235
(2) any farmer who grew tobacco in his/her own name and was registered with the National
Tobacco Company of Zambia Ltd.;
(3) any farmer who sold to the Dairy Produce Board;
(4) any farmer who bred, reared, and/or fattened cattle or poultry and sold them to the Cold
Storage Corporation of Zambia, Poultry Processing Company, Ltd. or to any licensed
butcher or supermarket;
(5) any farmer who reared and/or fattened pigs and sold them to the Zambia Pork Products
or the Cold Storage Corporation of Zambia or to any licensed butcher or supermarket;
(6) all hybrid poultry breeders; or
(7) all state farms operated by the Agriculture Division of ZIMCO and other agencies on a
commercial basis.
This definition has continued with minor adjustments until 1990/91. 5 The data series reviewed here
(1975/76 and later) falls entirely under the "modern" definition. Data users who have occasion to use
any of the pre-1975 data should be aware of the definition change. While it is reasonable to expect
the estimated number of commercial farmers to be somewhat greater after the more inclusive definition
was adopted, the published farm numbers are not available to make the comparisons. 6
Commercial farm data are published at the provincial and national levels by size groups. A
wide range of data are collected including: area, yield, production, and sales of major crops;
inventory, production, and sales of livestock and poultry; inventory of farm implements; pesticide and
fertilizer usage; production expenditures; and farm gate prices. To collect this much information
requires a very long survey instrument. It is not difficult to understand why relatively few commercial
farmers responded when they received such a long and complex questionnaire in the mail.
The CSO data on commercial farms forms the second longest continuous data series when
compared to the other data sets—from 1975 through 1989/90.' Unfortunately, the accuracy of these
data is questionable for two reasons: the accuracy of the list frame used and the low response levels
to the mailed questionnaires. While it was not possible to evaluate the data sets' accuracy directly,
examining the response rates to the mail survey for the years in which the data were obtained along
with the number of commercial farms identified to receive the mailed questionnaire indicates problems
in both areas (table 8.4). 8
The number of questionnaires sent to commercial farms increased sharply from 1,603 farms
in 1975/76 to 2,595 farms in 1976/77. The number then remained around 2,000 until it dropped
suddenly to 1,072 in 1982/83. It is not clear whether these variations were due to incomplete or
incorrect lists or were actually correct farm numbers. However, the short time between the sizable
gains and losses support the first conclusion. It would have taken a very serious economic event to
suddenly halve the number of commercial farms in Zambia.
5 No data have been compiled after 1989/90. However, since the 1990/91 Census, the CSO has supposedly changed their
emphasis from "commercial" to "large (>20 ha)" farmers. Whether this transition has actually been made is unknown.
6 Given the problems noted later in this paper concerning the commercial farm estimates, this change in definition may
have had no effect on the published data.
Both the 1988/89 and the 1989/90 data sets have yet to be published. However, preliminary computer summaries are
available.
Maize production estimates for commercial farms are available as far back as 1940. See the annex 8.2 for a complete
listing of the maize data series.