Bridging the Green Infrastructure Gap: Determinants of Renewable Energy PPP Financing in Emerging and Developing Economies

Abstract

This study analyses the factors influencing renewable energy infrastructure public–private partnership (PPP) financing, using data from 28 countries covering the period from 1996 to 2024. A composite institutional quality index was constructed using Principal Component Analysis (PCA). The analysis employs a panel econometric framework: the autoregressive distributed lag (ARDL) model to capture short- and long-term dynamics. The results highlight the significance of the time dimension on renewable energy PPP financing. In the short term, none of the predictor variables are significant, reflecting the inherently long-term character of renewable energy PPP investments. However, in the long term, gross domestic product per capita, inflation dynamics, efficiency in energy transmission, and institutional quality are identified as key determinants of renewable energy investment. The findings suggest that strengthening sector-specific regulatory frameworks and improving various aspects of institutional quality as defined by the World Governance Indicators can be important to attract private capital in energy PPPs. These institutional reforms, complemented by growth-oriented macroeconomic policies, would contribute to making renewable energy markets more attractive while reducing exposure to macroeconomic and institutional risks.

1. Introduction and Background

The change in climate is a serious threat to our civilisation. The consequences of climate change are already visible and will be calamitous if urgent action is not taken now. The United Nations [UN] Sustainable Development goal [SDG] 13 is a call for action to address climate change and its effects on humanity [1]. Ref. [2], in sync with [1], recognise that global warming currently represents one of the most significant dangers to our planet. In 2024, greenhouse gas emissions (GHG) increased, with carbon dioxide (CO₂) from fossil fuels growing by 0.8% to 37.4 billion tonnes. Atmospheric CO₂ reached a record high of 424.61 ppm [3]. Clearly, urgent mitigation measures are required. Scientific projections are that, globally, average surface temperatures may increase by more than 3 degrees Celsius during the century. Ref. [4] noted that the consequence of this climatic shift is not symmetric across countries. By and large, the adverse impact of change in climate is borne by poorer and developing countries. Three reasons are cited for this observation. Firstly, less economically developed nations are vulnerable to weather conditions because of the significant role of agriculture and water resources in their economies [5,6]. Conversely, more economically developed countries have a greater proportion of their economic activities in manufacturing value addition systems, which are generally more insulated from weather fluctuations and, consequently, climate change [2,4]. Second, and geographically, poorer countries are often located in hotter regions, where ecosystems are already near their limits. If these areas become even hotter, no models currently exist that can be followed. Emerging technologies and behaviours will need to be developed through trial and error. Third, adaptive capacity is weak in low-income countries due to limited access to modern technology and institutions that are incapable of safeguarding the populace against adverse climatic shifts [4,7].

According to [8], burning fossil fuels for energy and transportation is the main source of CO₂ and other greenhouse gas emissions. When burned, these fuels emit significant amounts of CO₂, methane, and nitrous oxide into the atmosphere. Quantitatively, the energy industry accounts for 66% of GHS emissions and approximately 80% of the total CO₂ emissions. Within the energy industry, electricity generation from fossil fuels is a major contributor to greenhouse gas (GHG) emissions, comprising a significant portion of global energy-related emissions [9]. Although electricity constitutes approximately 20% of final energy consumption, its production accounts for over 40% of all energy-related emissions [10]. The combustion of fossil fuels for electricity generation produces considerable quantities of CO₂, with coal-fired power plants emitting approximately 820 g of CO₂ per kilowatt-hour (kWh) and natural gas power plants emitting around 450 g of CO₂ per kWh [8,10].

Refs. [11,12] assert that renewable energies can be configured as resources with the capacity to reduce temperature increases and facilitate the transition to low-carbon economic models in emerging and poorer economies [EMDEs]. In fact, SDG 7 endeavours to achieve access to reliable, reasonably priced, modern and sustainable energy for all by 2030 [1]. More specifically, target 7. B outlines the following objective: “By 2030, to improve infrastructure and technology to deliver sustainable, modern energy to all developing countries—especially the least developed, small island, and landlocked nations—in alignment with their support programs”. Thus, enhancing the quality and quantity of energy infrastructure stock has become a key component of sustainable development policies in EMDEs [13,14]. Compared to advanced market economies, infrastructure in EMDEs lags considerably behind in quality, quantity and accessibility [15]. According to [13], the gap is notably wider in the energy sector. Researchers have observed that, in the last decade, EMDEs have witnessed noticeably high levels of economic growth while simultaneously confronting rapid urbanisation and population growth [16,17]. Others have recommended that EMDEs need to accelerate investment to rehabilitate, upgrade, and build new facilities to sustain the socio-economic growth prospects, meet the SDG targets by 2030, and adapt to climate change challenges. Maintaining high-quality energy infrastructure service delivery requires an appropriate composition of the infrastructure stock and sufficient maintenance.

However, a huge gap in financing exists in the energy sector. According to [18], estimates, globally, a total of USD 4.5 trillion is required if net zero GHG emissions are achievable by 2030. Yet, despite international efforts to decarbonise the energy sector, clean energy investments are estimated at USD 1.8 trillion, creating a financing gap of USD 2.7 trillion [18]. EMDEs are hardest hit by the scarcity of renewable energy finance, given the observation of [19] that international investments in renewable energy are asymmetrically distributed across countries, with the lion’s share of investments being channelled towards developed countries. Even though [20] reported a significant increase in the volume of investments in EMDEs, ref. [21] stated that only 12% of the financial flows are allocated to developing countries. EMDEs require approximately USD 1.6 trillion annually for clean energy financing [22].

With limited resources, the public sector in EMDEs, which is traditionally the major source of finance, cannot singlehandedly ensure adequate infrastructure funding and support activities that guarantee quality service provision [13,23]. Resultantly, in EMDEs, existing infrastructure facilities either need upgrading or modernisation. In some cases, the infrastructure has outlived its economic lifespan. Under hard capital market rationing, financing energy infrastructure in EMDEs is even more challenging, given that, due to population growth, energy demand has substantially increased [24]. International Development Agencies (IDA) and academics have recommended public-private partnerships (PPPs) as a practical approach to address the disparity between infrastructure demand and supply in EMDEs [25,26]. According to [8], mobilising renewable energy investments at a large scale requires high private sector involvement. In a climate-driven scenario postulated by [8], the private sector is expected to contribute approximately 70% of the clean energy investments.

Private financing of renewable energy PPPs remains surprisingly an understudied area of research in EMDEs [13,19]. Where efforts have been made to close this empirical research gap, the findings have provided mixed evidence. What determines private financing of energy PPPs has remained relatively unclear, as findings tend to vary with the sample period and frame, as well as the analytical framework of the study. For instance, ref. [27] utilised a global panel to examine financial market development factors’ impact on PPP project volumes in the energy sector. A dynamic panel generalised method of moments [GMM] approach was used. Notably, the energy sector is reported to be positively correlated with greater access to long-term financing.

Similarly, ref. [28] investigated the determinants of private sector participation in PPPs, analysing a panel of 37 purposively selected countries. A GMM model was fitted. Estimates indicate that while the banking sector has a positive, albeit statistically insignificant, effect on private investment, capital markets significantly promote private sector risk assumption. Nonetheless, ref. [16], an extension of [28], shows that developing countries with more advanced banking sectors attract greater investment in PPP projects. Unlike [28], it appears that the banking sector serves as the primary channel through which overall financial development positively influences private participation in energy PPP projects.

Other than financial market aspects, ref. [13] posited that the quality of institutions is a key predictor of investments in renewable energy PPPs. The researchers analysed country-level data from MENA countries. The results indicate that only regulatory quality [RQ] has significant impact on investments, with a 1% improvement in regulation resulting in a 3.7% increase in investment. Voice and accountability [VA], political stability [PS], government effectiveness [GE], rule of law [RL], and control of corruption [CC] generally did not exhibit a significant relationship with PPP investment volumes.

A similar study, ref. [19], used a logistic quantitative panel data analysis to investigate the determinants of cross-border private investment in renewable energy in developing countries. The conclusion was that, if the government introduces one renewable energy-friendly policy, the odds of additional investments are multiplied by a factor of 51%. At the macro level, only regulatory policies significantly influence investment decisions. When these interventions are implemented, the likelihood of investment increases by a factor of 22%. In this respect, this study aligns with [13], who reported that the quality of regulation is a strong driver of PPP investments.

Whilst [13,19] concur that the political environment is not a significant predictor of investments in energy PPPs, ref. [2] report a contrasting finding. Counter-intuitively, this study reports political stability and absence of terrorism as negatively and statistically significant. This finding is particularly interesting given that International Development Finance Institutions [IDFI] and notable prior researchers identified political stability as a key factor in attracting private investment in emerging and developing economies [25,29]. Nonetheless, this unexpected result may be interpreted through the lens of the complex dynamics inherent in political instability. Various forms of violence are often rooted in systemic inequalities, which may lead governments facing heightened instability to implement policy measures focused on socio-economic development to address these issues. For example, in [30], it is reported that poverty frequently precedes military coups d’état and the promise of better service delivery through infrastructure investments.

Beyond the institutional environment, a growing body of literature suggests that macroeconomic conditions are significant predictors of private sector participation in renewable energy PPP financing [2,19,31]. Ref. [2] employed a Tobit regression model on a panel of 63 developing countries to test the hypothesis that gross domestic product (GDP) levels influence private investment in renewable energy PPP infrastructure. Contrary to expectations, their findings revealed a statistically significant and negative relationship between GDP and private investments in renewable energy PPPs. This outcome implies that lower-GDP countries, often constrained by limited public financing capacity, may rely more on the private sector for financing [14,32]. However, in a subsequent study, ref. [31] reported a positive and statistically significant coefficient of GDP at the 5% level. This finding is not entirely unexpected given the susceptibility of energy PPPs in emerging and developing economies to demand-side risks [27,33]. These risks are typically mitigated in larger economies, where broader market bases can support large-scale energy PPP developments and ensure more stable returns for private investors.

Essentially, researchers are reporting conflicting findings on the predictive power of macroeconomic variables [2,13,19], institutional environment [2,13,19,29] and financial development factors [16,27,28] on private investment in renewable energy PPPs. Thus, a re-examination of the predictors of private investments in energy PPPs is justified. This study analyses the determining factors of private sector outlay in renewable energy PPPs in EMDEs. More specifically, the question of what determines private sector financing of renewable energy PPPs in EMDEs underpins this study. Understanding the predictors of private finance is fundamental for several reasons. First, it lays the foundation for monitoring and evaluating the progress made at national and international levels to tackle climate change through clean energy investments. Second, it informs decision-making about the effective design of policies and financial instruments with the potential to mobilise the highest amounts of capital. Third, the focus on the energy sector is crucial because energy infrastructures present the most significant disparity in infrastructure endowment within emerging economies compared with other infrastructure types. Closing this gap is thus a critical component if the SDG targets are to be met.

A panel auto-regressive distributed lag [ARDL] is used in this study. The ARDL framework is advantageous because it captures both the short and long-run dynamics of the relationship under investigation. To the best of our knowledge, this approach has scarcely been applied to understand the drivers of private financing of renewable energy PPPs in emerging and developing markets. In this way, this study extends the body of literature that is currently founded on frameworks that include, but are not limited to, censored regression [34], linear regression [2], and survey design [9].

Following the introduction, Section 2 presents an overview of the methodological framework. Section 3 discusses the findings of this study and Section 4 presents the concluding notes.

2. Methodology

2.1. Population and Sample

The sampling approach employed in this study was primarily purposive, reflecting the objective of identifying key determinants of private sector financing for renewable energy PPPs in selected developing and emerging countries. The sample was derived from the World Bank’s Public-Private Infrastructure Advisory Facility (PPIAF) database, a comprehensive repository that documents PPP investments across countries and time periods. The inclusion criterion was contingent on the availability and completeness of relevant data. Consequently, the final sample comprised 28 geographically dispersed countries: nine from Africa, ten from the Asia-Pacific region, three from Europe, five from Latin America, and one from North America. The data period is from 1996 to 2024. Given the intention to apply the panel Autoregressive Distributed Lag (ARDL) modelling framework, the sample size is deemed methodologically appropriate, as supported by [35]. The countries included in this study are Argentina, Bangladesh, Brazil, Cambodia, Chile, China, Côte d’Ivoire, Egypt, Ghana, India, Indonesia, Kenya, Malaysia, Mexico, Morocco, Nigeria, Pakistan, Peru, Philippines, Russia, Senegal, South Africa, Thailand, Tunisia, Turkey, and Vietnam. According to data from the PPIAF, these countries represent some of the most active renewable energy PPP markets globally, thereby providing a relevant empirical base for the analysis.

2.2. Data and Variables

This research benefited from several databases. The main data banks are the WDI, WGI, and PPIAF. These have underpinned many studies in developing and emerging countries [2,19,31]. From these databases, a full panel dataset that covers the period 1996 to 2024 was generated. The variables are described as follows: the loanable funds to private sector (financial resources availed to the private sector measured as a percentage of the gross domestic product), real GDP per capita (expressed in 2010 USD prices), the inflation rate (proxied through the consumer price index), the stock market capitalisation (yearend share price values in USD expressed as a percentage of GDP), transmission losses (a measure of quality and efficiency proxied by distribution losses in the power network as a share of total output). Private investors views inefficiency in power infrastructure as a risk factor that may increase project costs and subsequently reduce returns. Foreign direct investment is measured as the Net investment inflows for acquiring approximately 10% of voting stock in a foreign company to assume managerial control.

Institutional quality is proxied by the six measures of governance developed by [36] and reported by the WGI databank. These are GE, RL, PS, RQ, VA and CC. Collectively, the measures indicate the processes of government selection and monitoring, government capacity in effective and sound policy formulation and implementation, and the level of due regard for the institutions that govern the economy and society by the state and citizens [32]. In our analysis, the dependent variable “private participation in renewable energy project”, refers to the total financial commitments by the non-state sector. Refs. [2,13,19,29] defined the predicted variable this way. The sample renewable energy technology encompasses projects in hydro power generation, wind technology, solar energy, enhanced geothermal, biomass and waste management technologies. Stata/SE 14.2 is used for econometric estimations.

2.3. Principal Component Analysis (PCA)

The PCA is used in this study to generate a composite institutional index. This is important because the six governance variables presented above are highly correlated according to the literature [14,32]. More so, there is no consensus amongst researchers regarding which of the six has the strongest impact on private investment in renewable energy PPPs in emerging and developing markets. In fact, refs. [32,37,38] used the PCA to generate a composite institutional quality index. PCA converts correlated variables into independent principal components [37], reducing dataset dimensionality while retaining maximum variance and key information [37,38].

Table 1. Principal component analysis-eigenvalues.

Component	Eigenvalue	Proportion	Cumulative
Comp 1	3.853	0.642	0.642
Comp 2	0.999	0.167	0.809
Comp 3	0.576	0.096	0.905
Comp 4	0.288	0.048	0.953
Comp 5	0.171	0.029	0.982
Comp 6	0.111	0.019	1

shows the eigenvalues for governance variables: CC, RQ, RL, VA, PS, and GE. The first component explains 64.22% of the variation in the original dataset. The respective eigenvalue is 3.8530.

Table 2. Eigen vector loadings.

Variable	Comp 1	Comp 2	Comp 3	Comp 4	Comp 5	Comp 6
PS	0.057	0.991	0.086	0.015	0.049	0.065
VA	0.365	−0.114	0.897	−0.051	0.188	0.095
GE	0.458	−0.043	−0.372	0.019	0.794	0.136
RQ	0.450	−0.019	−0.074	0.828	−0.322	0.038
CC	0.481	0.045	−0.086	−0.296	−0.169	−0.801
RL	0.468	−0.019	−0. 186	−0.471	−0.445	0.569

presents the eigenvector loadings from the PCA. The first component (PC1) displays positive coefficients for all six governance dimensions. According to the analysis, this indicates that, in emerging and developing countries, governance quality indicators are positively associated with overall governance quality. Consequently, PC1 represents the primary information regarding governance in the original dataset [39]. In any case, the suitability of incorporating the PCA in the analytical mix of this study is supported by the Kaiser-Meyer-Olkin (KMO) metric reported in

Table 3. Kaiser-Meyer-Olkin Statistic.

Variable	KMO
Political stability	0.347
Voice and accountability	0.879
Government effectiveness	0.861
Regulatory quality	0.904
Control of corruption	0.816
Rule of law	0.856
Overall	0.854

. The PCA technique requires the KMO statistic to be greater than 0.5 [40,41]. This condition is satisfied in Table 3.

2.4. Auto-Regressive-Distributed Lag [ARDL] Model

This study aims to determine whether the determinants of PPPs identified in the literature are relevant to private financing of renewable energy PPPs in emerging and developing countries using a framework that captures both the short and long run dynamics. Empirical studies have either been methodologically anchored on a qualitative study approach [9,42] or static linear models [2,31]. A few are based on dynamic economic modelling [16,28]. The panel ARDL has been applied thinly in renewable energy PPP research. The panel data ARDL model requires that both the cross-sectional dimension (N) and the time dimension (T) exceed 25 as an initial requirement [43]. Additionally, confirming the absence of cross-sectional dependence within the dataset is essential. Most critically, for the practical application of the panel ARDL approach, the dependent variable should be stationarity after first differencing, while the independent and control variables should demonstrate stationarity either at level, at first difference, or as a combination thereof [43,44].

Panel unit root tests can generally be classified into two broad categories: first-generation and second-generation tests. First-generation tests assume that variables are cross-sectionally independent and have uncorrelated residuals in the error correction model [43,44]. Prominent examples of these tests include the Breitung [45] and Levin, Lin, and Chu [46] t-statistics, which assume a common unit root process across panels, as well as the Im, Pesaran, and Shin [47] W-statistic and Fisher-type tests based on Augmented Dickey–Fuller (ADF) and Phillips-Perron (PP) χ²-statistics, which allow for individual unit root processes. In contrast, second-generation tests consider cross-sectional dependence, that can result from common shocks or unobserved factors across panel units. Notable contributions in this category include the tests developed by Maddala and Wu [48], Pesaran [49], Breusch and Pagan [50], and Chudik and Pesaran [51], among others. These tests are designed to support inference when cross-sectional dependence occurs, which are common in macro-economic and financial data sets.

2.4.1. Cointegration Test

The ARDL model, as advanced by [49], was employed to achieve this study’s objectives. This modelling approach represents a generalisation of the conventional ARDL model commonly applied to time series data. Compared to traditional cointegration techniques such as the Engle-Granger and Johansen methodologies, the panel ARDL framework offers distinct advantages. Notably, it relaxes the strict requirement that all regressors must be non-stationary and integrated of the same order [38,52]. The panel ARDL model is applicable when variables are integrated of order zero [I(0)], order one [I(1)], or a mixture of both [44,49]. The panel ARDL approach effectively captures both short term dynamics and long-run equilibrium relationships via the embedded error correction mechanism. This feature allows for a more comprehensive analysis by simultaneously estimating the immediate and sustained effects of the explanatory variables on private sector participation in renewable energy PPPs within emerging and developing economies. Ref. [53] noted that the error correction component enables the model to retain information about the long-term relationship while accommodating short-term fluctuations. In the context of this study, the appropriate estimator between the pooled mean group [PMG], mean group [MG], or dynamic fixed effects [DFE] was selected based on the Hausman specification test, following the procedure outlined by [49].

2.4.2. Model Specification

This study seeks to examine the impact of domestic credit to the private sector, GDP per capita, inflation rate, stock market capitalisation, transmission losses, foreign direct investment, and institutions are significant predictors of private investments in renewable energy PPPs in emerging and developing economies. In accordance with [43,44], the following panel ARDL model is specified as follows:

$$\begin{array}{cc}\hfill {∆logEPI}_{it}=& {\delta }_0+{\delta }_1{EPI}_{it-1}+{\delta }_{2}log{FDI}_{it-1}+{\delta }_3{logGDP}_{it-1}+{\delta }_4{CPI}_{it-1}+{\delta }_5{SMK}_{it-1}+{\delta }_6{LTE}_{it-1}+{\delta }_7{PCX}_{it-1}+{\sum }_{i=0}^p{\delta }_{1i}∆{logEPI}_{it-1}\hfill \\ & +{\sum }_{i=0}^q{\delta }_{2i}{∆logFDI}_{it-1}+{\sum }_{i=0}^q{\delta }_{3i}{∆GDP}_{t-1}+{\sum }_{i=0}^q{\delta }_{4i}∆{CPI}_{it-1}+{\sum }_{i=0}^q{\delta }_{5i}∆{SMK}_{it-1}+{\sum }_{i=0}^q{\delta }_{6i}∆{LTE}_{t-1}\hfill \\ & +{\sum }_{i=0}^q{\delta }_{7i}∆{PCX}_{t-1}+{\epsilon }_{it}\hfill \end{array}$$

(1)

where ${EPI}_{it}$ is the level of private investment in renewable energy PPP for country i at time t, ${FDI}_{it}$ is the level of foreign direct investment for country i at time t, ${GDP}_{it}$ is the per-capita gross domestic product for country i at time t, ${CPI}_{it}$ is the consumer price index for country i at time t, ${SMK}_{it}$ is the stock market capitalisation for country i at time t, ${LTE}_{it}$ is an efficiency indicator that measures transmission losses in the energy sector for country i at time t, ${PCX}_{it}$ is a principal component generated governance index and ${\epsilon }_{it}$ is the error term. Equation (1) is used to test for cointegration. After identifying a long-run relationship, the error correction model (ECM) within a panel ARDL framework is specified as follows:

$$\begin{array}{cc}\hfill {∆logEPI}_{it}=& {\delta }_0+{\sum }_{i=0}^p{\delta }_{1i}∆{logEPI}_{it-1}+{\sum }_{i=0}^q{\delta }_{2i}{∆logFDI}_{it-1}+{\sum }_{i=0}^q{\delta }_{3i}{∆GDP}_{t-1}+{\sum }_{i=0}^q{\delta }_{4i}∆{CPI}_{it-1}+{\sum }_{i=0}^q{\delta }_{5i}∆{SMK}_{it-1}\hfill \\ & +{\sum }_{i=0}^q{\delta }_{6i}∆{LTE}_{t-1}+{\sum }_{i=0}^q{\delta }_{7i}∆{PCX}_{t-1}+\delta {ECT}_{it-1}+{\epsilon }_{it}\hfill \end{array}$$

(2)

Acronyms are defined as in Equation (1). AIC was used to select suitable lag lengths for the variables. Having outlined the fundamental tenets of the econometric framework being applied, the succeeding section presents the empirical findings.

3. Empirical Findings and Discussion

3.1. Descriptive Statistics

Table 4. Descriptive statistics.

Variable	Observations	Mean	Std. Dev.	Min	Max
EPI	812	1,082,614,993	2,840,000,000	0.000	38,416,500,000
CPI	812	176.254	741.462	0.700	14,581.000
DCR	812	45.141	34.892	4.722	194.674
FDI	812	2.916	2.502	−0.858	19.1496
GDP	812	4666.600	3636.665	496.770	14,713.570
PCX	812	0.000	1.000	−1.970	2.852
SMK	812	52.324	50.861	0.000	322.711
LTE	812	13.2898	5.6813	0.000	37.430

shows the descriptive statistics of the variables in the analysis. The mean value of renewable PPP investment in emerging and developing countries over the sample period is approximately USD1.1 billion. This observation corroborates [8] reporting that, despite a renewable energy funding gap in emerging and developing countries, private investors are now paying attention to this segment of the energy market. According to [16,28], the demand for renewable energy in emerging and developing markets is expected to remain strong due to high economic growth rates and climate-induced transitions in the same countries. The maximum annual private investment value is USD38,416 billion.

On the other hand, the assertion by [54] that, on average, emerging and developing countries face strong inflationary pressures is descriptively confirmed. The average consumer price index is 176.2547% with a maximum of 14.581%. Some countries, like Zimbabwe, experienced hyperinflationary conditions over the period under analysis. The CPI standard deviation is approximately 741.46%. The proportion of loanable funds allocated to the private sector as a proportion of GDP ranges between 34% and 194%. This reflects that loanable funds markets are thin in some countries and strong in others. The ratio of stock market capitalisation to GDP averages 52%. Refs. [16,28] argue that stock markets in developing countries are yet to mature and, as such, they struggle to raise finance for long gestation renewable energy PPP infrastructure projects. The descriptive statistics show that foreign direct investments normalised on GDP are very low: 2.9162%. This compounds the [8] report that emerging and developing countries generally struggle to attract investment funds, especially on a long-term basis. The average GDP per capita is USD 4666.6. This suggests that effective demand for renewable energy exists in emerging and developing countries. However, GDP per capita ranges between USD 496.7705 and USD 14,713.57. The PCA constructed ranges between −1.97 and 2.852, whilst the energy losses in transmission average 13.28%. The standard deviation of the variables provided evidence that the data is suitable for econometric analysis.

Table 5. Correlation coefficients.

	EPI	CPI	DCR	FDI	GDP	PCX	SMK	LTE
EPI	1
CPI	−0.019	1.000
DCR	0.016	−0.062	1
FDI	0.130	−0.078	0.145	1
GDP	0.082	−0.072	0.282	0.108	1
PCX	0.103	−0.125	0.140	−0.050	−0.134	1
SMK	0.033	0.063	0.395	0.009	0.125	0.366	1
LTE	−0.085	0.071	−0.522	−0.108	−0.386	−0.174	−0.271	1

presents the correlation coefficients between private investment in renewable energy PPP in emerging and developing countries with various independent variables. Variables that exhibit a strong linear relationship with PPP, as indicated by a high correlation coefficient, include foreign direct investment, institutional quality, gross domestic product per capita, and energy losses in transmission. This suggests that countries’ level of purchasing power, the governance environment and transmission losses may be significant predictors of private investment in energy PPP projects in emerging and developing economies. In contrast, the level of inflation, credit to the private sector as a fraction of gross domestic product and the stock market capitalisation do not seem to exhibit a strong association with energy PPP investments.

It is important to note that multicollinearity does not appear to present a concern, as the correlation coefficients among the independent variables remain relatively low.

3.2. Cross-Sectional Dependency

Besides multicollinearity, cross-sectional dependency is a key consideration when modelling panel data [44,53]. The approach to unit root testing depends on the cross-sectional dependency test outcome. Even though the Frees and Friedman tests are potentially applicable techniques, this study adopted Pesaran’s test of cross-sectional independence. This approach has been widely applied in applied econometrics [53].

Table 6. Pesaran’s test of cross-sectional independence test.

Pesaran’s test of cross-sectional independence statistic	1.365
Pesaran’s test of cross-sectional independence probability value	0.1723

reports Pesaran’s test of cross-sectional independence. Given a probability value greater than 5%, it implies that the data set has no cross-sectional dependence, in which case the first-generation unit root tests are applicable [44,53].

3.3. Unit Root Test

The Im-Pesaran–Shin (IPS) unit root test is applied on the series at levels and at first difference. The first-generation unit root tests assume that countries are independent of each other, with some heterogeneity expected [35,38]. The null hypothesis is that the panels possess a unit root, while the alternative hypothesis is that the panels are stationary. As shown in

Table 7. IM—Pesaran—Shin Unit Root Test.

IM—Pesaran—Shin (IPS) Unit Root Test
Variable	At Level	At First Difference	Conclusion
EPI	−18.2720 ***	−44.9640 ***	I(0)
CPI	−1.9188 **	−10.8502 ***	I(0)
DCR	−0.6297	−17.4247 ***	I(1)
FDI	−11.3939 ***	−30.5449 ***	I(0)
GDP	−5.0983	−11.6084 **	I(1)
PCX	−2.8614 ***	−21.4603 ***	I(0)
SMK	−9.8787 ***	−32.0459 ***	I(0)
LTE	−0.8562	−20.8826 ***	I(1)

*** and ** represent 1% and 5% levels of significance, respectively.

, the regression variables are either integrated of order zero I(0) or one I(1). Based on these results, the panel ARDL econometric model is appropriate for examining the phenomenon under study [35,38].

3.4. The Lag Structure

The optimal lag structure per country and per variable is reported (p, q, q, q, q, q, q, q) in

Table 8. Optimal lags (0,1,1,1,1,1,1,1).

COUNTRY	EPI	GDP	FDI	CPI	DCR	PCX	TLE	SMK
Argentina	0	1	1	2	1	1	1	1
Bangladesh	0	1	1	1	3	1	1	1
Brazil	0	1	1	1	1	1	1	1
Cambodia	0	1	1	3	1	1	2	1
Chile	1	2	2	3	2	1	1	1
Colombia	0	2	1	3	1	1	1	1
Egypt	0	1	1	2	3	1	2	1
Ghana	0	1	0	1	1	1	1	1
India	0	3	0	1	1	1	3	3
Indonesia	2	2	0	2	2	1	1	1
Kazakhstan	2	1	2	3	1	3	3	1
Kenya	3	1	1	1	2	1	1	1
Malaysia	1	1	1	1	1	1	1	1
Mexico	2	1	0	3	1	1	3	1
Morocco	2	1	3	1	2	1	1	1
Nigeria	1	3	2	2	1	1	1	2
Nigeria	0	2	0	0	2	1	0	1
Pakistan	1	1	1	3	2	1	3	2
Peru	1	3	1	2	1	1	1	1
Philippines	0	1	2	1	2	1	3	3
Russia	1	1	1	3	3	1	3	3
Senegal	0	1	0	2	2	1	3	1
South Africa	3	2	0	1	1	1	2	1
Sri Lanka	0	1	1	3	1	1	1	1
Thailand	0	1	2	3	1	1	1	2
Türkiye	1	1	2	3	1	1	1	1
Viet Nam	1	1	1	2	3	1	2	1
Zimbabwe	0	2	1	1	1	2	1	2
Model	0	1	1	1	1	1	1	1

. Akaike-Information-Criterion [AIC] incorporated the modal lag across the panel of countries and variables for further analytics. Across the seven predictor variables, the modal lag is one and zero for the predicted variables. Refs. [35,38] recommended the approach adopted in this study to determine the optimal lag structure for a panel ARDL model.

3.5. Cointegration Test

A panel cointegration test was conducted to mitigate the risk of spurious regression results. Evidence of cointegration indicates that the estimated relationships are stable and not driven by random trends [43,45]. The null hypothesis of the test posits the absence of cointegration.

This study employed both the Pedroni test to assess the presence of cointegration within the panel data. As reported in

Table 9. Pedroni Cointegration test.

Test Stats	Panel	Group
v	−0.829
rho	−0.3.34	−1.833
t	−17.675	−20.04
ADF	−10.444	−8.648

, a t-value greater than 1.96 confirms that the variables adjust to long-run equilibrium over time.

3.6. Panel ARDL Results

Having eliminated the risk of spurious regressions, using the Hausman test to finalise the pooled mean group [PMG] as the preferred model over the mean group [MG],

Table 10. Panel ARDL PMG—Estimator.

Variables	Short-Run Dynamics	Long-Run Dynamics
GDP	1.3123	2.433 **
	(12.115)	(1.109)
FDI	0.4305	0.336
	(0.653)	(0.289)
DCR	0.061	−0.029
	(0.096)	(0.024)
SMK	0.035	−0.007
	(0.035)	(0.005)
CPI	3.742	1.194 **
	(12.754)	(0.427)
LTE	0.363	−0.141 **
	(0.466)	(0.063)
PCX	1.021	1.085 ***
	(1.851)	(0.385)
Error correction term	−0.934 **
	(0.055)
Constant	9.010
	(1.328)
Number of observations	770
Number of groups	28
Log Likelihood	−2388.358

*** and ** represent significance at 1% and 5%, respectively.

presents the panel ARDL estimation results.

The parameter estimates that, in the long run, the economy’s size and the populace’s respective purchasing power, at the 5% level, are significant predictors of private investment in renewable energy in emerging and developing countries. The relationship between per-capita gross domestic product is strong and linear. This finding resonates well with empirical literature that claims that projected higher demand and clientele ability to pay are appealing to private investors in energy markets [16,19,28]. According to [27], demand risk is minimal in a well-functioning economy, and such economies tend to attract high volumes of private investments. Nonetheless, our findings contradict those of [31], who reported a negative and significant relationship between gross domestic product per capita and private investment in renewable energy PPPs. The fact that our study adopted an econometric approach that captures both the short-run and long-run PPP market dynamics potentially explains the differences in findings.

This study reveals a positive and statistically significant predictive effect of inflation on renewable energy public–private partnerships (PPPs) at the 1% level. While a long-run positive relationship between inflation and private investment in infrastructure may initially appear counterintuitive—given that inflation is often regarded as a risk factor that discourages investment [55,56], there are contextual mechanisms that help explain this outcome. Many infrastructure projects, particularly in energy, transport, utilities, and healthcare sectors, operate under tariff structures or concession agreements that are indexed to inflation. Over time, moderate inflation enhances nominal revenues through higher tolls, user charges, or electricity tariffs, thereby strengthening projected cash flows and improving the financial viability of projects. Consequently, investors may view infrastructure assets as a partial hedge against inflation, reinforcing their willingness to commit capital [57].

This study shows that the PCA-generated institutional quality index is highly significant at 1% as a predictor of private investment in energy PPPs. This finding corroborates the strand of literature that holds that countries with robust institutions are more successful in mobilising private capital for infrastructure because investors perceive lower risks of expropriation, renegotiation, or arbitrary policy changes [13,19]. In renewable energy projects, where capital requirements are substantial and payback periods are long, institutional quality becomes critical to mitigate risks associated with political interference, tariff adjustments, and regulatory instability [14,26,32]. Also, in contrast with [16,28], our results indicate that at the 5% significance level, and in the long run, energy transmission and distribution losses dissuade private investment in renewable energy PPPs. Given that the level of distribution losses is sometimes viewed as a proxy for the quality of public investment, the finding implies that private investors incorporate the minimum levels of efficiency in their investment strategies.

The regression model results further reveal that the capital market variable, the stock market capitalisation, and the bank market variable, domestic credit to the private sector, are not significant in the short or long run. This can be explained by the fact that in most developing and emerging economies, capital markets are shallow and illiquid and as such they have limited capacity to consistently finance long-term infrastructure projects. However, in [16,27,28,32] the financial sector is strongly and positively related to private investment. It is frequently contended in the development finance discourse that the sophistication of a country’s financial system plays a significant role in influencing private investors’ decisions to engage in its energy projects [58,59]. Similarly, this study reported the level of foreign direct investment as insignificant in the short and long run. This is counterintuitive given that, according to [38], international capital flows are the pivotal for development in emerging and developing countries.

4. Conclusions and Policy Recommendations

In conclusion, this study underscores the importance of the time dimension in renewable energy PPP financing. The absence of significant predictor variables in the short run reflects the inherently long-term nature of renewable energy PPP investments. Conversely, in the long run, gross domestic product per capita, inflation dynamics, energy transmission efficiency, and governance quality emerge as highly significant determinants of renewable energy investment in emerging and developing economies. By contrast, foreign direct investment and indicators of financial market development—such as stock market capitalisation and domestic credit to the private sector—are not significant in the short or long run.

This study makes two principal contributions to the literature. Methodologically, it applies to the panel ARDL econometric framework, which has seldom been employed in analyses of renewable energy PPPs in emerging and developing contexts, thereby capturing both short- and long-run dynamics. Empirically, by employing an up-to-date dataset, this study enriches the relatively scarce body of scholarship on renewable energy PPPs in these markets. This contribution is timely given the central role of affordable and clean energy in achieving the UN Sustainable Development Goals (SDGs) by 2030 and Africa’s Agenda 2060.

From a policy perspective, the findings provide governments and policymakers with robust empirical evidence on the interplay between institutional quality, the efficiency of energy distribution networks, and broader macroeconomic conditions in shaping renewable energy PPP finance. The results suggest that public policy should prioritise strengthening sector-specific regulatory frameworks and the broader qualitative dimensions of institutions—including government effectiveness, regulatory quality, rule of law, control of corruption, voice and accountability, and political stability. Governments in emerging and developing countries can prioritise the development of sound, predictable, transparent and enforceable laws that govern the aspects of PPP contracts such as the award of contracts, dispute resolution, and tariff structures, among other aspects. This enhances the governance risk profile of PPP projects as frequent and abrupt policy reversals undermine investor confidence.

Growth-oriented macroeconomic management strategies should complement these institutional reforms. This is significant because stable economic environment fundamentally supports renewable energy PPPs. Considering this, governments can aim to implement policies and strategies that foster broad-based economic growth. Such measures can enhance the capacity of private firms and households to engage in and sustain PPP financing through increased energy consumption and a greater willingness to pay. Collectively, such measures would enhance the attractiveness of renewable energy markets while mitigating risks associated with macroeconomic volatility and institutional fragility.

Limitations of This Study and Future Research Directions

To address the limitations associated with small sample sizes in empirical analyses, this study employed a macro-level perspective of the renewable energy sector within emerging and developing economies. However, this approach may be limited, as predictor variables might not exert uniform effects across various energy sub-sectors. Accordingly, where data permits, future research could concentrate on individual energy sectors or specific geographic regions and compare those findings to the present study. Moreso, the PCA index can be decomposed into its components for in-depth analysis of the role of intuitions in PPP finance. Application of other econometric methodologies such as the GMM framework can enrich future research.