The Economics of Desertification, Land Degradation, and Drought

Box 3.2—Measuring land degradation
Determination of soil erosion rates provides the basis for analysis in the majority of the studies (Enters 1998).
According to Oldeman (1996), erosion is the most important driver of land degradation. As such, a variety of
methods and models exists to quantify the extent of degradation by determining soil erosion. A widely used
approach to predicting soil erosion is the Universal Soil Loss Equation (USLE; Wischmeier and Smith 1978).
The equation predicts mean annual soil loss from various variables, such as the erosivity of rainfall, the
erodibility of the soil, the length and slope of the soil, crop cover and management factors, and a conservation
practices factor. Originally, values were derived from data for the U.S. Midwest; however, because that data
cannot be assumed to be representative elsewhere, the equation has to be adapted to the sites where it shall be
applied. It is quite data demanding to fit the equation to local conditions; therefore, simpler empirical models,
such as the Soil Loss Estimation Model for Southern Africa (SLEMSA), were developed. The USLE was
developed further into the Revised Universal Soil Loss Equation (RUSLE) 46 (Renard et al. 1991). The Water
Erosion Prediction Model (WEPP) allows for more complexity (Laflen, Lane, and Foster 1991) but is even more
data demanding than RUSLE and USLE.
Other possible approaches assess the impact of erosion experimentally on fields or in a laboratory, and some
studies used assessments of soil erosion based on local expertise and subjective assessments (see, for example,
Alfsen et al. 1996; McKenzie 1994).
There have been various attempts to econometrically measure the soil erosion–productivity
relationship. A common solution is to establish a direct relationship by using simplified yield
functions that have topsoil depth as a dependent variable 47 (Gunatilake and Vieth 2000). Key
equations that link the economic behavioral model with the biophysical system of land degradation
are production functions that include the effect of changes in soils due to land degradation.
Many existing studies have estimated the impact of conservation measures on productivity
and have compared it to the impact of degrading (nonconserving) agricultural practice. Most studies
have indicated a positive effect of conservation measures on farm income or profitability. Byiringiro
and Reardon (1996) found that an increase in soil conservation investment per hectare from low to
high increases the marginal value product of land by 21 percent. Kaliba and Rabele (2004) analyzed a
positive impact of conservation measures and concluded that farmers gain more from soil
conservation measures than from using inorganic fertilizer alone. Other studies have found positive
impacts of conservation under certain conditions, such as plot size and slope (Adegbidi, Gandonou,
and Oostendorp 2004), rainfall conditions (Bekele and Drake 2003), type of conservation measure
(Kassie et al. 2008; Bravo-Ureta et al. 2006), and a reduced variability of yields (Kassie et al. 2008).
All of these studies analyzed the on-site effects of land-degrading and land-conserving measures.
Because degradation problems tend to be site specific and the adoption of soil conservation measures
depends on the decisions of individual farmers, most case studies are applied at the farm level (Lutz,
Pagiola, and Reiche 1994). To assess the extent of degradation, most studies limit themselves in their
analysis to the impact of certain processes of land degradation on agricultural yields, such as soil
erosion rates, 48 nutrient depletion, 49 soil compaction, or salinization.
Adoption Models
Adoption of sustainable land management techniques and investments in soil conservation practices
depends not only on monetary profits but also on factors that affect a more general definition of
benefits. Formal analysis conducted by McConnell (1983) shows how it may be optimal for farmers to
make production choices in which rates of soil depletion exceed what would be socially optimal.
Inefficiencies in capital markets, for instance, may truncate farmers’ planning horizons, thus
introducing soil-depleting biases, or it may affect farmers’ rates of time discount, so that it exceeds
46
Recent application of RUSLE has been done by Pender et al. 2006; Nkonya et al. 2008a).
47
The relationship between topsoil depth and crop yield can be estimated using the Mitscherlich-Spillman production
function, exponential functional forms (Lal 1981), and various other functional forms (Ehui 1990; Walker 1982; Taylor &
Young 1985; Pagiola 1996; Bishop & Allen 1989).
48
Measured as loss of soil (in tons per hectare per year).
49
Often assessed as nutrient balances (see, for example, Stoorvogel et al. 1993; Stocking 1986; Smaling et al. 1996;
Craswell et al. 2004).
74