Agriculture Reference
In-Depth Information
degradation in terms of the provisioning services (crop yield) and does not include
impact of land degradation on other ecological services.
9.2.2 r
eLatIonShIp
between
L
and
d
egradatIon
and
I
tS
m
ajor
d
rIverS
To better understand the relationship between land degradation and its drivers, we
use the first difference econometric method.* The drivers of land degradation are
divided into two major groups: the proximate causes, that is, the primary or direct
causes of land degradation, and the underlying causes, which affect land degradation
indirectly. Using change of the NDVI as an indicator of land degradation or improve-
ment (Bai et al. 2008), the following theoretical first difference model represents the
key proximate and underlying causes of land degradation, which can change over
time:
ΔNDVI = β
0
+ β
1
Δprecip + β
2
ΔAgI + β
3
Δpop + β
4
ΔPov + β
5
Δgovt
(9.2)
+ β
6
Δroad + β
7
ΔGDP + β
8
ΔGDP
2
+ β
9
Δrurs + β
10
Δtenure + ε
i
where Δ = first difference, Δ =
X
t
= 2
-
X
t
= 1
, where
t
1
= baseline period and
t
2
= end
line period; pop = population density; precip = precipitation; pov = poverty; AgI =
agricultural intensification; road = access to roads; GDP = gross domestic product;
govt = government effectiveness; rurs = rural services (e.g., agricultural extension
services); tenure = land tenure; and ε
i
= normally distributed random error. Of the
causes of land degradation discussed in Nkonya et al. (2011a), topography is not
included because it does not change over time and therefore cannot be included in
the difference model.
However, due to data availability, not all causes of land degradation that change
over time will be included. Global level poverty, access to road networks, and other
socioeconomic underlying causes of land degradation are either not available as time
series or not available at all. Poverty and road density are available in only one
period and will not be included in the empirical model. The variables included in
the theoretical model are given in
Table 9.1
. Two proximate causes of land deg-
radation, namely, precipitation and agricultural intensification, are included in the
model. Precipitation represents the biophysical characteristics, while agricultural
intensification represents land management on agricultural land. Three underlying
causes, namely, government effectiveness, GDP, and population density, are used
in the model. Government effectiveness—government's capacity to implement poli-
cies with independence from political pressures and with respect to the rule of law
(Kaufmann et al. 2010)—will represent the institutions, which play a key role in
land management. Government effectiveness indicator ranges from -2.5 for weak
government effectiveness to 2.5 for strong government effectiveness (Kaufmann et
al. 2010). The relationship between change in GDP and the NDVI is presented using
a quadratic relationship between economic growth and the environment. The envi-
ronmental Kuznet curve (EKC) (Kuznet 1955; Grossman and Krueger 1991; Dinda
*
For a detailed discussion of the drivers of land degradation, see Nkonya et al. (2011a).
Search WWH ::
Custom Search