Do Institutions Moderate the Energy-Growth Aliation? Evidence from Sub Sahara Africa Countries

: This paper aims to look into the role of institutional quality in regulating energy and growth affiliation. 3 The countries of Sub-Saharan Africa (SSA) are studied from 1990 to 2019. CSD and SH tests were used to verify 4 cross-sectional dependency and slope homogeneity properties. CIPS and CADF were used to investigate 5 stationarity features. The Westerlund bootstrap cointegration test was used to analyze the long-tenure equilibrium 6 affiliation among the variables and confirm cointegration in the extended period. To examine the long-short term 7 performance between the variables, the CS-ARDL approach is used. To analyze the flow of causation, the study 8 used the DH causality process. The findings reveal that energy has a negative and significant impact on growth. 9 In both terms, industrialization and population have a negative and positive impact on growth, respectively. The 10 DH heterogenous causality study reveals the mixed effect, i.e. one-way causal associations between growth and 11 institutional quality, two-way causal associations between energy and population, and no causation with 12 industrialization. Furthermore, institutional quality as a moderating variable harms growth. To achieve long- 13 period growth, states should expand investment in renewable energy sectors, create well-resourced institutions, 14 and plan for renewable energy development, according to this empirical research.

study, which intends to look into the possibilities of using energy to anticipate US GDP growth. The study's 104 context is enhanced because the United States is a highly industrialized economy with high energy usage. The empirical review identifies a subset of empirical literature separated according to techniques and focuses on 106 renewable or non-renewable energy sources. (Narayan & Doytch, 2017) examines renewable and non-renewable 107 energy sources to show that renewables support the neutrality hypothesis, whereas non-renewables fully endorse 108 the input, growth, and conservative notions.

134
There is a single route link between GDP growth and renewable energy, according to research of (Zafar 135 et al., 2019) results. The findings are also supported by (Ocal & Aslan, 2013) and (Xu, 2016

163
The study aims to provide empirical results based on the above hypotheses to investigate the moderating 164 effects of institutions on the energy-growth relationship. This study assumes that growth is a function of energy 165 consumption in workable institutions, growing industrialization, and population factors. The empirical model is 166 estimated using panel data methodology to see a systematic relationship between institutional performance and 167 the degree of energy and growth. Panel data methods allow researchers to account for individual heterogeneity 168 and eliminate the possibility of misleading findings. The research uses CSD, SH, Friedman, CS-ARDL, and DH 169 causality methods for the empirical model to discuss possible country-specific unobserved heterogeneity: Where β0 infers an unseen time-invariant discrete effect; β1, and β4, separately, capture the moderating 172 effect of IQ indicators on the energy-growth link with some control variables thus industrialization and population; 173 εit is the error sign, which is usually considered to have an average of zero (0) and variance of σ2; I is the number 174 of states I = (1, 2, 3..., N); and t is the time setting (t = 1, 2, 3..., T) 175

176
Cross-sectional reliance was used to analyze sectional characteristics, slope homogeneity, and the 177 Friedman test to determine the presence of heterogeneity along the sequence, among other econometric 178 approaches used in the study. To discuss the long-tenure alliance and causation, the Westerlund and Edgerton

182
The study checks for the existence of CSD before estimating the cointegrating relationship between 183 growth and its fundamentals. In the case of countries from the same geographical area, this problem can be 184 particularly acute. Shocks may also be transmitted between countries with similar economic systems, resulting in   198 where 3ij is a sample evaluation of the remaining sets that are correlative 199 3ij= 3ij=

200
And 3it is the dimension of uit in Eq (4). LM is asymptotically distributed as on the presumption of null interest. 201 χ2 with N (N-1)/2 DF. However, when N is large and T is small, this test is prone to substantial size biases, 202 expected in systematic applications. This is because the LM figures are not suitably concerned with finite T, and 203 when N is large, the bias will probably worsen. A counter-proposal was made, which was as follows:

238
Due to substantial CSD, the dynamics of the economic development process in any country may be comparable.

239
As a result, N is chosen relative to T to gauge the slope's homogeneity. (Swamy, 1970) gives a basis for the 249 The bias-adjusted type of the Δ̃ exam can be modelled in the following way: 252 where E (zĩT) =k and Var (zĩT) = (2k (T -k-1)) ⁄ (T +1) indicate the mean and variance, respectively.

254
Traditional panel root unit tests like ADF are unsuccessful when the sequence is CSD and heterogeneous.

278
Where dt stands for deterministic elements while pi and qi stand for lag lengths and lead orders vary across CSs.

279
The two group-average test statistics Gt and Ga, as well as the two-panel test statistics Pt and Pa, can be described

333
The following is the null declaration and alternative assumption for the calculated panel statistics     shock, it will spread to the others. Table 1 also displays SH tests conducted using (Swamy, 1970)

374
The failure of the homogeneous slope hypothesis suggests that enforcing the slope homogeneity 375 limitation will lead to erroneous results.

379
The analysis verifies the nonstationary property and order of integration of the variables as a first step.

380
Countries have different features, and the panels may contain CSD, resulting in inaccurate and unfair results.

381
(Pesaran, 2007) proposed two-unit root tests for dealing with CD ambiguity: IPS cross-sectional (CIPS) and the null proposition is dismissed at 1% since the variables reflect the stationary effects at first difference. As a 385 result, the analyses for CIPS and CADF are close. In conclusion, the results show that the order of integration of 386 the individual series is either I(0) or I(1), but not I(2) and that there is significant error CSD, indicating that the 387 Panel CS-ARDL(p, q) method is sufficient.     Table 5, we can see that, although the coefficient of IND retains the negative sign implied 416 by economic theory, it is not essential. This could be because the financial systems of many SSA countries have 417 been weakened by large structural fiscal deficits and volatile monetary and exchange rate policies.

418
Furthermore, energy stimulates growth, and affiliation is incompatible when moderating with 419 institutional quality. This means that a 1% rise in energy demand would result in a 1.9116% reduction in growth.

420
Energy's negative impact on growth reflects the risk of inefficient and wasteful energy use at the regional level.

421
From an economic standpoint, this result is counter-intuitive since the coefficient of the energy variable is negative, 422 suggesting that further energy delays growth. However, as previously mentioned, several factors are at play, 423 including inefficient energy usage, which SSA countries are attempting to address. As a result, energy 424 conservation should be used to save energy and reduce pollutant emissions. This empirical evidence backs up proof contradicts the observational evidence discovered by (Omri, 2013).

427
In the short and long term, population density can be seen to have a positive and substantial impact on

469
The findings show that there is a two-route causal link between population and growth. On the one side, 470 when the difference in time and between regions is considered, the result shows that growth positively impacts 471 the population. This finding is critical for decision-makers at the regional level as they develop policies to revive 472 the population. On the other hand, the degree of growth was negatively affected by population as a dependent 473 variable, indicating a situation similar to the post-revolutionary SSA society's transition era, when population 474 decline could have favoured growth. The established causality relationships need a more nuanced approach to 475 understand further how the relationships between the two variables stand, as seen above. Studies from 476 (Mahmoudinia et al., 2020) and (Jemna, 2015) confirmed this hypothesis, indicating that population and growth 477 are causally mutual. (Jemna, 2015) aim to illustrate causality between growth and fertility by using the VAR 478 methodology and the approach. The empirical discoveries express a bidimensional causality relationship between 479 fertility and growth and that each variable's invention has a long-term effect on the other. (Mahmoudinia et al., 480 2020) stated that when GDP growth and capital stock are dependent variables, panel cointegration and causality techniques indicate a long-run relationship. In the long term, population growth has a positive and statistically significant effect on growth. Also, for OIC countries, the short-run bidirectional relationship between population and growth has been agreed upon. Theoretically, population growth is a national saving that expands the 484 economy's potential. Given these considerations, it can be argued that population growth is a catalyst for growth 485 rather than a hindrance. As a result of this negative view of the population and its power, economic policymakers 486 must eliminate significant economic barriers by reforming systems, improving management capacity, and 487 enacting appropriate monetary and fiscal policies.

488
In another scenario, the findings revealed a non-causal relationship between GR and IND. As a result, 489 neither growth nor industrialization seems to be mutually exclusive. A more significant services sector than a 490 manufacturing sector, on the other hand, does not refute Kaldor's arguments about the latter's growth-enhancing 491 properties; instead, it is a product of a maturing economy (Kaldor, 1966). Manufacturing productivity, on the 492 other hand, has been questioned as a source of growth. (Timmer & De Vries, 2009;Vries et al., 2014) show that 493 recent growth acceleration episodes in developed countries were driven by productivity improvements in the 494 services sector, rather than manufacturing, by implementing a new system of productivity accountability. On the 495 other hand, the scientists argue that there is no causal link between these results and growth.

496
Finally, on the causal front, the findings show a one-way causal connexion between overall institutional efficiency and growth. The causality runs from institutional quality to growth, suggesting that higher institutional 498 quality contributes to higher growth, but higher economic progress does not always lead to higher institutional 499 quality. This contradicts the common perception that institutions and growth have a two-way relationship. Better 500 institutions contribute to more significant growth, and more remarkable growth necessitates the creation of higher- test indicate that there is unidirectional causality between institutional efficiency and growth. As a result, they 505 concluded that improving institutional quality in these developing countries is essential to ensure high economic 506 growth. (Kilishi et al., 2013) conducted an empirical study in Sub-Saharan Africa to determine if institutions 507 matter for regional growth and, if so, which ones matter the most. Their findings show that institutions matter for 508 economic activities, with regulatory quality appearing the most relevant. They suggest that improving regulatory 509 quality could improve the region's economic performance.

510
Furthermore, (Siddiqui & Ahmed, 2009) used the extended tenure technique and the causality test to 511 investigate the affiliation between institutional efficiency and economic performance. The results of their 512 cointegration test show that institutional efficiency and growth have a long-term relationship. Furthermore, the 513 findings of their causality test reveal that the causality between institutional quality and growth is unidirectional, 514 with the causality flowing from institutional quality to growth.

525
there is insufficient evidence to suggest that institutional efficiency influences economic growth.

526
Industrialization has a negative and negligible relationship with growth; the only variable that positively 527 affects growth in both terms is population.
The "feedback hypothesis" suggests a two-way causality between energy and growth, as well as a one-529 way causal route between growth and institutional quality, with the flow beginning from institutional 530 quality to growth for the significant variables of interest. There is a "neutrality hypothesis" for the control 531 variables; according to this study, there is no statistically significant causation between growth and 532 industrialization, although there is a two-way causality with population.

533
The empirical findings also indicate the following consequences for public policy:

534
I. An interregional energy production plan should be established quickly and include a fixed rate of energy 535 increase that corresponds to SSA countries' economic growth rate.

536
II. SSA countries should focus on developing energy conservation awareness among individuals and 537 businesses and investing in new renewable and sustainable energy sources.
automatic on/off operations) is one option for enhancing renewable and sustainable national energy output in SSA countries.
institutions function independently.