated has changed over time. In 1975-76, data were disaggregated by Central,
Southern, and "Copperbelt and others" categories, along with a national total. In 1981-82, this
division was changed to Central, Lusaka, Southern, and "Copperbelt and others," a system of
classification that remained unchanged through 1989-90. In order to facilitate statistical computations
and conserve on reporting, two regionally disaggregated time series were generated from these data:
northwest representing the data from "Copperbelt and others" and including the Western,
Northwestern, Copperbelt, Luapula, and Northern provinces; and southeast including the Central,