The role of development and financial risk in agricultural energy consumption: an empirical analysis based on country panel data and smooth transition regression

O papel do desenvolvimento e do risco financeiro no consumo de energia agrícola: uma análise empírica com base em dados de painel de países e regressão de transição suave

Abstract

Financial development facilitates farmers' access to credit resources, enabling the adoption of innovative technologies for energy management. On the other hand, long-term planning for agricultural businesses is expected to reduce various types of risks, including financial risk, thus promoting production. Therefore, the aim of this study is to empirically investigate the relationship between development and financial risk with energy consumption in the agricultural sector in Iran and selected developing countries. For this purpose, financial development indicators such as the ratio of deposits and bank facilities to agricultural value-added, as well as economic, financial, and political risk indicators, have been used for two groups of oil-exporting and non-oil-exporting developing countries during the years 2010 to 2020. In order to analyze the data, the Smooth Transition Regression model was used, which allows for examining non-linear relationships between variables. The findings suggest significant differences in the average per capita agricultural energy consumption between the studied oil-exporting and non-oil-exporting countries. Other research results indicate the superiority of non-linear specification relationships between variables and their threshold behavior. Financial development has an inverse effect on per capita energy consumption, and an increase in various types of risk (financial, economic, and political) is associated with an increase in per capita energy consumption in the agricultural sector. The magnitude of the impact of variables in non-oil-exporting countries is generally estimated to be higher than in oil-exporting countries.

Keywords:
energy consumption; financial development; smooth transition regression; agriculture

Resumo

O desenvolvimento financeiro facilita o acesso dos agricultores a recursos de crédito, permitindo a adoção de tecnologias inovadoras para gestão de energia. Por outro lado, espera-se que o planejamento de longo prazo para negócios agrícolas reduza vários tipos de riscos, incluindo o risco financeiro, promovendo assim a produção. Portanto, o objetivo deste estudo é investigar empiricamente a relação entre desenvolvimento e risco financeiro com o consumo de energia no setor agrícola no Irã e em países em desenvolvimento selecionados. Para esse propósito, indicadores de desenvolvimento financeiro, como a proporção de depósitos e facilidades bancárias para valor agregado agrícola, bem como indicadores de risco econômico, financeiro e político, foram usados para dois grupos de países em desenvolvimento exportadores e não exportadores de petróleo durante os anos de 2010 a 2020. Para analisar os dados, foi usado o modelo de Regressão de Transição Suave, que permite examinar relações não lineares entre variáveis. Os resultados sugerem diferenças significativas no consumo médio de energia agrícola per capita entre os países exportadores e não exportadores de petróleo estudados. Outros resultados de pesquisas indicam a superioridade das relações de especificação não lineares entre variáveis e seu comportamento limite. O desenvolvimento financeiro tem um efeito inverso no consumo de energia per capita, e um aumento em vários tipos de risco (financeiro, econômico e político) está associado a um aumento no consumo de energia per capita no setor agrícola. A magnitude do impacto das variáveis nos países não exportadores de petróleo é geralmente estimada como sendo maior do que nos países exportadores de petróleo.

Palavras-chave:
consumo de energia; desenvolvimento financeiro; regressão de transição suave; agricultura

1. Introduction

Agriculture, as one of the most important economic sectors, holds a special place in the economies of all countries, especially those in development. The agricultural sector contributes to sustainable development and poverty reduction in rural areas by creating job opportunities. The shortage of financial resources and lack of access to credit have always been among the most significant challenges in the development of the agricultural sector in developing countries. Risk or uncertainty also plays an undeniable role in the growth and development of economic activities, including agriculture (Ahangari and Kamranpour, 2016).

On the other hand, the emphasis on improving energy efficiency in economic activities, considering the scarcity of this valuable resource and input, is highlighted in national and international documents as well as one of the Sustainable Development Goals (SDGs). Research indicates that the relationship between financial development and energy consumption in the agricultural sector in developing countries is complex and depends on various factors. On one hand, financial development can significantly enhance energy infrastructure and energy resources in the agricultural sector. With increased access to financial resources, agricultural organizations can easily secure the necessary funds for improving infrastructure and energy supply. Furthermore, financial development can contribute to the advancement of energy-related technologies, leading to increased efficiency and reduced energy consumption in the agricultural sector. From another perspective, financial development, due to increased access to credits and financial resources, can lead to the expansion of agricultural businesses (horizontally, vertically, or both), and this can result in increased energy consumption in the agricultural sector (Sadorsky, 2010; Ahangari and Kamranpour, 2016). The relationship between financial development and energy consumption can be influenced by various political and economic risks. Democracy and political stability have a significant impact on financial development (Girma and Shortland, 2008).

The relationship between financial development and energy consumption has been addressed in various forms in previous studies. Ansari and Yakani (2013) examined the impact of financial market development on the growth of the agricultural sector in Iran during the period 1386-1346 using the Autoregressive Distributed Lag (ARDL) model. The study's results indicated that the stock market has a negative effect on the growth of the agricultural sector, while this effect was reported as positive for the money market.

Ahangari and Kamranpour (2016) investigated the influence of financial development and value-added on energy consumption in the industrial and agricultural sectors of Iran during the period 1976-2013 using the Bounds Testing Approach (ARDL bounds test). The findings suggest that both in the short term and the long term, in both sectors, financial development and value-added growth lead to an increase in energy consumption. In fact, the results indicate a positive correlation between financial development and growth in both the industrial and agricultural sectors.

Shahbaz and colleagues (Shahbaz et al., 2013) examined the impact of financial market development on the growth of the agricultural sector in Pakistan during the years 1971-2011 using modern econometric time series models. Their study results indicate a positive effect of financial market development on the growth of the agricultural sector.

Hayakawa and colleagues (Hayakawa et al., 2013) investigated the relationship between financial and political risk and foreign direct investment (FDI) in ninety developing countries, including sixty countries during the period 1985-2017. Their research findings showed that international investors pay attention to financial and political risk as an important indicator in investment returns. The initial level of political risk in countries does not adequately explain foreign direct investment, but it is the change in political risk that leads to an increase or decrease in the level of foreign direct investment. On the other hand, changes in gross domestic product, not its absolute amount, attract foreign direct investment.

Vijaykumar and colleagues (Vijaykumar et al., 2018) conducted a study to estimate the impact of risk on capital attraction in the agricultural sector. In this research, they utilized national panel data for a 30-year period ending in 2015. The results of this study indicate that the country risk ranking had a significant effect on investment in agricultural activities.

In a brief summary of the research background, it can be claimed that in previous studies, the impact of financial development and risk on economic variables such as investment and energy consumption has mostly been examined in linear relationships. Therefore, the current research can be considered as a novel approach, both in terms of the applied model and the studied sample and data used.

2. Theoretical Background and Research Methodology

Considering the objective of the present study, which is to examine the relationship between development and financial risk on energy consumption in the agricultural sector, the theoretical foundations of the subject and the econometric model used will be introduced below.

2.1. Financial development and energy consumption

As Sadorsky (2011) points out, financial development may have different effects on energy consumption through various pathways. Having more financial resources provides the opportunity for business development and expansion, which in turn leads to an increase in demand for productive inputs, including energy. On the other hand, financial development can facilitate the adoption of innovative technologies, which are expensive but reduce energy consumption. It should be noted that financial development can lead to investments in renewable energy resources, essentially creating a substitution effect in energy sources. This recent effect should be interpreted in the context of sustainable energy resource utilization and environmental preservation.

2.2. Risk and energy consumption

The presence of risk, of any kind, be it financial, economic, or political, acts as a barrier to long-term planning and policymaking by producers and economic actors, often leading them to focus on short-term matters. Therefore, one can expect a reduction in investment in new technologies and ultimately an increase in energy consumption. Shahbaz et al. (2013) have pointed out that a stable financial environment enhances the performance of financial institutions and fosters financial innovations. Consequently, easier access to financial resources becomes feasible, naturally allowing for investment in energy-efficient technologies. Stable economic and financial environments can assist productive sectors in accessing the stock market and banking system, significantly influencing energy consumption. Since the protection of property rights and investment security is closely related to political stability, the political risk of a country can also have a significant impact on financial development and energy consumption.

2.3. Econometric model

Based on the study by Chu (2021) and González et al. (2005) and the theoretical foundations discussed, in this study, a Panel Smooth Transition Regression (PSTR) model is specified as follows (where L represents the natural logarithm):

$$\begin{array}{l}{\text{LAEC}}_{\text{it}}={\text{a}}_{\text{i}}+{\text{b}}_1{\text{LFD}}_{\text{it}}+{\text{c}}_1{\text{LRISK}}_{\text{it}}+{\text{d}}_1{\text{LCPI}}_{\text{it}}+{\text{e}}_1{\text{LGDP}}_{\text{it}}+{\text{f}}_1{\text{LUP}}_{\text{it}}+\\ \left({\text{b}}_2{\text{LFD}}_{\text{it}}+{\text{c}}_2{\text{LRISK}}_{\text{it}}+{\text{d}}_2{\text{LCPI}}_{\text{it}}+{\text{e}}_2{\text{LGDP}}_{\text{it}}+{\text{f}}_2{\text{LUP}}_{\text{it}}\right)\text{*g}\left({\text{LRISK}}_{\text{it}};\gamma ,\theta \right)+{\text{e}}_{\text{it}}\end{array}$$ (1)

All variables and their data sources are introduced in Table 1. It should be noted that to avoid multicollinearity between two financial development indices (BD and BL), these two variables were replaced with a single component, FD, using Principal Component Analysis (PCA). The same approach was applied to the three desired risk indices, and the variable RISK was incorporated into the model. In addition to the two focal variables (financial development and risk), three other variables, including the Consumer Price Index, Gross Domestic Product per capita, and the urban population percentage, were also included in the model based on previous studies. An increase in the Consumer Price Index implies the presence of inflationary conditions, which, considering the interrelation of agriculture with other economic sectors, might lead to an increase or decrease in demand for agricultural products and, consequently, agricultural inputs (including energy). An increase in Gross Domestic Product per capita is interpreted as an improvement in the overall income level, thereby increasing households' demand for goods and, consequently, inputs. Regarding the other variable, the urban population percentage, it is expected that an increase in the urban population share would lead to increased demand for foodstuffs and, naturally, an increased demand for agricultural inputs.

Table 1
Definition of research variables.

Variable	Symbol	Type	Definition			Source
Agricultural energy consumption per capita	AEC	Dependent		The ratio of agricultural energy consumption (kilogram of crude oil equivalent) to agricultural workers				World Bank
The ratio of the value of bank deposits to the value-added of agriculture	BD	Independent		The ratio of the total value of bank deposits to the value-added of agriculture (unit free)					World Bank
The ratio of the value of granted facilities to agricultural activities to the value-added of agriculture	BL	Independent		The ratio of the total value of facilities granted to agricultural activities to the value-added of agriculture (unit free)					World Bank
Economic risk index	ER	Independent		In the range of zero to fifty					ICRG*
Urban Population	UP	Independent		The percentage of urban people					WDI
Financial risk index	FR	Independent		In the range of zero to fifty					ICRG
Political risk index	PR	Independent		In the range of zero to hundred					ICRG
Consumer price index	CPI	Instrumental variable		Base year 2000					WDI**
GDP per capita	GDP	Instrumental variable		in USD					WDI

• ^*
  International Country Risk Guide;
• ^**
  World Development Indicators.

In Equation 1, the function g(.) represents the transfer function that indicates how the transition from one regime to another occurs based on the threshold variable (RISK). This function is continuous and bounded between zero and one.

The general form of the above function with a logistic specification is as follows:

$g\left(LRIS{K}_{it};\gamma ,\theta \right)={\left[1+\exp \left(-\gamma \left(LRIS{K}_{it}-\theta \right)\right)\right]}^{-1}$ (2)

In this equation, the parameter γ represents the slope of the transfer function and indicates the speed of transition from regime one to regime two. The parameter θ is the threshold value.

2.4. Research data

Given the significant volatility of the global oil market and the role of oil revenues in the economies of oil-producing countries (such as Iran), the present study seeks to examine the relationship between the mentioned variables for two groups of oil-developing countries (oil-producing and oil-exporting) and non-oil countries. Considering the subsidy policies common in oil-producing countries for fuel carriers, it is expected that the type and intensity of the relationship between variables in these two groups will differ significantly. This classification of countries can be considered as part of the innovation of the current research.

Based on this, and considering the availability of the required data, two groups of developing countries were selected for the study: oil-developing countries, including Iran, Bahrain, Egypt, Kuwait, Malaysia, Nigeria, Oman, Qatar, Saudi Arabia, and Tunisia, and non-oil countries, consisting of Armenia, Croatia, Czech Republic, Estonia, Poland, Portugal, Romania, Serbia, Slovakia, and Ukraine. The research data covers the annual time series from 2010 to 2020.

3. Results and Discussion

3.1. Descriptive statistics

Descriptive statistics for the research variables have been reported for the two groups of countries under study in Tables 2 and 3.

Table 2
Descriptive statistics of research variables (oil countries).

Variable	UP (%)	GDP (million USD)	CPI	PR	FR	ER	BL	BD	AEC (million barrel oil equivalent)
Average	73	6520	125	51	29	24	24	31	453
Standard deviation	0.86	0.92	0.89	0.77	0.62	0.83	0.75	0.51	0.76
Maximum	95	8511	153	72	40	43	45	69	712
Minimum	53	2165	42	10	11	9	11	23	362

Table 3
Descriptive statistics of research variables (non-oil countries).

Variable	UP	GDP	CPI	PR	FR	ER	BL	BD	AEC
Average	56	4520	136	62	35	32	41	42	274
Standard deviation	0.72	0.82	0.82	0.82	0.74	0.54	0.96	0.43	0.53
Maximum	71	5620	163	79	43	40	66	56	524
Minimum	46	2141	56	23	15	16	19	27	212

By comparing the two tables above, it can be easily observed that the average per capita energy consumption in the agricultural sector in the studied oil-producing countries is significantly higher than that in non-oil countries. The reason for this can be attributed to two factors, including cheap energy in these countries and the relatively small number of agricultural workers, especially in the countries around the Persian Gulf. The banking financial development indices, including the deposit-to-GDP ratio and agricultural value-added-to-loans ratio, are higher in non-oil countries. The levels of risk indices are also higher in non-oil countries, indicating greater stability in the economic, financial, and political environments of this group of countries.

3.2. Inferential statistics

As previously mentioned, in this study, based on research conducted by Beck and Levine (2002), Jalil and Feridun (2010), Kim and Park (2016), and Ouyang and Li (2018) to avoid the problem of multicollinearity between independent variables, the Principal Component Analysis (PCA) method was used to transform the two desired financial development indices (BD, BL) into one component. The obtained results are shown in Table 4.

Table 4
The results of the principal component analysis for financial development variables.

Major components	Characteristic root	Variance percentage	Cumulative variance percentage
First component	4.52	91.25	91.25
Second component	0.64	8.75	100
		PC1
Variable		BD	BL
Factor		0.54	0.48

Based on the above table, two components can be presented for the two discussed financial development variables. The first component alone contains 91.25% of the total variation of the two variables, while this figure is 8.75% for the second component. Therefore, the new variable, namely PC₁, can be used as a suitable and representative substitute for the two financial development variables in the regression model. In addition, the numerical information of the new variable is calculated based on the data of BD and BL and with coefficients of 0.54 and 0.48. The calculated variable is represented by FD in Equation 1.

Given the nature of the research data, which is of the panel type, it is necessary to test the stationarity of the research variables before estimating the regression model. For this purpose, three tests of the first generation of panel data unit root tests, including Fisher-DF, Levin-Lin-Chu, and Im-Pesaran-Shin tests, were used. Then, due to the possibility of correlation among variables across time periods (countries) and their nonlinear behavior, cross-sectional dependence tests such as the Pesaran’s Cross-sectional Dependence (CSD) test and the Likelihood Ratio (LR) test were used. Due to the confirmation of cross-sectional dependence and nonlinear behavior, the second-generation panel unit root tests were also used to improve accuracy. The obtained results can be seen in Tables 5, 6, and 7.

Table 5
CSD Test Results.

Variable	LUP	LGDP	LCPI	LPR	LFR	LER	LBL	LBD	LAE
Statistics	153.19	112.89	107.48	139.11	145.36	109.59	134.25	101.12	87.25
P-Value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

• Note: The letter L at the beginning of the variables symbol denotes natural logarithm, the null hypothesis of the test is absence of correlation.

Table 6
The results of testing the nonlinear behavior of variables.

Variable	LUP	LGDP	LCPI	LPR	LFR	LER	LBL	LBD	LAEC
Statistics	52.43	61.35	26.49	19.81	97.11	42.71	36.41	12.24	7.49
P-Value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

• Note: The letter L at the beginning of the symbol of variables denotes natural logarithm. The null hypothesis or the test is the linear behavior of variables.

Table 7
The results of unit root tests (first and second generations).

KSS	CADF	CIPS	IPS	LLC	Fisher-ADF
-11.12*	-7.24*	-4.23*	-23.45*	-15.21*	64.25*	LAEC
-9.24*	-4.45*	-3.02*	-19.25*	-23.41*	54.12*	LBD
-8.43*	-3.62*	-7.24*	-24.46*	-18.22*	36.81*	LBL
-6.42*	-5.94*	-5.12*	-32.87*	-45.71*	42.64*	LER
-6.35*	-5.87*	-4.08*	-26.49*	-26.39*	59.88*	LFR
-7.88*	-4.65*	-3.11*	-34.55*	-21.02*	24.11*	LPR
-5.56*	-4.75*	-4.25*	-26.38*	-1.12	55.44*	LCPI
-4.73*	-1.25	-5.47*	-29.45*	-19.76*	62.73*	LGDP
-1.11	-6.63*	-4.33*	-17.43*	-23.12*	91.49*	LUP

• Note: The letter L at the beginning of the symbol of variables denotes natural logarithm; The null hypothesis of all the tests is non-stationarity of the variables.
• ^*
  Significant at the one percent level.

Tables 5 and 6 indicate the presence of cross-sectional (country-level) correlation and nonlinear behavior of the variables in the study period. For this reason, the results of the first-generation panel data unit root tests alone cannot be relied upon, and it is necessary to use second-generation tests alongside them. Therefore, both first and second-generation tests were used in the present study, and the results obtained are reported in Table 7. It is observed that, except for two cases, all other computed statistics are significant at the 1% significance level. Therefore, the null hypothesis of stationarity or the presence of a unit root in the variables is not acceptable.

In the next step, the superiority of the nonlinear model over the linear one was considered when specifying the regression model. Assuming that risk indices can have a threshold effect and justify nonlinear specification, the necessary tests were conducted. In Table 8, three statistics: LM, LMF, and LR, which have been used in previous studies, were employed as judgment criteria for choosing between linear and nonlinear specifications. These statistics were separately calculated for two groups of countries and presented in the table. In the mentioned table, all computed statistics are significant at the 1% significance level. Therefore, the null hypothesis of linear specification of relationships between variables is rejected in favor of nonlinear specification. In other words, the economic, financial, and political risk indices practically create a break in the regression line and, therefore, it is not possible to explain the relationships between variables for the entire time period based on a single regression.

Table 8
The results of tests of superiority of non-linear model.

Threshold variable	LER	LFR	LPR
Model number	1	2	3
Oil countries LM	271.56*	255.02*	263.44*
${\text{LM}}_{\text{f}}$	32.24*	29.49*	24.22*
LR	311.12*	259.27*	325.89*
Non-oil countries LM	45.56*	35.11*	29.08*
${\text{LM}}_{\text{f}}$	14.12*	9.97*	8.44*
LR	67.77*	55.82*	47.36*

• Note:The null hypothesis is the superiority of the linear model and the alternative hypothesis is the existence of a transfer function (the superiority of the non-linear model).
• ^*
  Significant at the one percent level.

Considering the above points and as mentioned in the previous section, the panel data regression model based on smooth transition data, which is commonly used in similar studies, was considered for specifying and estimating the relationships between variables. The results of estimating the discussed regression model (Equation 1) for two groups of countries are reported in Tables 9 and 10.

Table 9
The results of the estimation of panel smooth transfer regression (PSTR) for oil countries.

Threshold variable	LER	LFR	LPR
Model Number	1	2	3
b1	-0.012**	0.035**	-0.063***
c1	0.125***	0.856***	0.025*
d1	-0.008**	-0.048**	-0.011***
e1	0.423**	0.712***	0.413**
f1	0.365**	-0.174**	0.212**
b2	-0.045*	-0.152***	0.021*
c2	-0.076***	-0.195**	0.185***
d2	0.145**	0.014*	-0.102**
e2	-0.187**	0.047**	-0.025***
f2	-0.057*	0.041	-0.112**
$\theta$	3.561	3.176	4.532
$\gamma$	29.653	54.239	42.187

• ^*
  10%;
• ^**
  5%;
• ^***
  1%.

Table 10
The results of the estimation of panel smooth transfer regression (PSTR) for non-oil countries.

Threshold variable	LER	LFR	LPR
Model Number	1	2	3
b1	-0.025**	0.044**	-0.142**
c1	0.152**	0.741**	0.087*
d1	-0.025**	-0.114**	-0.069***
e1	0.512**	0.682**	0.506**
f1	0.311***	-0.219**	0.293**
b2	-0.053**	-0.163**	0.067**
c2	-0.08**	-0.163**	0.123***
d2	0.163**	0.052*	-0.125**
e2	-0.113**	0.118***	-0.152***
f2	-0.098**	0.164	-0.146**
$\theta$	4.125	3.938	5.326
$\gamma$	35.265	41.571	56.289

• ^*
  10%;
• ^**
  5%;
• ^***
  1%.

In the above table, the regression coefficients with index 1 refer to regime 1, and the total coefficients with indices 1 and 2 indicate regime 2. For example, coefficient $b_1$ represents the impact of financial development on per capita agricultural energy consumption in regime 1, but the sum of $b_1$ and $b_2$ represents the impact of financial development on the dependent variable in regime 2. In model number one, where the economic risk variable is considered as the threshold, firstly, the threshold value of this variable in the logarithmic form is equal to 561.3. Secondly, if LER is greater than this number (meaning moving towards greater economic stability), the rate of the financial development variable's effect on per capita energy consumption in agriculture will be -0.057. This implies that with an increase in economic stability (risk reduction), the increase in financial development indices will have a stronger negative effect on per capita energy consumption in the agricultural sector. Similarly, based on the values of coefficients $c_1$ and $c_2$, it can be claimed that an increase in risk and instability in the selected oil countries has a positive and significant effect on per capita agricultural energy consumption. However, with increased economic stability (transition to regime 2), the size of the above effect decreases because the coefficient $c_2$ is negative.

Regarding the interpretation of the other estimated coefficients, similar explanations can be provided. Based on this, the effect of the Consumer Price Index (CPI) on per capita energy consumption in regime 1 is negative. This is because as costs increase, producers will be compelled to manage the consumption of energy resources. Additionally, with an increase in per capita income and countries becoming wealthier, there is an increase in demand for food items, which can lead to an increase in per capita energy consumption in the agricultural sector. Finally, considering the estimated coefficient for the urban population percentage variable, it can be expected that with an increase in the ratio of urban households to rural households, there will be an increase in demand for agricultural products and, consequently, an increase in energy consumption.

The findings presented in Table 10 are also interpretable, just like Table 9. One point that can be inferred from comparing the results of the two groups of countries is that generally, the impact of variables is stronger in the non-oil countries group. This can be attributed to these countries' economies being less susceptible to global oil market fluctuations due to various political, economic, and other shocks. In summary, considering the results obtained from estimating the model for two groups of countries, we can point out the role of risk indices (economic, financial, and political) in the nonlinear response of per capita energy consumption in agriculture to changes in the independent variables (financial development indices and control variables).

4. Conclusion

The present research aimed to investigate the impact of development and financial risk on agricultural energy consumption in a group of developing countries. For this purpose, the Smooth Transition Regression (STR) model, which takes into account the existence of nonlinear relationships between variables, was employed using panel data. In this regard, two financial development indices were considered: the ratio of bank deposits to agricultural value-added and the ratio of lending facilities to agricultural value-added. Additionally, in the context of risk, three risk indices, including financial, economic, and political risks, were incorporated into the modeling.

Following initial investigations and the confirmation of the presence of nonlinear effects of independent variables on the dependent variable (agricultural energy consumption), the STR model, as specified in Equation 1, was estimated. The most significant findings suggest that with increased economic stability (risk reduction), the increase in financial development indices will have a more pronounced decreasing effect on per capita energy consumption in the agricultural sector. Likewise, based on the obtained results, it can be argued that an increase in risk and instability in the targeted oil-dependent countries has a positive and meaningful impact on agricultural energy consumption. However, with increased economic stability (entering the second regime), the intensity of this effect diminishes.

On the other hand, the effect of the Consumer Price Index (CPI) on per capita energy consumption in the first regime was estimated to be negative. This is because as costs increase, producers will inevitably resort to managing the use of energy resources. Furthermore, with the rise in per capita income and countries becoming wealthier, demand for foodstuffs is stimulated, which can lead to an increase in per capita energy consumption in the agricultural sector. Ultimately, it can be expected that with an increase in the proportion of urban households compared to rural households, demand for agricultural products and consequently energy consumption will increase.

Based on the research findings, it is recommended that proper allocation and direction of credit and bank facilities to the agricultural sector, strengthening stable financial policies to attract foreign investment in advanced equipment and energy-efficient technologies, expanding and consolidating sustainable economic relations with global countries, and avoiding hasty decisions and actions for changes in political and economic policies should be given due attention.

References

• AHANGARI, A. and KAMRANPOUR, S., 2016. The effect of financial development and added value on energy consumption in industrial and agricultural sectors of Iran. Quarterly Journal of Applied Economic Studies of Iran, vol. 5, no. 19, pp. 269-286. In Farsi.
• ANSARI, Y. and YAKANI, A.H., 2013. The effect of the development of financial markets on the development of the agricultural sector in Kohgiluyeh and Boyer Ahmad provinces. Economic Research and Agricultural Development of Iran, vol. 3145, pp. 529-535. In Farsi.
• BECK, T. and LEVINE, R., 2002. Industry growth and capital allocation: does having a market- or bank-based system matter? Journal of Financial Economics, vol. 64, no. 2, pp. 147-180. http://doi.org/10.1016/S0304-405X(02)00074-0
  » http://doi.org/10.1016/S0304-405X(02)00074-0
• CHU, L. K., 2021. Economic structure and environmental Kuznets curve hypothesis: new evidence from economic complexity. Applied Economics Letters, vol. 28, no. 7, pp. 612-616. http://doi.org/10.1080/13504851.2020.1767280.
• GIRMA, S. and SHORTLAND, A., 2008. The political economy of financial development. Oxford Economic Papers, vol. 60, no. 4, pp. 567-596.
• GONZÁLEZ, A., TERÄSVIRTA, T. and VAN DIJK, D., 2005. Panel smooth transition regression models Stockholm: School of Economics, SEE/EFI Working Paper Series in Economics and Finance, no. 6.
• HAYAKAWA, K., KIMURA, F. and LEE, H.H., 2013. How does country risk matter for foreign direct investment? The Developing Economies, vol. 51, no. 1, pp. 60-78. http://doi.org/10.1111/deve.12002
  » http://doi.org/10.1111/deve.12002
• JALIL, A. and FERIDUN, M., 2010. The impact of growth, energy and financial development on the environment in china: a cointegration analysis. Energy Economics, vol. 33, no. 2, pp. 284-291. http://doi.org/10.1016/j.eneco.2010.10.003
  » http://doi.org/10.1016/j.eneco.2010.10.003
• KIM, J. and PARK, K., 2016. Financial development and deployment of renewable energy technologies. Energy Economics, vol. 59, pp. 238-250. http://doi.org/10.1016/j.eneco.2016.08.012
  » http://doi.org/10.1016/j.eneco.2016.08.012
• OUYANG, Y. and LI, P., 2018. On the nexus of financial development, economic growth, and energy consumption in China: new perspective from a GMM panel VAR approach. Energy Economics, vol. 71, pp. 238-252. http://doi.org/10.1016/j.eneco.2018.02.015
  » http://doi.org/10.1016/j.eneco.2018.02.015
• SADORSKY, P., 2010. The impact of financial development on energy consumption in emerging economies. Energy Policy, vol. 38, no. 5, pp. 2528-2535. http://doi.org/10.1016/j.enpol.2009.12.048
  » http://doi.org/10.1016/j.enpol.2009.12.048
• SADORSKY, P., 2011. Financial development and energy consumption in central and eastern European frontier economies. Energy Policy, vol. 39, no. 2, pp. 999-1006. http://doi.org/10.1016/j.enpol.2010.11.034
  » http://doi.org/10.1016/j.enpol.2010.11.034
• SHAHBAZ, M., SHAHBAZ SHABBIR, M. and SABIHUDDIN BUTT, M., 2013. Effect of financial development on agricultural growth in Pakistan: new extensions from bounds test to level relationships and Granger causality tests. International Journal of Social Economics, vol. 40, no. 8, pp. 707-728. http://doi.org/10.1108/IJSE-01-2012-0002
  » http://doi.org/10.1108/IJSE-01-2012-0002
• VIJAYKUMAR, J., RASHEED, A., and TONDKAR, R., 2018. Country risk and foreign direct investment: an empirical investigation multinational business review. Multinational Business Review, vol. 17, no. 3, pp. 181-204.

Publication Dates

• Publication in this collection
  25 Nov 2024
• Date of issue
  2024

History

• Received
  21 July 2024
• Accepted
  02 Sept 2024