612
F. Abdelradi and D. Yassin
CC impact ecosystems and as a result cause more reduced ecosystem services
like lower production of oxygen, lower biodiversity, loss of soil and trees. There
is a growing empirical work on the impact of CC on forests ecosystems in Africa.
As well as the livelihoods of the forest-dependent communities’ activities that rely
on vegetation services [39]. On the other hand, food security is impacted by CC in
the form of reduced food production due to changes in rainfall and temperatures,
which has led to increased demand for groundwater and decreased groundwater
recharge. For example, Funk et al. [40] have shown that for the period 1996–2003, a
reduction in rainfall of 50–150 mm that has caused a reduction in maize and sorghum
production across most of the eastern African countries. Correspondingly, for energy
consumption, it is expected to decrease because of warmer winters periods. However,
there is a substantial increase in the temperature during the summer together with
better climate circumstances in northern Europe, may cause a decrease in tourism
during the summer and an increase in spring and autumn.
5 Methodology
There are two main approaches extensively used in the literature to measure technical
efficiency: parametric SFA and non-parametric DEA. To take care of the identification problem caused by non-parametric models, SFA is used as an alternative. The
SFA approach, was commenced simultaneously by Aigner et al. [41]; Meeusen et al.
[42], makes a distinction between exogenous shocks outside the control and inefficiency of the firm. In contrast to DEA, SFA takes into account random noise and
allows for conventional tests of hypotheses. Additionally, SFA requires the specification of a functional form for the production function and a distributional assumption
for the inefficiency term. Agricultural production is stochastically determined since
it is affected by climatic changes, and the agricultural production literature is likely
affected by measurement and variable omission errors [43–45], choosing a robust
model is vital to consider these problems.
The literature extended the efficiency analysis from cross-section to panel estimation, which has helped to relax some of the assumptions and consider closer realistic
characterization technical efficiencies. Pitt and Lee [46] were the first to extend
the cross-section specification to panel data. They have developed ML estimation
approach using Normal-Half Normal stochastic frontier model. Battese and Coelli
[47] have extended Pitt and Lee [46] model to a Normal-Truncated Normal specification. Schmidt and Sickles [48] have shown that the estimation of a time-invariant
inefficiency frontier model could be carried out through conventional fixed-effects
estimation approaches. However, invariant inefficiency has been questioned especially with empirical applications [49]. Cornwell et al. [50] have introduced a solution
to this issue by proposing a quadratic specification model with individual-specific
slope parameters. This quadratic specification permits a unit specific pattern over
time of inefficiency. However, it needs a large number of parameters to be estimated.
Lee and Schmidt [51] proposed a different specification that is more parsimonious
F. Abdelradi and D. Yassin
CC impact ecosystems and as a result cause more reduced ecosystem services
like lower production of oxygen, lower biodiversity, loss of soil and trees. There
is a growing empirical work on the impact of CC on forests ecosystems in Africa.
As well as the livelihoods of the forest-dependent communities’ activities that rely
on vegetation services [39]. On the other hand, food security is impacted by CC in
the form of reduced food production due to changes in rainfall and temperatures,
which has led to increased demand for groundwater and decreased groundwater
recharge. For example, Funk et al. [40] have shown that for the period 1996–2003, a
reduction in rainfall of 50–150 mm that has caused a reduction in maize and sorghum
production across most of the eastern African countries. Correspondingly, for energy
consumption, it is expected to decrease because of warmer winters periods. However,
there is a substantial increase in the temperature during the summer together with
better climate circumstances in northern Europe, may cause a decrease in tourism
during the summer and an increase in spring and autumn.
5 Methodology
There are two main approaches extensively used in the literature to measure technical
efficiency: parametric SFA and non-parametric DEA. To take care of the identification problem caused by non-parametric models, SFA is used as an alternative. The
SFA approach, was commenced simultaneously by Aigner et al. [41]; Meeusen et al.
[42], makes a distinction between exogenous shocks outside the control and inefficiency of the firm. In contrast to DEA, SFA takes into account random noise and
allows for conventional tests of hypotheses. Additionally, SFA requires the specification of a functional form for the production function and a distributional assumption
for the inefficiency term. Agricultural production is stochastically determined since
it is affected by climatic changes, and the agricultural production literature is likely
affected by measurement and variable omission errors [43–45], choosing a robust
model is vital to consider these problems.
The literature extended the efficiency analysis from cross-section to panel estimation, which has helped to relax some of the assumptions and consider closer realistic
characterization technical efficiencies. Pitt and Lee [46] were the first to extend
the cross-section specification to panel data. They have developed ML estimation
approach using Normal-Half Normal stochastic frontier model. Battese and Coelli
[47] have extended Pitt and Lee [46] model to a Normal-Truncated Normal specification. Schmidt and Sickles [48] have shown that the estimation of a time-invariant
inefficiency frontier model could be carried out through conventional fixed-effects
estimation approaches. However, invariant inefficiency has been questioned especially with empirical applications [49]. Cornwell et al. [50] have introduced a solution
to this issue by proposing a quadratic specification model with individual-specific
slope parameters. This quadratic specification permits a unit specific pattern over
time of inefficiency. However, it needs a large number of parameters to be estimated.
Lee and Schmidt [51] proposed a different specification that is more parsimonious