Next Article in Journal
Evaluating the Use of Rice Husk Ash for Soil Stabilisation to Enhance Sustainable Rural Transport Systems in Low-Income Countries
Previous Article in Journal
Writing Is Coding for Sustainable Futures: Reimagining Poetic Expression Through Human–AI Dialogues in Environmental Storytelling and Digital Cultural Heritage
Previous Article in Special Issue
Higher Education Institutions as Leverage for Backing up SMEs’ Efforts to Meet SDG 9
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Industrial Diversification in Emerging Economies: The Role of Human Capital, Technological Investment, and Institutional Quality in Promoting Economic Complexity

by
Sinazo Ngqoleka
1,
Thobeka Ncanywa
2,
Zibongiwe Mpongwana
1 and
Abiola John Asaleye
1,*
1
Faculty of Economic and Financial Sciences, Walter Sisulu University, Private Bag X1 UNITRA, Mthatha 5117, South Africa
2
Directorate of Research Development and Innovation, Walter Sisulu University, Private Bag X1 UNITRA, Mthatha 5117, South Africa
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(15), 7021; https://doi.org/10.3390/su17157021 (registering DOI)
Submission received: 5 July 2025 / Revised: 28 July 2025 / Accepted: 29 July 2025 / Published: 1 August 2025

Abstract

This study examines the role of human capital, technological investment, and institutional quality in promoting economic complexity in South Africa, with implications for sustainable development and the strategic role of Small and Medium Enterprises. Motivated by the growing importance of productive sophistication for long-term development in emerging economies (notably SDG 8 and SDG 9), the study examines both long-run and short-run dynamics using the Autoregressive Distributed Lag approach, with robustness checks via Fully Modified Least Squares, Dynamic Least Squares, and Canonical Cointegration Regression. Structural Vector Autoregression is employed to assess the persistence of shocks, while the Toda–Yamamoto causality test evaluates causality. The results reveal that institutional quality significantly enhances economic complexity in the long run, while technological investment exhibits a negative long-run impact, potentially indicating absorptive capacity constraints within industries. Though human capital and income per capita do not influence complexity in the long run, they have short-term effects, with income per capita having the most immediate influence. Variance decomposition shows that shocks to technological investment are essential for economic complexity, and are the most persistent, followed by human capital and institutional quality. These findings show the need for institutional reforms that lower entry barriers for SMEs in industries, targeted innovation policies that support upgrading, and human capital strategies aligned with driven industrial transformation. The study offers insights for policymakers striving to influence structural drivers to advance sustainable industrial development and achieve the SDGs.

1. Introduction

The ability of nations to diversify their industrial base and move up the technological processes is an important determinant of long-term economic growth [1,2]. It is more pressing for emerging economies such as South Africa, which continue to attempt to overcome structural dependence on low-complexity, resource-based exports despite decades of post-apartheid economic reforms [3,4]. Industrial diversification, the expansion into more sophisticated economic activities, is widely recognised as a pathway to enhanced productivity, employment generation, and sustained competitiveness [5,6]. The concept of economic complexity, as measured by the Economic Complexity Index (ECI), provides a framework for assessing a country’s productive capabilities and knowledge intensity embedded in its export structure. Unlike conventional growth models that rely on factor accumulation, economic complexity stresses the importance of capabilities, knowledge, and institutional quality in enabling countries to move into more complex sectors. For South Africa, which has historically exhibited a modest level of economic complexity relative to other developing economies, it is essential to know what drives or constrains its industrial diversification, both for academically important and substantial policy relevance.
The present study is warranted on four grounds as follows: there exists a gap in the literature, with minimal empirical investigation of the determinants of economic complexity within Sub-Saharan Africa, particularly in South Africa, despite its strategic role on the continent; South Africa continues to face persistent structural and policy challenges, including low levels of industrial diversification, high unemployment, and implementation bottlenecks in its industrial policy frameworks, issues that demand empirical study; existing research often neglects interrelated drivers of economic complexity, such as human capital formation, technological investment, and institutional quality; finally, this study directly aligns with several Sustainable Development Goals (SDGs), and these motivations are explained below.
Despite the growing academic interest in economic complexity as a framework to promote structural transformation and development potential, there remains a noticeable empirical void concerning its drivers in Sub-Saharan Africa. Much of the existing literature has focused predominantly on high-income or rapidly industrialising economies in Asia and Latin America, using a panel data set or on growth [7,8,9,10], while others have focused on sustainability and environmental quality [11,12,13]. This approach limits such analyses in developing economies, especially where unemployment and inadequate capital formation are recognised as problems and presumed economic upgrading is important, by overlapping structural constraints. In emerging economies, industrial advancement is rarely driven by a single factor but could be a result of education systems, technological capability, and institutional strength. The African region, despite facing some of the most pressing development challenges, has received limited attention.
South Africa presents an as-yet underexamined study: as the most industrialised and diversified economy on the continent, it holds strategic importance for regional integration, trade, and innovation diffusion. South Africa continues to face structural and policy constraints that hinder its economic transformation. In 2023, the industry sector, including manufacturing, mining, construction, electricity, water, and gas, accounted for only 24.6 per cent of GDP, significantly down from 36 per cent in the 1990s [14]. In 2023, South Africa’s manufacturing sector contributed approximately 13 per cent of GDP [15]. While unemployment remains exceptionally high, with the official rate at 32.1 per cent in the third quarter of 2024 and youth unemployment for those aged 15–24 nearing 60.2 per cent, the situation remains a pressing socio-economic concern [16]; these figures show a structural disconnect between South Africa’s industrial base and its capacity to absorb its labour force. Moreover, government-led industrial strategies such as the Industrial Policy Action Plan (IPAP) and the Economic Reconstruction and Recovery Plan (ERRP) have been hampered by governance inefficiencies, institutional fragmentation, and poor coordination across innovation and education systems, undermining their effectiveness [17]. Consequently, empirical investigation into the binding constraints on industrial diversification and structural upgrading may help in informing more effective policies to improve the situation.
Finally, this study aligns with several United Nations Sustainable Development Goals (SDGs), strengthening its relevance on policy and global development impact. Through the determinants of economic complexity, particularly human capital formation, technological advancement, and institutional quality, the study contributes directly to SDG 8 (Decent Work and Economic Growth), which emphasises productive employment and sustainable economic development [18,19,20,21]. It also supports SDG 9 (Industry, Innovation and Infrastructure), which calls for inclusive industrialisation and innovation-driven growth, both of which are foundational to building knowledge-based economies [22,23,24]. The emphasis on governance and institutional effectiveness supports SDG 16 (Peace, Justice and Strong Institutions) [25], recognising that strong regulatory and institutional environments are necessary for enabling structural transformation.
Based on the following, the main objective of this study is to empirically investigate the key determinants of economic complexity in South Africa, with a focus on human capital development, technological investment, institutional quality, and trade openness. Specifically, the study aims to achieve the following:
i.
Quantify the long-term impact of human capital, technological investment, and institutional quality on economic complexity.
ii.
Analyse the short-run effect of human capital, technological investment, and institutional quality on economic complexity.
iii.
Examine the direction of causality to identify whether improvements in human capital, institutions, and technology drive economic complexity or vice versa.
iv.
Analyse the dynamic response and magnitude of shocks to key determinants such as human capital, technological investment, and institutional quality on economic complexity in South Africa
The remainder of the paper is structured as follows: Section 2 reviews the theoretical exposition and development of hypotheses. Section 3 outlines the methodological framework. Section 4 presents the empirical results and discussion of findings. Section 5 concludes with policy recommendations.

2. Theoretical Exposition and Development of Hypotheses

2.1. Theoretical Exposition

Economic complexity refers to the amount of productive knowledge in an economy and the ability of a country to produce a diverse and sophisticated set of goods. As Hidalgo and Hausmann [26] conceptualise, complexity is a function of resource endowments that arise from a network of capabilities that allow countries to participate in global value chains with increasingly complex products. The Economic Complexity Index quantifies these capabilities and has proven to be a strong predictor of economic growth [27]; this suggests that long-term development hinges on the accumulation of know-how and productive diversity, both of which are influenced by structural variables, including human capital, technological investment, and institutional quality.
Human capital is foundational to economic complexity because it provides the cognitive and technical skills required to produce, adapt, and innovate goods [28]. The augmented Solow model posits that human capital accumulation raises the steady-state level of income by enhancing labour productivity [29]. Regarding this complexity, human capital facilitates the absorption of existing knowledge and enables domestic innovation and the creation of new products; empirical studies have confirmed this link [30,31,32]. Technological investment complements human capital and promotes the application of knowledge in production. Neo-Schumpeterian theories argue that innovation-driven growth requires both technological capacity and absorptive capabilities [33]. Technological advancements allow firms to improve productivity, enter new industries, and produce technological export goods. Ávila [34] emphasises the role of absorptive capacity, defined as the ability to recognise, assimilate, and apply new external knowledge, which is related to the quality of human capital. As such, investments in R&D yield higher returns in environments where skilled employment and institutions support innovation diffusion.
Institutional quality plays a role in determining whether human and technological resources can be effectively mobilised [35]. Institutions influence economic performance by establishing the rules, incentives, and enforcement mechanisms that govern economic behaviour. According to Hussain, Bhatti, Ahmad, and Nawaz [36], inclusive institutions that secure property rights, ensure contract enforcement, and encourage investment are foundational for long-run development. Recent empirical evidence supports the view that institutional quality positively correlates with economic complexity [7,37]. Vu [7] finds that governance indicators are robust predictors of a country’s ability to produce and export complex products. Strong institutions reduce transaction costs, increase investor confidence, and facilitate long-term planning, factors essential for sustaining innovation and complexity growth.

2.2. Empirical Literature Gap and Development of Hypotheses

Several studies have examined the determinants and implications of economic complexity. A vast number of studies have focused on environmental sustainability [11,12,13]. Others have focused on panel studies and on growth [27,38,39,40]. For example, Vu [7] employed a panel of 115 countries to examine the role of institutional quality in influencing economic complexity. Rivera et al. [8] extended this by focusing on the relationship between human capital institutionalisation and globalisation in developed economies and Latin America. Other contributions, such as Rafei et al. [11] and Feng et al. [12], investigated the implications of economic complexity for environmental quality and sustainability. Meanwhile, Zhu and Li [9] examined the joint effect of economic complexity and human capital on economic growth, and Nguea [10] showed the mediating role of ICT, human capital, and FDI in the relationship between demographic dividends and economic complexity. Although these studies provide insights, the literature remains dominated by cross-country panel analyses, with limited attention given to single-country case studies, particularly within developing economies. Moreover, the specific role of technological investment as a key driver of structural transformation and innovation on economic complexity is still under-examined.
Based on the identified gaps in the empirical literature, particularly the underrepresentation of single-country studies in developing economies and the limited focus on technological investment, this study develops and tests a set of hypotheses within a developing economy, using South Africa as a case study. South Africa presents a compelling case due to its relatively high level of economic complexity in Sub-Saharan Africa, its ongoing investments in research and innovation, and its institutional and structural dualities. Moreover, South Africa’s national development agenda places strategic emphasis on technological advancement and industrial upgrading, yet empirical evidence on how these efforts translate into higher economic complexity remains scarce. Thus, this study fills the gap and provides policy-relevant insights for South Africa and other emerging economies seeking to move up the global value chain. The hypotheses are as follows:
H1. 
In the long run, increases in human capital, technological investment, and institutional quality have a statistically significant and positive impact on economic complexity.
This hypothesis is grounded in endogenous growth theory and economic complexity theory, which posit that productive knowledge accumulates through education, innovation, and governance quality [26,41]. Empirical studies suggest that long-term structural transformation is conditional on these inputs [7,37].
H2. 
In the short run, variations in human capital, technological investment, and institutional quality exert significant and positive effects on economic complexity.
While the long-run drivers of economic complexity are well established, short-run dynamics remain less examined. The short-run effects are likely to capture transitory impacts such as policy changes, institutional reforms, or technological adoption shocks. It is expected that changes in these explanatory variables will affect economic complexity [13,40]. For instance, while technology shocks may manifest quickly through digital penetration or innovation activities, human capital improvements may show a delayed effect due to training and labour market absorption frictions. Similarly, institutional reforms may require time before influencing productive capabilities and complexity.
H3. 
There is causality from human capital, institutional quality, and technological investment to economic complexity.
These hypotheses are rooted in both theoretical and empirical ambiguity in the literature. While many studies assume that human capital, institutions, and technology are exogenous drivers [42,43,44,45], some suggest feedback effects wherein higher complexity induces institutional reforms or technological spillovers [38,46].
H4. 
Economic complexity exhibits a delayed response to shocks in human capital, technological investment, and institutional quality, with institutional shocks having the most persistent effects.
This hypothesis acknowledges developing economies’ developmental process through institutional duality, high skill inequality, and uneven technological diffusion. The literature on shocks suggests that the magnitude and persistence of shocks vary across structural drivers, and economic complexity may be particularly sensitive to governance quality and other economic factors [7,39].

3. Methodological Approach

3.1. Theoretical Framework and Model Specification

We begin by conceptualising economic complexity as a function of human capital, technological investment, and institutional quality, along with income control. We assume a log-linear production structure of complexity, consistent with endogenous growth theory and capability accumulation models:
E C I t = A H M C t α 1 T E C t α 2 I N S t α 3 G D P t α 4
In Equation (1), E C I is economic complexity, H M C is human capital, T E C is technological investment, I N S is institutional quality, and G D P is income per capita.
I n E C I t = I n A + α 1 I n H M C t + α 2 I n T E C t + α 3 I n I N S t + α 4 I n G D P t + ε t
In Equation (2), α 1 , α 2 , α 3 , and α 4 are elasticities to be estimated, while A is a constant reflecting unobserved productivity or structural characteristics; this specification is grounded in a Cobb–Douglas-style function, treating economic complexity as a quasi-output of accumulated intangible capital. To achieve objectives 1 and 2, which are to investigate the long-run and short-run impacts, Equation (2) is estimated using auto autoregressive distributed lag (ARDL). The ARDL ( p , q 1 , q 2 , q 3 , q 4 ) framework is given as follows:
Δ I n E C I t = ϕ 0 + i = 1 p ϕ 1 i Δ I n E C I t i + j = 0 q 1 ϕ 2 j Δ I n H M C t j + k = 0 q 2 ϕ 3 k Δ I n T E C t k + l = 0 q 3 ϕ 4 l Δ I n I N S t i + m = 0 q 4 ϕ 1 i Δ I n G D P t m + λ 1 I n E C I t 1 + λ 2 I n H M C t 1 + λ 3 I n T E C t 1 + λ 4 I n I N S t 1 + λ 5 I n G D P t 1 + ψ E C T + μ t
In Equation (3), the first row of terms (Δ) captures the short-run effects of changes in economic complexity and its determinants. The second row of level terms represents the lagged level variables, which are used to estimate the long-run relationship among the variables. The coefficient on ψ on E C T is expected to be negative and significant to confirm the presence of a long-run equilibrium relationship. Given the moderate sample size associated with quarterly data spanning from 1996 to 2023 (approximately 112 observations), the Akaike Information Criterion (AIC) is preferred for lag selection in ARDL modelling. AIC is less stringent in penalising additional parameters, thereby allowing for a more flexible dynamic structure, which is particularly advantageous in small to medium samples.
Since the series is integrated of orders (1) and (0), we used the Toda–Yamamoto (TY) procedure for causality; if we assume X t and Y t series, the framework is given as follows:
Y t = α o + i = 1 p α i Y t i + i = 1 p β i X t i + j = p + 1 p + d max γ j Y t j + j = p + 1 p + d max δ j X t j + e t
In Equation (4), p is the optimal lag length from a VAR model. d max is the maximum order of integration among X t and Y t . The X variable can have a similar model to Equation (4) and four outcomes are possible as follows: unidirectional causality for either X or Y, bidirectional causality, and independence.
The shock effect of human capital, technological investment, and institutional quality on economic complexity is examined using the structural vector autoregression (SVAR); the series are orderly in the framework from empirical and theoretical justification. Institutional quality is placed first, consistent with the new institutional economics literature, which posits that stable and effective institutions influence the long-term incentives and governance frameworks underpinning human capital formation, income growth, and technological progress [47,48]. Human capital is ordered second, as it is influenced by institutional quality and plays a foundational role in enhancing productivity and absorptive capacity but does not contemporaneously alter institutional structures [49]. Income per capita follows, reflecting its dependence on both education and institutional quality, while acting as a resource base that supports technological investment. Technological investment is positioned fourth, as it reacts to income levels and skill availability and can change more rapidly than structural variables. Finally, economic complexity is treated as the most endogenous variable, capturing the cumulative effects of institutional quality, human capital, income, and technological capacity, in line with the economic complexity framework [26]. The structural form of the SVAR model is written as follows:
A o Y t = i = 1 p A i Y t i + ε t
Furthermore Equation (5) can be expressed as follows:
a 11 0 0 0 0 a 21 a 22 0 0 0 a 31 a 32 a 33 0 0 a 41 a 42 a 43 a 44 0 a 51 a 52 a 53 a 54 a 55 I n I N S t I n H M C t I n G D P t I n T E C t I n E C I t = i = 1 p a 11 a 12 a 13 a 14 a 15 a 21 a 22 a 23 a 24 a 25 a 31 a 32 a 33 a 34 a 35 a 41 a 42 a 43 a 44 a 45 a 51 a 52 a 53 a 54 a 55 I n I N S t i I n H M C t i I n G D P t i I n T E C t i I n E C I t i + ε 1 t ε 2 t ε 3 t ε 4 t ε 5 t
In Equation (6), A 0 is the contemporaneous structural matrix, with recursive (lower triangular) restrictions imposed for identification. A i are lagged structural coefficient matrices of order p. ε t is the vector of orthogonal structural shocks, assumed to have a diagonal covariance matrix. Finally, we carried out a robustness check using dynamic least squares and modified least squares.

3.2. Preliminary Analyses and Information About the Series

Prior to the implementation of the structural VAR and ARDL models, a series of preliminary analyses were conducted to assess the statistical properties and dynamic relationships among the variables. Descriptive statistics are first employed to examine the distributional features; this is complemented by a correlation analysis to examine initial associations between economic complexity and its key determinants: human capital, technological investment, institutional quality, and income per capita. To establish the time series properties of the data, unit root tests are performed using both the Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) approaches to determine the order of integration. Recognising the possibility of long-run equilibrium relationships among the variables, the Johansen cointegration test is also applied, allowing for the identification of one or more cointegrating vectors in the multivariate system; these procedures are essential for guiding the most appropriate technique to use, ensuring that the assumptions of stationarity and cointegration are met before proceeding with the structural and dynamic modelling frameworks.
This study spans the period from 1996 Q1 to 2023 Q4, constrained by the availability of economic complexity data, which begins in 1995. As the original data were annual, they were converted to quarterly frequency using the quadratic match sum method, a widely accepted approach for temporal disaggregation [50,51]. Economic complexity data were obtained from the Atlas of Economic Complexity, Centre for International Development, Harvard University. GDP per capita (constant local currency units) was sourced from the World Bank’s national accounts data, while income data were retrieved from OECD National Accounts. Government expenditure on education (as a percentage of total government expenditure), used to proxy human capital, was obtained from the UNESCO Institute for Statistics (UIS). Technological investment is represented by research and development expenditure (% of GDP), also sourced from UIS. Lastly, institutional quality is proxied by Regulatory Quality, drawn from the World Bank’s Worldwide Governance Indicators (WGI). All data used in the analysis are country-specific and pertain exclusively to South Africa.

4. Empirical Results and Discussion of Findings

4.1. Preliminary Results

Table 1 presents summary statistics for the main variables. The Economic Complexity Index (ECI) has a mean of 0.3028 and a standard deviation of 0.1919, suggesting variation in South Africa’s productive structure. GDP per capita shows limited dispersion (SD = 0.0493), indicating relative income stability. Human capital (HMC) displays a consistently high average (1.2609) with low variability, while institutional quality (INS) shows a wider spread (SD = 0.2979), indicating uneven governance performance. Technological investment (TEC) is negative on average (−0.1767), with moderate variability, pointing to relatively low innovation inputs. Correlation results indicate that ECI is strongly positively associated with institutional quality (r = 0.8392) but negatively correlated with GDP per capita (r = −0.6653), suggesting a potential structural disconnect between income and economic complexity. Weak correlations are observed between ECI and both human capital (r = −0.038) and technological investment (r = −0.1166), warranting further empirical investigation.
Table 2 reports the stationarity results based on the Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests under three specifications: constant, constant with trend, and none. In the ADF test, the Economic Complexity Index (ECI) is stationary at level under the constant and trend and none specifications at the 1% significance level. Human capital (HMC) and technological investment (TEC) are stationary at level under the constant specification at the 10% significance level. All variables achieve stationarity at first difference across all specifications. Similarly, the PP test confirms that ECI is stationary at level under the constant and trend specification at the 1% level, with all variables becoming stationary at first difference in all cases. Given these outcomes, objectives 1 and 2 are estimated using the Autoregressive Distributed Lag (ARDL) approach, with Fully Modified OLS (FMOLS), Dynamic OLS (DOLS), and Canonical Cointegration Regression (CCR) employed for robustness checks. Objective 3 is addressed using the Toda–Yamamoto causality approach, while Objective 4 applies the Structural Vector Autoregression (SVAR) model.
Table 3 presents the ARDL bounds test results and model specification. The computed F-statistic is 3.5104. Compared with the critical bounds, this value exceeds the upper bound at the 10% level (I (1) = 3.09) and the 5% level (I (1) = 3.49), indicating evidence of a long-run relationship among the variables at these significance levels. However, it falls below the 2.5% and 1% upper bounds, suggesting marginal evidence of cointegration at stricter thresholds. Model selection was guided by the Akaike Information Criterion (AIC), with the top 20 candidate models summarised in Appendix A Figure A1. The optimal specification selected is ARDL (3, 2, 2, 2, 2).

4.2. Autoregressive Distributed Lag Result

Table 4 presents the ARDL estimates for the long-run and short-run dynamics of economic complexity in South Africa. In the long run, institutional quality (INS) and technological investment (TEC) are statistically significant at the 1% and 5% levels, respectively. Specifically, a 1% increase in institutional quality is associated with a 0.0488% rise in economic complexity, indicating the role of governance structures in sustaining productive diversification. Conversely, a 1% increase in technological investment is associated with a 0.1627% decline in economic complexity, a counterintuitive result possibly because of inefficient deployment, technological mismatch, or absorptive capacity constraints in the economy. Other variables, human capital (HMC) and income per capita (GDP), are not significant in the long run, suggesting their effects are either short-term or operate indirectly through other channels; these finding contradicts endogenous growth theory, which posits that productive knowledge accumulates through education and innovation is important to promote development, although the implication of governance quality aligned with the study’s findings [41,44,52,53]. Empirical studies suggest that long-term structural transformation is conditional on these inputs [7,37].
In the short run, all variables significantly affect economic complexity at the 1% level. Notably, a 1% increase in GDP leads to a 2.19% increase in economic complexity, indicating strong contemporaneous responsiveness of economic complexity to income growth. Institutional quality also remains positively associated (0.3375%), while human capital (−0.7151%) and technological investment (−1.1693%) show negative short-run effects. These may reflect time lags in returns from education and technology or transitional inefficiencies; these outcomes align with the literature, which shows that the short-run effects are likely to capture transitory impacts such as policy changes, institutional reforms, or technological adoption shocks. It is expected that changes in these explanatory variables will affect economic complexity [13].
The error correction term (CointEq (−1)) is negative (−0.0803) and significant at the 1% level, confirming the existence of a stable long-run relationship. Approximately 8% of any short-run disequilibrium is corrected within a period, strengthening the model’s convergence behaviour. The model displays goodness of fit (R2 = 0.9944; adjusted R2 = 0.9935) and overall explanatory power (F-statistic = 1103.937, p < 0.01). The Durbin–Watson statistic of 2.086 also indicates no serial correlation. The model passes key diagnostic tests. The Breusch–Godfrey LM test indicates no serial correlation (p = 0.5062), and the ARCH test confirms homoskedasticity (p = 0.8352). Structural stability is verified, supporting the validity of the model specification as shown in Figure A2 in Appendix A. Although the Jarque–Bera test indicates residuals are not normally distributed (shown in Figure A3), this does not undermine model reliability. As emphasised in prior literature [54], the ARDL framework is robust to non-normality, especially in large samples with stable dynamics and valid residual diagnostics.
These findings provide support for Hypothesis 1, confirming that institutional quality and technological investment exert significant long-run effects on economic complexity. The positive long-run elasticity of institutional quality strengthens the role of governance in enabling structural economic transformation, particularly through mechanisms that promote stability, transparency, and accountability. However, the negative coefficient for technological investment, while counterintuitive, raises important questions about efficiency, sectoral targeting, and absorptive capacity related to technological expenditures. This suggests that merely increasing investment levels is insufficient without strategic alignment to national capabilities and developmental goals, a principle central to pragmatic sustainability, which emphasises the adaptive implementation of policy interventions. Hypothesis 2, concerning short-run dynamics, is also supported. All variables demonstrate statistically significant influence in the short term, with GDP exhibiting the most pronounced immediate effect. The capacity of income fluctuations to support structural upgrades is shown in the result, particularly where economic momentum can be rapidly mobilised. However, the observed negative elasticities for human capital and technological investment in the short run suggest the presence of transitional frictions. These may include adjustment costs, mismatches in skills and technologies, or short-term reallocations that temporarily detract from productive complexity. Such findings point to the importance of sustained, phased investments in education and innovation systems, components of sustainable communities that deliver cumulative benefits over time rather than immediate gains. Therefore, the findings emphasise the need for an integrated, long-term strategy that bridges short-run responsiveness with long-run developmental vision.

4.3. Toda–Yamamoto Causality Approach

Table 5 presents the results of the Toda–Yamamoto causality test, which assesses the directional relationships among variables while avoiding pre-testing biases associated with traditional Granger causality. The null hypothesis of “no causality” was rejected only in two cases: Gross Domestic Product (GDP) causes Institutional Quality (INS), and Human Capital (HMC) causes Institutional Quality. In both cases, the causality is unidirectional, indicating that past values of income per capita (GDP) and human capital (HMC) help predict changes in institutional quality, but not vice versa. No causal link is found from Human Capital, Institutional Quality, or Technological Investment to Economic Complexity Index (ECI) under this test. These results suggest that Institutional Quality acts as an influence from economic growth and human development, rather than a primary initiator of structural economic change, in the short to medium term. The absence of causality from Human Capital, Institutional Quality, or Technological Investment to Economic Complexity implies that the transmission mechanisms from these structural drivers to economic complexity are not immediate or linear and may be mediated through other channels or operate predominantly in the long run; this aligns with the earlier ARDL findings, where only Institutional Quality and Technological Investment were significant in the long-run equation. The contrast suggests that while these variables contribute to long-term structural transformation, their short-run predictive power (as captured in causality tests) may be limited by lagged effects, endogeneity, or transitional rigidities.
The hypothesis positing causal relationships from Human Capital, Institutional Quality, and Technological Investment to Economic Complexity is not supported by the Toda–Yamamoto test results. Although the ARDL model reveals long-run associations, the short-run Granger-causality analysis fails to establish direct predictive linkages. This asymmetry between short- and long-term dynamics suggests that the transformation of human capital and institutional capacity into productive sophistication is not immediate but rather contingent upon extended time horizons, potential interaction effects, and the presence of enabling or complementary policy frameworks. These findings indicate the importance of adopting a pragmatic sustainability approach, which recognises that sustainable development outcomes such as structural transformation and innovation-driven growth often require gradual, iterative processes rather than rapid causal impacts. Moreover, the delayed influence of these foundational variables further reveals the need for long-term policy commitments aimed at building sustainable communities through integrated investments in education systems, institutional trust, and innovation. Short-run policy evaluations may underestimate the latent, compounding effects of reforms that are essential to achieving sustained economic complexity and inclusive development. The findings are not aligned with the studies that stressed that institutions and technology are exogenous drivers [42,43]. Some suggest feedback effects wherein higher complexity induces institutional reforms or technological spillovers [38,46].

4.4. Structural Vector Autoregression Results

Table 6 reports the results of variance decomposition derived from the Structural Vector Autoregression (SVAR) model, while the corresponding impulse response functions are illustrated in Figure A4 in Appendix A. The analysis decomposes the forecast error variance of Economic Complexity Index (ECI) over a 10-period horizon, focusing on the relative contributions of shocks from Technological Investment (TEC), Human Capital (HMC), and Institutional Quality (INS). The findings indicate that technological investment shocks account for the largest share of forecast error variance in ECI, rising from 0.19% in period 2 to approximately 1.71% by period 10; this suggests that the influence of technology on economic complexity becomes progressively more prominent over time, though it remains moderate in absolute terms. Human capital shocks contribute the second-largest variance, increasing from 0.09% in period 2 to 1.21% by period 10. Interestingly, institutional quality shocks contribute the least, with values rising from just 0.003% to 0.40% over the same horizon. These results reveal that economic complexity in South Africa responds gradually to structural shocks, supporting the notion of a delayed transmission mechanism. The relatively low and slow-building contributions of these variables show the structural nature of economic complexity, which evolves through cumulative capability formation rather than short-term impulses.
While technological investment emerges as the most influential among the three examined factors, its sustained impact shows the necessity for continuous and long-term R&D commitments to drive advancements in economic complexity; this aligns with the principles of pragmatic sustainability, where incremental innovation becomes a pathway toward development. The contribution of human capital, although comparatively slower to manifest, stresses the importance of enduring investments in education and skills development, core elements in promoting sustainable communities capable of adapting to economic demands. Contrary to prevailing assumptions, institutional quality demonstrated the least immediate effect on productive sophistication. This unexpected outcome may suggest that institutional reforms require longer gestation periods to translate into measurable improvements in economic complexity. Alternatively, the influence of governance may be indirect, operating through its effect on investment, firm-level behaviour, and societal trust. In this regard, the findings call for improvements in institutional frameworks, particularly in striving for sustainable transitions through layered, interdependent mechanisms of change.
The stated hypothesis that economic complexity exhibits a delayed response to shocks in human capital, technological investment, and institutional quality, with institutional shocks having the most persistent effects, is only partially supported. The delayed response aspect is validated by the gradual accumulation of variance shares across all variables. However, technological investment, not institutional quality, contributes the largest and most persistent share over the 10-period horizon. Institutional shocks, while persistent, exhibit the lowest magnitude, suggesting that their effect, though stable, is less direct or slower to internalise within the productive structure. The findings are also supported by previous literature, which suggests that the magnitude and persistence of shocks vary across structural drivers, and economic complexity may be particularly sensitive to governance quality and other economic factors [7,39].

4.5. Robustness Checks Using FMOLS, DOLS, and CCR

Table A1 in Appendix A presents robustness checks for the long-run estimations using Fully Modified Ordinary Least Squares (FMOLS), Dynamic Ordinary Least Squares (DOLS), and Canonical Cointegration Regression (CCR). These alternative estimators account for potential endogeneity and serial correlation, thereby enhancing the credibility of the ARDL-based findings. Across all three estimations, the results confirm that Institutional Quality (INS) is positively and significantly associated with Economic Complexity Index (ECI), while Technological Investment (TEC) is negatively and significantly related to ECI. In the FMOLS estimation, the coefficient for institutional quality is 0.6527, while that of technological investment is −1.6932. Similarly, in the DOLS model, institutional quality has a coefficient of 0.6249, and technological investment is estimated at −1.5823. The CCR model produces nearly identical results, with institutional quality estimated at 0.6507 and technological investment at −1.6813. In all three models, the coefficients for Human Capital (HMC) and Income per capita (GDP) are statistically insignificant, in line with the ARDL long-run results. Moreover, each model reports an R-squared and adjusted R-squared value exceeding 0.50, indicating a good level of explanatory power. The consistent results across multiple estimators enhance the robustness and reliability of the long-run findings. They reaffirm the role of institutional quality in promoting economic complexity, while also confirming the persistent negative association between technological investment and complexity, possibly due to inefficiencies or misalignment in technology deployment.

5. Conclusions and Policy Recommendations

This study was motivated by the growing recognition that economic complexity, a measure of a country’s capacity to produce and export sophisticated and knowledge-intensive products, is central to long-term structural transformation and sustainable development. Despite the increasing policy focus on innovation and governance, limited empirical work has examined how these structural factors influence economic complexity in emerging economies. To fill this gap, the study pursued four objectives in South Africa as follows: it assessed the long-run impact of human capital, institutional quality, and technological investment on economic complexity; it examined the short-run effects of income per capita, human capital, institutional quality, and technological investment on economic complexity; it evaluated whether economic complexity exhibits a delayed response to shocks originating from these structural drivers; and finally, it investigated the causal relationships between human capital, institutional quality, technological investment, and economic complexity. To achieve these objectives, the study employed the Autoregressive Distributed Lag model to estimate both long-run and short-run relationships. Robustness checks were conducted using Fully Modified OLS, Dynamic OLS, and Canonical Cointegration Regression. Structural Vector Autoregression was used for impulse response and variance decomposition analysis, while the Toda–Yamamoto causality framework was used to test for directional causality.
The empirical findings reveal that in the long run, institutional quality emerged as a statistically significant and positive determinant of economic complexity, suggesting that stable and effective governance structures facilitate the development of complex production capabilities. Technological investment, by contrast, showed a negative and statistically significant relationship, implying that current innovation expenditures may be inefficient, poorly managed, or undermined by weak absorptive capacity. Human capital and income per capita were not significant in the long-run equation, indicating that their impact on complexity may be indirect or delayed. In the short run, all variables, income per capita, human capital, institutional quality, and technological investment, exhibited statistically significant effects on economic complexity. Income per capita showed the largest elasticity, suggesting that macroeconomic gains can lead to immediate improvements in productive sophistication. However, human capital and technological investment had negative short-run coefficients, which may reflect transitional inefficiencies or lagged effects in translating investments to benefit the economy.
The SVAR-based variance decomposition analysis shows the delayed structural adjustment. Over a ten-period horizon, shocks from technological investment accounted for the largest share of forecast error variance in economic complexity, followed by shocks from human capital and then institutional quality; this finding suggests that while all three drivers matter over time, their influence accumulates slowly, indicating the need for sustained and coordinated structural policy interventions. Surprisingly, the results from the Toda–Yamamoto causality tests did not support direct causality from human capital, institutional quality, or technological investment to economic complexity. However, it found unidirectional causality from income per capita and human capital to institutional quality, suggesting that improvements in governance may, in fact, be driven by economic and educational progress rather than preceding them. While the ARDL model shows a significant relationship between institutional quality and economic complexity, the Toda–Yamamoto causality test does not confirm a short-run causal link. This suggests that institutional reforms may influence complexity through longer-term, indirect channels rather than immediate predictive effects. Thus, institutional quality remains important, but its impact unfolds gradually over time.
Based on these findings, the study recommends that in South Africa and other emerging economies with similar structural characteristics, strengthening institutional quality remains a very important aspect to focus on in enhancing economic complexity. Reforms aimed at improving the regulatory environment, ensuring judicial independence, and enhancing public sector efficiency are likely to yield long-term benefits for industrial transformation and SMEs. The negative impact of technological investment stresses the need for better coordination between innovation spending and sectoral development strategies. R&D investments must be increased and aligned with domestic industrial priorities, which is supported by mechanisms that ensure effective diffusion and commercialisation. The results suggest that human capital formation, while essential, is not sufficient. Education and training systems must meet the skill demands of high-complexity industries. Moreover, given the strong short-run impact of income per capita, periods of economic growth should aim to promote structural upgrading through public-private partnerships and industrial policies. Finally, the slow-moving nature of these structural transformations points to the need for long-term policy planning and institutional coordination, enhancing absorptive capacity between governance and innovation, and introducing reforms to improve South Africa’s economic complexity in a sustained and inclusive manner.
This study contributes to the ongoing discourse on sustainable development by empirically examining the long- and short-run determinants of economic complexity, with a focus on human capital, institutional quality, and technological investment. Our findings support the long-term importance of institutional strength and innovation in promoting structural transformation, while also highlighting short-run sensitivities to income fluctuations and transitional frictions in education and technology deployment. In doing so, the study advances the objectives of several Sustainable Development Goals, particularly SDG 4 (Quality Education), SDG 8 (Decent Work and Economic Growth), SDG 9 (Industry, Innovation and Infrastructure), and SDG 16 (Peace, Justice and Strong Institutions).
Despite the importance of this study, it is worth mentioning that this study does not investigate interaction effects between the structural drivers, such as whether institutional quality moderates the impact of technological investment or human capital on economic complexity. Future research could address these gaps by incorporating interaction terms or employing nonlinear models such as threshold regression or structural equation modelling to better capture the interdependencies among policy variables.

Author Contributions

Conceptualization, S.N., T.N. and Z.M.; Methodology, T.N. and A.J.A.; Software, T.N. and A.J.A.; Validation, T.N. and Z.M.; Formal analysis, T.N. and Z.M.; Investigation, S.N., T.N., Z.M. and A.J.A.; Resources, S.N., T.N. and A.J.A.; Data curation, T.N., Z.M. and A.J.A.; Writing—original draft, S.N. and T.N.; Writing—review & editing, T.N., Z.M. and A.J.A.; Visualization, T.N.; Supervision, T.N. and Z.M.; Project administration, S.N., T.N. and Z.M.; Funding acquisition, A.J.A. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix A

Figure A1. Model selection model.
Figure A1. Model selection model.
Sustainability 17 07021 g0a1
Figure A2. Stability test.
Figure A2. Stability test.
Sustainability 17 07021 g0a2
Figure A3. Normality test.
Figure A3. Normality test.
Sustainability 17 07021 g0a3
Figure A4. Impulse response function. Source: authors’ computation.
Figure A4. Impulse response function. Source: authors’ computation.
Sustainability 17 07021 g0a4
Table A1. Robustness check.
Table A1. Robustness check.
Fully Modified Least Squares (FMOLS)
Dependent Variable: ECI
VariableCoeff.Std. Errort-StatisticProb.
HMC−0.40930.5555−0.73670.4629
INS0.65270.064510.11270.0000
TEC−1.69320.3671−4.61160.0000
GDP0.05330.42040.12680.8993
C0.01492.17880.00680.9945
R-squ.0.8841Mean dependent var0.3024
Adj. R-squ.0.8797S.D. dependent var0.1927
S.E. of reg.0.0668Sum squared resid0.4734
Long-run var.0.0164
Dynamic Least Squares (DOLS)
Dependent Variable: ECI
VariableCoeff.Std. Errort-StatisticProb.
HMC−0.62460.5896−1.05940.2918
INS0.62490.06859.11070.0000
TEC−1.58230.3897−4.06000.0001
GDP−0.02850.4461−0.06400.9491
C0.71642.31480.30950.7575
R-squ.0.8773Mean dependent var0.3028
Adj. R-squ.0.8727S.D. dependent var0.1919
S.E. of reg.0.0684Sum squared resid0.5015
Long-run var.0.0185
Canonical Cointegrating Regression (CCR)
Dependent Variable: ECI
VariableCoeff.Std. Errort-StatisticProb.
HMC−0.44610.5271−0.84620.3993
INS0.65070.063510.23500.0000
TEC−1.68130.3506−4.79420.0000
GDP0.04920.41320.11910.9054
C0.08392.14340.03910.9689
R-squ.0.8844Mean dependent var0.3024
Adj. R-squ.0.8800S.D. dependent var0.1927
S.E. of reg.0.0667Sum squared resid0.4723
Long-run var.0.0164
Source: authors’ computation.

References

  1. Li, J.; Imran, A. Sustainable Transitions: Navigating Green Technologies, Clean Energy, Economic Growth, and Human Capital for a Greener Future. Sustainability 2025, 17, 3446. [Google Scholar] [CrossRef]
  2. Ncanywa, T.; Dyantyi, N.; Asaleye, A.J. Empowerment Through Entrepreneurship: A Mixed-Methods Analysis of Social Grants and Economic Sufficiency. Economies 2025, 13, 107. [Google Scholar] [CrossRef]
  3. Bell, J.; Goga, S.; Mondliwa, P.; Roberts, S. Structural Transformation in South Africa: Moving Towards a Smart, Open Economy for All. CCRED Working Paper No. 9/2018. 2018. Available online: https://ssrn.com/abstract=3269732 (accessed on 1 July 2025).
  4. Mpiti, V.S.; Ncanywa, T.; Asaleye, A.J. Do Educators’ Demographic Characteristics Drive Learner Academic Performance? Examining the Role of Gender, Qualifications, and Experience. Educ. Sci. 2025, 15, 487. [Google Scholar] [CrossRef]
  5. Ferraz, D.; Falguera, F.P.S.; Mariano, E.B.; Hartmann, D. Linking Economic Complexity, Diversification, and Industrial Policy with Sustainable Development: A Structured Literature Review. Sustainability 2021, 13, 1265. [Google Scholar] [CrossRef]
  6. Al-Masri, R.; Ibrahim, M. 66-Integrating Green Finance, Economic Complexity, and Renewable Energy for Sustainable Development in Asia. J. Energy Environ. Policy Options 2025, 8, 66–74. [Google Scholar]
  7. Vu, T.V. Does institutional quality foster economic complexity? The fundamental drivers of productive capabilities. Empir. Econ. 2022, 63, 1571–1604. [Google Scholar] [CrossRef]
  8. Rivera, B.; Leon, M.; Cornejo, G.; Florez, H. Analysis of the Effect of Human Capital, Institutionality and Globalization on Economic Complexity: Comparison between Latin America and Countries with Greater Economic Diversification. Economies 2023, 11, 204. [Google Scholar] [CrossRef]
  9. Zhu, S.; Li, R. Economic complexity, human capital and economic growth: Empirical research based on cross-country panel data. Appl. Econ. 2017, 49, 3815–3828. [Google Scholar] [CrossRef]
  10. Nguea, S.M. Does Demographic Dividend Enhance Economic Complexity: The Mediating Effect of Human Capital, ICT, and Foreign Direct Investment. J. Knowl. Econ. 2024, 15, 19678–19699. [Google Scholar] [CrossRef]
  11. Rafei, M.; Esmaeili, P.; Balsalobre-Lorente, D. A step towards environmental mitigation: How do economic complexity and natural resources matter? Focusing on different institutional quality level countries. Resour. Policy 2022, 78, 102848. [Google Scholar] [CrossRef]
  12. Feng, Q.; Usman, M.; Saqib, N.; Mentel, U. Modelling the contribution of green technologies, renewable energy, economic complexity, and human capital in environmental sustainability: Evidence from BRICS countries. Gondwana Res. 2024, 132, 168–181. [Google Scholar] [CrossRef]
  13. Abid, I.; Gafsi, N. Economic Complexity, Environmental Sustainability, and Technological Integration in Saudi Arabia: Analyzing Long-Term Trends. Int. J. Energy Econ. Policy 2025, 15, 669–683. [Google Scholar] [CrossRef]
  14. African Development Bank Group. African Economic Outlook 2024. Driving Africa’s Transformation. The Reform of the Global Financial Architecture; African Development Bank Group: Abidjan, Côte d’Ivoire, 2024. [Google Scholar]
  15. World Bank. World Development Indicators: Manufacturing, Value Added (% of GDP)—South Africa; World Bank: Washington, DC, USA, 2024. [Google Scholar]
  16. Statistics South Africa. Statistics South Africa on the Official Unemployment Rate in the Third Quarter of 2024; Statistics South Africa: Pretoria, South Africa, 2024. [Google Scholar]
  17. Edwards, L. Trade and Industrial Policy for South Africa’s Future. Policy Paper 31. ERSA Working Paper Series. 2024. Available online: https://ersawps.org/index.php/working-paper-series/article/view/137 (accessed on 1 July 2025).
  18. Soliman, P.E.N.; Beram, R.S. Sustainable Development Goals and Unemployment: Worldwide Evidence. J. Knowl. Econ. 2025. [Google Scholar] [CrossRef]
  19. Zehri, C.; El Amin, B.M.; Kadja, A.; Inaam, Z.; Sekrafi, H. Exploring the nexus of decent work, financial inclusion, and economic growth: A study aligned with SDG 8. Sustain. Futur. 2024, 7, 100213. [Google Scholar] [CrossRef]
  20. Magida, N.; Ncanywa, T.; Sibanda, K.; Asaleye, A.J. Human Capital Development and Public Health Expenditure: Assessing the Long-Term Sustainability of Economic Development Models. Soc. Sci. 2025, 14, 351. [Google Scholar] [CrossRef]
  21. Asaleye, A.J.; Garidzirai, R.; Ncanywa, T. Electricity availability and access for economic opportunities: A disaggregated analysis of gender gaps and implications for pro-poor growth. Energy Strat. Rev. 2025, 59, 101723. [Google Scholar] [CrossRef]
  22. Mohamed, B.H.; Ari, I.; Al-Sada, M.b.S.; Koç, M. Strategizing Human Development for a Country in Transition from a Resource-Based to a Knowledge-Based Economy. Sustainability 2021, 13, 13750. [Google Scholar] [CrossRef]
  23. Asaleye, A.J.; Ojo, A.P.; Olagunju, O.E. Asymmetric and shock effects of foreign AID on economic growth and employment generation. Res. Glob. 2023, 6, 100123. [Google Scholar] [CrossRef]
  24. Asaleye, A.J.; Ogunjobi, J.O.; Ezenwoke, O.A. Trade openness channels and labour market performance: Evidence from Nigeria. Int. J. Soc. Econ. 2021, 48, 1589–1607. [Google Scholar] [CrossRef]
  25. Küfeoğlu, S. Emerging Technologies: Value Creation for Sustainable Development; Springer Nature: London, UK, 2022; Available online: https://library.oapen.org/handle/20.500.12657/57360 (accessed on 28 July 2025).
  26. Hidalgo, C.A.; Hausmann, R. The building blocks of economic complexity. Proc. Natl. Acad. Sci. USA 2009, 106, 10570–10575. [Google Scholar] [CrossRef]
  27. Zhang, C.; Waris, U.; Qian, L.; Irfan, M.; Rehman, M.A. Unleashing the dynamic linkages among natural resources, economic complexity, and sustainable economic growth: Evidence from G-20 countries. Sustain. Dev. 2024, 32, 3736–3752. [Google Scholar] [CrossRef]
  28. Bambi, P.D.R.; Pea-Assounga, J.B.B. Unraveling the interplay of research investment, educational attainment, human capital development, and economic advancement in technological innovation: A panel VAR approach. Educ. Inf. Technol. 2024, 30, 3309–3341. [Google Scholar] [CrossRef]
  29. Ogbeifun, L.; Shobande, O.A. A reevaluation of human capital accumulation and economic growth in OECD. J. Public Aff. 2022, 22, e02602. [Google Scholar] [CrossRef]
  30. Bye, B.; Fæhn, T. The role of human capital in structural change and growth in an open economy: Innovative and absorptive capacity effects. World Econ. 2022, 45, 1021–1049. [Google Scholar] [CrossRef]
  31. Guo, R.; Ning, L.; Chen, K. How do human capital and R&D structure facilitate FDI knowledge spillovers to local firm innovation? a panel threshold approach. J. Technol. Transf. 2022, 47, 1921–1947. [Google Scholar] [CrossRef]
  32. Aliasghar, O.; Sadeghi, A.; Rose, E.L. Process innovation in small- and medium-sized enterprises: The critical roles of external knowledge sourcing and absorptive capacity. J. Small Bus. Manag. 2023, 61, 1583–1610. [Google Scholar] [CrossRef]
  33. Henrekson, M.; Johansson, D.; Karlsson, J. To Be or Not to Be: The Entrepreneur in Neo-Schumpeterian Growth Theory. Entrep. Theory Pract. 2023, 48, 104–140. [Google Scholar] [CrossRef]
  34. Ávila, M.M. Competitive Advantage and Knowledge Absorptive Capacity: The Mediating Role of Innovative Capability. J. Knowl. Econ. 2022, 13, 185–210. [Google Scholar] [CrossRef]
  35. Arshed, N.; Hanif, N.; Aziz, O.; Croteau, M. Exploring the potential of institutional quality in determining technological innovation. Technol. Soc. 2022, 68, 101859. [Google Scholar] [CrossRef]
  36. Hussain, A.; Bhatti, M.A.; Ahmad, T.I.; Nawaz, M.A. The Role of Secure Property Rights in Driving Economic Growth: A Review of Recent Evidence. Curr. Trends Law Soc. 2024, 4, 119–129. [Google Scholar] [CrossRef]
  37. Ahmad, M.; Ahmed, Z.; Majeed, A.; Huang, B. An environmental impact assessment of economic complexity and energy consumption: Does institutional quality make a difference? Environ. Impact Assess. Rev. 2021, 89, 106603. [Google Scholar] [CrossRef]
  38. Chatzistamoulou, N.; Kounetas, K.; Tsekouras, K. Technological hierarchies and learning: Spillovers, complexity, relatedness, and the moderating role of absorptive capacity. Technol. Forecast. Soc. Change 2022, 183, 121925. [Google Scholar] [CrossRef]
  39. Gomez-Gonzalez, J.E.; Uribe, J.M.; Valencia, O.M. Does economic complexity reduce the probability of a fiscal crisis? World Dev. 2023, 168, 106250. [Google Scholar] [CrossRef]
  40. Stojkoski, V.; Kocarev, L. The Relationship Between Growth and Economic Complexity: Evidence from Southeastern and Central Europe 2017. Available online: https://mpra.ub.uni-muenchen.de/77837/ (accessed on 1 July 2025).
  41. Aktug, M.; Polat, S.; Basoglu, A. Fiscal Policy as a Driver of Economic Complexity: Evidence from Selected OECD Countries. J. Knowl. Econ. 2025. [Google Scholar] [CrossRef]
  42. Duh, M.; Primec, A. Investigating the Effectiveness of Endogenous and Exogenous Drivers of the Sustainability (Re)Orientation of Family SMEs in Slovenia: Qualitative Content Analysis Approach. Sustainability 2024, 16, 7285. [Google Scholar] [CrossRef]
  43. Epicoco, M. Technological Revolutions and Economic Development: Endogenous and Exogenous Fluctuations. J. Knowl. Econ. 2021, 12, 1437–1461. [Google Scholar] [CrossRef]
  44. Oloni, E.; Asaleye, A.; Abiodun, F.; Adeyemi, O. Inclusive growth, agriculture and employment in Nigeria. J. Environ. Manag. Tour. 2017, 8, 183. [Google Scholar] [CrossRef]
  45. Asaleye, A.J.; Strydom, K. Gender differential in economic participation, foreign aid and productivity: Pathway to broad-based growth. Socio-Econ. Plan. Sci. 2025, 101, 102250. [Google Scholar] [CrossRef]
  46. Krammer, S.M. Do good institutions enhance the effect of technological spillovers on productivity? Comparative evidence from developed and transition economies. Technol. Forecast. Soc. Change 2015, 94, 133–154. [Google Scholar] [CrossRef]
  47. Dias, J.; Tebaldi, E. Institutions, human capital, and growth: The institutional mechanism. Struct. Change Econ. Dyn. 2012, 23, 300–312. [Google Scholar] [CrossRef]
  48. Okolo, C.V.; Wen, J.; Susaeta, A. Maximizing natural resource rent economics: The role of human capital development, financial sector development, and open-trade economies in driving technological innovation. Environ. Sci. Pollut. Res. Int. 2024, 31, 4453–4477. [Google Scholar] [CrossRef] [PubMed]
  49. Doré, N.I.; Teixeira, A.A. The role of human capital, structural change, and institutional quality on Brazil’s economic growth over the last two hundred years (1822–2019). Struct. Change Econ. Dyn. 2023, 66, 1–12. [Google Scholar] [CrossRef]
  50. Chow, G.C.; Lin, A.-L. Best Linear Unbiased Interpolation, Distribution, and Extrapolation of Time Series by Related Series. Rev. Econ. Stat. 1971, 53, 372–375. [Google Scholar] [CrossRef]
  51. Denton, F.T. Adjustment of Monthly or Quarterly Series to Annual Totals: An Approach Based on Quadratic Minimization. J. Am. Stat. Assoc. 1971, 66, 99–102. [Google Scholar] [CrossRef]
  52. Oladipo, O.A.; Iyoha, F.; Fakile, S.A.; Asaleye, A.J.; Eluyela, F.D. Do government taxes have implications on manufacturing sector output? Evidence from Nigeria. J. Manag. Inf. Decis. Sci. 2019, 22, 181–190. Available online: https://www.researchgate.net/publication/334646219 (accessed on 1 July 2025).
  53. Asaleye, A.J.; Lawal, A.I.; Popoola, O.; Alege, P.O.; Oyetade, O.O. Financial integration, employment and wages nexus: Evidence from Nigeria. Montenegrin J. Econ. 2019, 15, 141–154. [Google Scholar] [CrossRef]
  54. Nkoro, E.; Uko, A.K. Autoregressive Distributed Lag (ARDL) cointegration technique: Application and interpretation. J. Stat. Econom. Methods 2016, 5, 63–91. [Google Scholar]
Table 1. Descriptive statistics and correlation analysis.
Table 1. Descriptive statistics and correlation analysis.
Descriptive Statistics
ECIGDPHMCINSTEC
Mean0.30284.84671.26090.3724−0.1767
Median0.33514.86761.26720.4523−0.1741
Maximum0.67384.89611.30960.8293−0.0905
Minimum−0.03584.75111.2026−0.2228−0.2933
Std. Dev.0.19190.04930.02610.29790.0572
Skewness−0.1195−0.7364−0.3825−0.4939−0.3738
Kurtosis2.24741.97742.53192.11442.1622
Jarque-Bera2.909815.0033.75388.21485.8839
Sum33.92542.8341141.229841.7133−19.7977
Sum Sq. Dev.4.08870.26990.07579.85060.3641
Obs.112112112112112
Correlation Analysis
ECIGDPHMCINSTEC
ECI1
GDP−0.66531
HMC−0.038−0.20791
INS0.8392−0.4651−0.21561
TEC−0.11660.4755−0.53920.32281
Source: authors’ computation.
Table 2. Stationarity result.
Table 2. Stationarity result.
Augmented Dickey–Fuller Test
ConstantConstant and TrendNoneConclusion
SeriesCoeff.Prob.Coeff.Prob.Coeff.Prob.
ECI−1.54870.5050−3.94690.0135−2.16280.0300Mixture
D(ECI)−3.42040.0124−4.29680.0026−3.01100.0029
GDP−1.93720.3142−0.51750.98131.00400.9162I (1)
D(GDP)−3.03190.0352−3.68140.0279−2.86020.0046
HMC−2.59280.0977−2.81270.1962−0.10330.6457Mixture
D(HMC)−5.40280.0000−5.48560.0001−5.42650.0000
INS0.33710.9792−2.16000.5062−0.95180.3024I (1)
D(INS)−4.00300.0020−4.45430.0028−3.67560.0003
TEC−2.67300.0822−2.72450.2292−0.90280.3228Mixture
D(TEC)−3.57030.0080−3.87090.0166−3.57920.0005
Phillips–Perron Test
ConstantConstant and TrendNoneConclusion
SeriesCoeff.Prob.Coeff.Prob.Coeff.Prob.
ECI−0.78010.8205−4.11620.0080−1.00650.2806Mixture
D(ECI)−5.36330.0000−5.11800.0003−5.06580.0000
GDP−2.16030.2220−0.24010.99141.73720.9798I (1)
D(GDP)−5.04140.0000−4.98180.0004−4.94840.0000
HMC−2.49440.1196−2.40490.3750−0.20010.6120I (1)
D(HMC)−6.53120.0000−8.29210.0000−6.60840.0000
INS0.64260.9903−1.64850.7672−0.98450.2893I (1)
D(INS)−5.88450.0000−5.71170.0000−6.21980.0000
TEC−1.97980.2953−1.73390.7296−0.89250.3274I (1)
D(TEC)−4.87780.0001−4.86820.0007−4.89520.0000
Source: authors’ computation.
Table 3. ARDL bound test and model selection.
Table 3. ARDL bound test and model selection.
F-Bounds Test Null Hypothesis: No Levels Relationship
Test StatisticValueSignif.I (0)I (1)
F-statistic3.510310%2.203.09
k45%2.563.49
2.5%2.883.87
1%3.294.37
Model Selection: ARDL: 3, 2, 2, 2 and 2
Source: authors’ computation.
Table 4. ARDL long- and short-run results.
Table 4. ARDL long- and short-run results.
Long-run Behaviour
Dependent Variable: ECI
VariableCoefficientStd. Errort-StatisticProb.
HMC−0.04460.0823−0.54280.5886
INS0.04880.01782.73200.0075
TEC−0.16270.0670−2.42700.0171
GDP0.02870.05880.48780.6268
C−0.10720.3099−0.34600.7301
Short Dynamics
Dependent Variable: ECI
VariableCoefficientStd. Errort-StatisticProb.
D(HMC)−0.71510.1428−5.00760.0000
D(INS)0.33750.05765.85880.0000
D(TEC)−1.16930.1918−6.09620.0000
D(GDP)2.18520.54094.03950.0001
CointEq(−1)−0.08030.0215−3.73830.0003
R-squared0.9944Adjusted R-squared.0.9935
F-statistic1103.937Prob (F-statistic)0.0000
Durbin–Watson stat2.0861
Diagnostics Checks
Breusch–Godfrey Serial Correlation LM Test:
F-statistic0.6859Prob.0.5062
Obs R-squared1.6188Prob. Chi-Square.0.4451
Heteroskedasticity Test: ARCH
F-statistic0.0434Prob.0.8352
Obs R-squared0.0443Prob. Chi-Square.0.8333
Stability TestStable
Normality TestNot normally distributed
Source: authors’ computation.
Table 5. Toda–Yamamoto causality result.
Table 5. Toda–Yamamoto causality result.
Null HypothesisMWALD Statisticsp-ValueDecision
INS does not Granger-cause GDP2.46160.2921Accept
GDP does not Granger-cause INS5.13680.0767Reject
HMC does not Granger-cause GDP0.75820.6845Accept
GDP does not Granger-cause HMC0.80680.668Accept
ECI does not Granger-cause GDP1.21690.5442Accept
GDP does not Granger-cause ECI2.01630.3649Accept
TEC does not Granger-cause GDP0.30060.8604Accept
GDP does not Granger-cause TEC1.93680.3797Accept
ECI does not Granger-cause HMC0.45430.7968Accept
HMC does not Granger-cause ECI0.45580.7962Accept
ECI does not Granger-cause INS0.44370.801Accept
INS does not Granger-cause ECI3.70690.1567Accept
ECI does not Granger-cause TEC1.23830.5384Accept
TEC does not Granger-cause ECI0.31810.8529Accept
HMC does not Granger-cause INS4.6890.0959Reject
INS does not Granger-cause HMC1.89810.3871Accept
HMC does not Granger-cause TEC1.15560.5611Accept
TEC does not Granger-cause HMC0.55480.7577Accept
INS does not Granger-cause TEC1.1120.5735Accept
TEC does not Granger-cause INS0.84360.6559Accept
Source: authors’ computation.
Table 6. Variance decomposition from the SVAR.
Table 6. Variance decomposition from the SVAR.
Variance Decomposition of INS
PeriodS.E.INSHMCGDPTECECI
20.04734999.445980.5246990.0001030.0262760.002944
40.07593295.099754.7738170.0582560.0220320.046144
60.09666688.5573410.693330.3567350.260390.132204
80.11338583.1363314.717240.8360351.0661540.244239
100.12723579.4927616.604341.3537912.1488110.400298
Variance Decomposition of HMC
PeriodS.E.INSHMCGDPTECECI
20.0169610.3897589.095680.1405260.2854270.08862
40.0228417.27757189.433720.4961832.1699470.62258
60.0239967.08798986.464660.5361694.7780471.133137
80.0243958.0509183.722020.610146.3910021.225926
100.0246168.60970782.266290.8711577.0423291.210518
Variance Decomposition of TEC
PeriodS.E.INSHMCGDPTECECI
20.0141790.371590.9233941.4125897.102580.189859
40.0249830.2136631.9028392.64829394.345560.88964
60.032170.1366883.637633.63224891.136971.456467
80.0366370.2546515.6404334.37315988.028641.703114
100.0393810.7437167.364214.92855585.253361.71016
Source: authors’ computation.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ngqoleka, S.; Ncanywa, T.; Mpongwana, Z.; Asaleye, A.J. Industrial Diversification in Emerging Economies: The Role of Human Capital, Technological Investment, and Institutional Quality in Promoting Economic Complexity. Sustainability 2025, 17, 7021. https://doi.org/10.3390/su17157021

AMA Style

Ngqoleka S, Ncanywa T, Mpongwana Z, Asaleye AJ. Industrial Diversification in Emerging Economies: The Role of Human Capital, Technological Investment, and Institutional Quality in Promoting Economic Complexity. Sustainability. 2025; 17(15):7021. https://doi.org/10.3390/su17157021

Chicago/Turabian Style

Ngqoleka, Sinazo, Thobeka Ncanywa, Zibongiwe Mpongwana, and Abiola John Asaleye. 2025. "Industrial Diversification in Emerging Economies: The Role of Human Capital, Technological Investment, and Institutional Quality in Promoting Economic Complexity" Sustainability 17, no. 15: 7021. https://doi.org/10.3390/su17157021

APA Style

Ngqoleka, S., Ncanywa, T., Mpongwana, Z., & Asaleye, A. J. (2025). Industrial Diversification in Emerging Economies: The Role of Human Capital, Technological Investment, and Institutional Quality in Promoting Economic Complexity. Sustainability, 17(15), 7021. https://doi.org/10.3390/su17157021

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop