Adoption and intensity of adoption of conservation farming practices in Zambia

2.3. Adoption of CF in Zambia
Despite its promotion over the last fifteen years, adoption of CF in Zambia is relatively
limited. 7 Haggblade and Tembo (2003) report that 20% of CF farmers in the 2002/3 season
were spontaneous adopters, with the 80% majority practicing CF as a condition for receiving
subsidised input packages. CFU reports that around 170,000 farmers had adopted CF on part
or all of their land in 2011.
Adoption tends to be incremental and partial in Zambia. Umar et al. (2011) found that almost
all farmers (out of 129 interviewed) practice both conventional and conservation farming on
different plots. Haggblade and Tembo (2003) reported that 0.25 ha of carefully managed
basin-planting CF can provide a minimal food security safety net for a family of four.
Primary constraints to adoption in Zambia are the use of crop residues for other purposes,
labor constraints and the limited potential to grow cover crops during the dry season. Of these
three constraints, a number of authors argue that labor constraint is the major constraint to CF
adoption in Zambia (Umar et al. 2011; Baudron et al. 2007; Haggblade and Tembo 2003).
The labor constraint manifests itself during land preparation and weeding. Preparation of the
planting basins is highly labor intensive and the hiring of labor is rarely feasible due to
unaffordable daily wages at peak times (also because hiring is not widely accepted culturally)
(Baudron et al. 2007; Mazvimavi 2011). Weeding requirements tend to be higher on CF plots
(in the absence of herbicide use) creating another labor constraint (Umar et al. 2011).
Findings related to land and labor are supported by Chomba (2004). His study, based on
nationwide post-harvest (1998/99-1999/2000) and supplemental household surveys
(1999/2000) covering 2,524 farmers in Eastern, Southern, Central and Lusaka Provinces,
found that household size and land size positively influenced adoption rates of CF during the
1998-2000 seasons. He also found that distance to markets and extension services were
important. This may be particularly so given that this study uses data collected early in the
promotion of CF in Zambia.
Nyanga et al. (2011) surveyed 469 farmers in 12 districts (in Southern, Central, Western, and
Eastern provinces) in an effort to understand their perceptions of climate change and
attitudinal and knowledge-based drivers of CF adoption. 8 The authors of this mainly
qualitative study documented a widespread awareness of increased climate variability. There
was a positive correlation between perception of increased climate variability and the use of
CF, but no correlation between attitudes towards climate change itself and CF. Interestingly,
the authors found a widespread expectation of subsidy, input packages or material rewards
for uptake of CF, which they argued has developed as a result of previous program’s use of
such incentives. This is concordant with a finding of Baudron et al. (2007), who reported that
50% of farmers dis-adopt CF if they no longer qualify for such incentives.
Most of these studies on the adoption of CF in Zambia rely on small samples purposefully
selected from regions where CF was promoted, which risks confounding the effects of
various interventions on adoption. The studies that rely on cross sectional data cannot control
for potential endogeneity caused by time-invariant unobservable household characteristics
that effect adoption. A limited number of studies that use multi-year data either fail to make
7
Throughout the paper, we use the term CA when referring to the general literature, but CF for the case in
Zambia.
8
Nyanga et al. (2011) use a proxy indicator for adoption of CA and consider any farmer with some area under
minimum tillage (both planting basins and animal draft power ripping) as a “CA adopter.”
5