1.3 Research questions
This study’s aim was to examine the extent to how electricity access influences industrial output in Uganda,
while addressing the following research questions:
1) What is the effect of electricity consumption on
industrial output?
2) What is the effect of education on industrial output?
3) How does labour employed influence industrial
output.
To answer the above research questions, the study
used a Cointegrated Vector Autoregression (CVAR) as
an estimation technique. Secondary data (time series
data) on the inputs (electricity consumption, labour,
education, and industrial output) were extracted from
the industries for a period of 10 years (2009–2019).
Additionally, we propose relevant policy recommendations.
1.4 Theoretical framework
In analyses of the role played by electricity in production process, growth models (exogenous growth
models such as the Solow’s growth model, the RamseyCass-Koopman’s model, the Diamond overlapping
generation model, as well as the endogenous growth
models) have been extended. Besides capital, labour,
and knowledge (technology), electricity has been
explicitly modelled as an argument for the economic
output (Alley et al. 2016; Akekere & Yousuo 2013;
Danish et al. 2015; Maxim 2014; Palit & Bandyopadhyay 2016; Yuan, Kang et al. 2008). This research has
implications for economic theory. The output elasticity
of electricity is nearly triple its share of inputs in production. The marginal revenue product of electricity is
nearly triple the marginal revenue products of labour
and capital inputs at equilibrium. Electricity consumption, akin to R&D, has a much larger role in industrial
output than postulated in production theory. Differences in the output elasticities between labour and
the labour employed functions raise additional debates
about the adequacy of the human capital theory in
explaining industrial output.
2 REVIEW OF LITERATURE
Most studies on the effects of electricity as a component of energy on output and economic growth
frequently assume that energy enters the production
function independently or in a Hicks-neutral fashion (see Akekere & Yousou 2013; Samuel & Lionel
2013; Stern 2004; Yuan et al. 2008). Limited attention has been paid to which energy may enter the
production functions. Given the fact that most industrial machines depend on electricity to perform, the
capital-augmenting factor of energy performance in a
production model appears to offer a promising way of
appraising the role of electricity on industrial output.
Studies on industrial electricity access have attracted
a lot of information, however studies are limited with
deviating evidence, and there has not been a consensus
over the effects of electricity on industrial output.
In Pakistan, Qazi et al. (2012) used the Johansen
cointegration approach based on VAR to conduct a
study on the relationship between disaggregate energy
consumption and industrial output. The study covered the period 1972–2010. There were three results
obtained from the analysis. The results showed a
positive long-term relationship between disaggregate
energy consumption and industrial output. On the
other hand, evidence of a unidirectional causality
was observed running from electricity consumption
to industrial output.
In another study, Husain and Lean (2015) used
a demand function to investigate the relationship
between electricity consumption, output, and price
in Malaysia using time series data for the period
1978–2011. In the long term, electricity consumption,
output, and price were found to be cointegrated. Evidence of a positive relationship was found between
electricity consumption and manufactured output.
In the same vein, Alley (2016) investigated the
effect of electricity supply, industrialization, and economic growth using evidence from Nigeria. The
results showed that electricity supply had a significant
positive effect on industrial output. This result contrasts with that of the OLS. Likewise, Olarinde and
Omojolaibi (2014) examined electricity consumption,
institutions, and economic growth in Nigeria for the
period 1980–2011 using causality using the Autoregressive Distributed Lag (ARDL) model and WALD
test approach, and found a positive direct relationship between institutions, electricity consumption, and
economic growth.
In another study, Sun and Anwar (2015) in Singapore used a trivariate vector autoregressive framework
and found a positive relationship between electricity
consumption and industrial production using monthly
data from 1983 to 2014. Mawejje and Mawejje (2016)
in Uganda found a long-term causality running from
electricity consumption to industry at the sectoral level
(macro level); a unidirectional short-term causality
running from the services sector to electricity consumption; and neutrality in the agricultural sector
using a vector error correction techniques.
Ettah , (2017) also found that Nigeria’s national output is significantly and consistently improved by an
available and sustained electricity supply, using the
vector error correction mechanism (VECM). Similarly, Alby, Dethier, and Straub (2013) found that in
developing countries with a high frequency of power
outages, electricity-intensive sectors have a low proportion of small firms since only large firms are able to
invest in generators to mitigate the effects of outages.
On the other hand, Akekere and Yousou (2013)
found that electricity supply negatively affects industrial output in Nigeria. Nwajinka et al. (2013) also
found that that other things being equal electricity supply has no significant impact on industrial productivity
in Nigeria.
Whereas empirical findings on the nexus differ
across countries, the differences may be due to the
heterogeneity of infrastructural and other characteristics between the countries. The literature documenting
86
This study’s aim was to examine the extent to how electricity access influences industrial output in Uganda,
while addressing the following research questions:
1) What is the effect of electricity consumption on
industrial output?
2) What is the effect of education on industrial output?
3) How does labour employed influence industrial
output.
To answer the above research questions, the study
used a Cointegrated Vector Autoregression (CVAR) as
an estimation technique. Secondary data (time series
data) on the inputs (electricity consumption, labour,
education, and industrial output) were extracted from
the industries for a period of 10 years (2009–2019).
Additionally, we propose relevant policy recommendations.
1.4 Theoretical framework
In analyses of the role played by electricity in production process, growth models (exogenous growth
models such as the Solow’s growth model, the RamseyCass-Koopman’s model, the Diamond overlapping
generation model, as well as the endogenous growth
models) have been extended. Besides capital, labour,
and knowledge (technology), electricity has been
explicitly modelled as an argument for the economic
output (Alley et al. 2016; Akekere & Yousuo 2013;
Danish et al. 2015; Maxim 2014; Palit & Bandyopadhyay 2016; Yuan, Kang et al. 2008). This research has
implications for economic theory. The output elasticity
of electricity is nearly triple its share of inputs in production. The marginal revenue product of electricity is
nearly triple the marginal revenue products of labour
and capital inputs at equilibrium. Electricity consumption, akin to R&D, has a much larger role in industrial
output than postulated in production theory. Differences in the output elasticities between labour and
the labour employed functions raise additional debates
about the adequacy of the human capital theory in
explaining industrial output.
2 REVIEW OF LITERATURE
Most studies on the effects of electricity as a component of energy on output and economic growth
frequently assume that energy enters the production
function independently or in a Hicks-neutral fashion (see Akekere & Yousou 2013; Samuel & Lionel
2013; Stern 2004; Yuan et al. 2008). Limited attention has been paid to which energy may enter the
production functions. Given the fact that most industrial machines depend on electricity to perform, the
capital-augmenting factor of energy performance in a
production model appears to offer a promising way of
appraising the role of electricity on industrial output.
Studies on industrial electricity access have attracted
a lot of information, however studies are limited with
deviating evidence, and there has not been a consensus
over the effects of electricity on industrial output.
In Pakistan, Qazi et al. (2012) used the Johansen
cointegration approach based on VAR to conduct a
study on the relationship between disaggregate energy
consumption and industrial output. The study covered the period 1972–2010. There were three results
obtained from the analysis. The results showed a
positive long-term relationship between disaggregate
energy consumption and industrial output. On the
other hand, evidence of a unidirectional causality
was observed running from electricity consumption
to industrial output.
In another study, Husain and Lean (2015) used
a demand function to investigate the relationship
between electricity consumption, output, and price
in Malaysia using time series data for the period
1978–2011. In the long term, electricity consumption,
output, and price were found to be cointegrated. Evidence of a positive relationship was found between
electricity consumption and manufactured output.
In the same vein, Alley (2016) investigated the
effect of electricity supply, industrialization, and economic growth using evidence from Nigeria. The
results showed that electricity supply had a significant
positive effect on industrial output. This result contrasts with that of the OLS. Likewise, Olarinde and
Omojolaibi (2014) examined electricity consumption,
institutions, and economic growth in Nigeria for the
period 1980–2011 using causality using the Autoregressive Distributed Lag (ARDL) model and WALD
test approach, and found a positive direct relationship between institutions, electricity consumption, and
economic growth.
In another study, Sun and Anwar (2015) in Singapore used a trivariate vector autoregressive framework
and found a positive relationship between electricity
consumption and industrial production using monthly
data from 1983 to 2014. Mawejje and Mawejje (2016)
in Uganda found a long-term causality running from
electricity consumption to industry at the sectoral level
(macro level); a unidirectional short-term causality
running from the services sector to electricity consumption; and neutrality in the agricultural sector
using a vector error correction techniques.
Ettah , (2017) also found that Nigeria’s national output is significantly and consistently improved by an
available and sustained electricity supply, using the
vector error correction mechanism (VECM). Similarly, Alby, Dethier, and Straub (2013) found that in
developing countries with a high frequency of power
outages, electricity-intensive sectors have a low proportion of small firms since only large firms are able to
invest in generators to mitigate the effects of outages.
On the other hand, Akekere and Yousou (2013)
found that electricity supply negatively affects industrial output in Nigeria. Nwajinka et al. (2013) also
found that that other things being equal electricity supply has no significant impact on industrial productivity
in Nigeria.
Whereas empirical findings on the nexus differ
across countries, the differences may be due to the
heterogeneity of infrastructural and other characteristics between the countries. The literature documenting
86