201
Principal time series data are drawn from publications of the CSO annual "Agricultural and
Pastoral Production" reports, one each for commercial and noncommercial farms. A quite complete
commercial farms series is available from 1975 to 1989. The noncommercial farms series begins in
1971, was interrupted from 1977 to 1981, and resumes for 1982 through 1985. Commercial farms
data are generally divided by farm area size categories (0—199 hectares, 200-1,999 hectares, 2,000+
hectares). Data on noncommercial farms are not broken down by size.
Noncommercial farmers are defined in these " Agricultural and Pastoral Production" reports
simply to be those farmers who are not commercial farmers. In order to determine who qualifies as
a noncommercial farmer, the 1973 noncommercial farms report defines commercial farmers as:
(a) a farmer who sold maize worth K600 or more at the line of rail prices;
(b) a farmer who grew virginia [sic] or burley tobacco in his own name and was
registered with the Tobacco Board of Zambia;
(c) a farmer who sold dairy products to, and was registered with, the Dairy Produce
Board;
(d) a farmer who had title to land.
This definition has been amended over time. By the time of the 1982 commercial farms report, the
commercial farms classification had been amended to include farms that sold to marketing boards any
crops exceeding the value of 150 90kg bags of maize, farms that bred or sold livestock to licensed
firms or boards, and state farms.
This classification and separate enumeration of commercial and noncommercial farms poses
problems for analysis. For example, as economic conditions change from year to year, a farmer may
choose to produce and market tobacco, thereby leaving the noncommercial classification and entering
the commercial farms classification. There is also some question as to the reliability of the population
lists for each classification, particularly since the mid-1980s when the state reduced its role in the
marketplace and as a consequence reduced its ability to monitor market activity. For these reasons,
it is difficult to determine whether increases in production of commercial crops is due to expansion
of output by existing commercial crops producers, to marketed output by farmers previously classified
as noncommercial, or to new market entrants. Better analysis would be possible if there were a more
reliable and stable long-term program of farm data collection.
Data on commercial farms production in Zambia are available from 1975 to 1989, and are
broken down by province and farm area size classification. These data are based on responses to an
annual mailed census sent on average to about 3,000 farming households nationwide. (Response rates
were generally low; e.g., 34 percent in 1978.)
The survey is intended to cover all the commercial farmers in the country and
questionnaires are mailed on the basis of postal addresses. The response has however
been quite poor. Many of the returned questionnaires were only partially completed
requiring many imputations for individual reports to be summarized on a
representative basis (CSO 1989a, p. iii).
A consultant's report to USAID (Scott 1990) notes, "The [commercial farms] list used is
known to be incomplete and the response rate is below 20 percent, so that the purpose of this annual
operation is not clear."