Exploring the Dynamics of Human Capital and Economic Growth in Central Asian Countries: An Empirical Analysis Using Panel Data

Abstract

This study examines the relationship between human capital and economic growth in five Central Asian countries: Uzbekistan, Kazakhstan, Turkmenistan, Kyrgyzstan and Tajikistan. We have selected the period from 2000 to 2023 for our analysis. This study evaluates the impact of various components of human capital—namely education, health expenditure, life expectancy, child mortality and labour productivity—on gross domestic product growth, employing panel data regression models based on data from the World Bank’s World Development Indicators. The empirical examination employed pooled ordinary least squares, fixed-effects (FE) and random-effects models, with the final model selected via the Hausman test, which indicated that the FE model was the most appropriate fit. We conducted several diagnostic tests to assess the model’s reliability; the modified Wald test indicated heteroscedasticity, whereas the Wooldridge test confirmed the absence of autocorrelation in the panel data. Furthermore, the absence of multicollinearity among the variables was validated through the variance inflation factor test. Based on these findings, we recommend that regional policymakers prioritize human capital development, allocate resources effectively and formulate economic policies grounded in institutional reforms.

Keywords

Human capital economic growth Central Asia panel data education expenditure health

1. Introduction

Human capital encompasses the knowledge, skills, competencies and overall well-being of individuals and serves as a fundamental cornerstone of modern economic growth (Cappelli et al., 2023). In today’s rapidly evolving landscape, both human capital and technological advancement are essential forces driving economic development, as they synergistically enhance one another’s effectiveness. Notably, investments in these areas not only boost productivity and operational efficiency but also promote sustainable development. This dynamic interplay underscores the critical importance of skilled human resources, which are as vital as physical assets and technological innovations in driving economic progress (Sulisnaningrum et al., 2022). According to the principles of endogenous growth theory, human capital is a pivotal driver of economic growth. It is often argued that innovation improves production efficiency, underscoring the importance of knowledge, expertise and experiential learning for achieving long-term sustainable development (Aydin and Bozatli, 2023). Ideally, human capital transcends the traditional concept of a mere production factor; it is recognized as a strategic asset that continually evolves and renews itself. In doing so, it contributes to the structural enhancement of the economic system, particularly in a knowledge-based economy (Altiner & Toktaş, 2017).

2. Literature Review

Adam Smith, a prominent figure in economic thought, underscored the critical importance of knowledge, skills and experience in enhancing production efficiency. He argued that a skilled workforce is a fundamental driver of economic growth, suggesting that the productivity gains from such a workforce significantly contribute to a nation’s prosperity (Smith, 1776). Building on this idea, economist Theodore Shultz placed strong emphasis on human capital. He viewed investments in education, healthcare and migration as essential components for fostering economic development. Schultz sought to demonstrate that these investments yield significant returns, thereby enhancing long-term economic benefits (Schultz, 1961). In a similar vein, Gary Becker introduced a more quantitative approach by incorporating human capital into economic analysis. He proposed that human capital should be viewed as a financial investment, much like physical capital. Through precise mathematical models, Becker meticulously explored the intricate relationship between human capital and production, demonstrating how an educated and skilled workforce can drive economic success (Becker, 1964). Lucas emerged as a pioneering thinker in endogenous growth theory, positing that economic growth can occur spontaneously through the accumulation of human capital and the transfer of knowledge. Lucas contended that when society invests in education, it not only enriches individuals but also accelerates broader societal growth, leading to a more dynamic economy (Lucas, 1988). Lastly, Paul Romer developed a model that closely links technological progress to human capital. He highlighted the vital role that human capital plays in fostering innovation, asserting that strategic investments in education and scientific research are crucial for driving economic growth through endogenous factors. Romer’s insights suggest that a well-educated workforce is essential for fostering an environment conducive to innovation and technological advancement (Romer, 1990).

Several classical theoretical models provide a basis for analyzing the impact of human capital on economic growth, through which its role in enhancing production efficiency, driving technological progress and promoting sustainable economic growth is systematically examined. This study examines classical approaches, including the neoclassical growth model and endogenous growth theories, as well as their extensions, from a scientific and theoretical perspective.

The Solow model is a widely used neoclassical approach to analyzing economic growth that assumes economic growth is driven by three main factors: labour (L), physical capital (K) and technological progress (A). The model assumes that economic growth in the long run is driven primarily by technological progress, as the marginal productivities of labour and capital decline. The production function of the Solow model has the following general form:

Y (t) = F (K (t), A (t) L (t))

Here, Y(t) is the total output at time t, K(t) is physical capital, L(t) is labour, and A(t) is the technological level. The function is usually expressed in the Cobb–Douglas form:

Y = A \times K^{\propto} \times L^{1 - \propto}

In this model, technological progress is treated as exogenous, meaning it is determined outside the economic system rather than within it. Human capital is not treated as a separate, independent factor of production in the model, a limitation of the model. One of the main theoretical conclusions of the Solow model is that long-term economic growth can be achieved only through technological progress, since investments in physical capital lead to a state of finite profitability that declines over time (Gemmell, 1995). Therefore, changes in the capital-to-labour ratio do not permanently guarantee growth rates, and sustainable economic growth can be achieved only through continuous technological improvement.

Becker (1964) interprets this concept not as a simple factor of production but as a form of personal investment made by individuals. He considers human capital a form of wealth generated by factors such as education, healthcare, vocational training and skills development. According to Becker, such investments not only increase individual income but also enhance society’s overall economic well-being by increasing labour productivity. This approach is fundamentally different from the traditional Solow neoclassical model, which does not treat human capital as a separate factor and regards exogenous technological progress as the primary source of growth. On the other hand, Becker’s model emphasizes human capital as a central factor at both the microeconomic and social levels, enabling us to interpret economic growth as a more complex and multifaceted process (Xie & Yang, 2020).

Nonetheless, Lucas (1988) asserts that human capital serves as a fundamental and dynamic driver of economic growth. Unlike the Solow model, the Lucas model integrates human capital directly into the production function. In this framework, economic growth is fuelled by the time and resources that economic agents allocate to education, which encompasses the development of knowledge and skills. In essence, economic agents’ investments in knowledge enhance their productivity and contribute to overall economic growth. Lucas also highlights the social externalities associated with human capital, indicating that an individual’s level of education can positively influence the productivity of other economic agents. Thus, the Lucas model provides a more comprehensive analysis of the role of human capital in economic growth, considering both individual and societal perspectives. It further attributes technological progress primarily to the accumulation of human capital, aligning with the core tenet of endogenous growth theory: that growth originates from internal economic mechanisms.

The interconnection between human capital and economic growth has been extensively examined through empirical studies using panel data from many countries (Barro, 2001; Hanushek & Woessmann, 2008). Research in this area employs dynamic panel approaches, such as system generalized method of moments (GMM), to assess the impact of human capital on growth, while accounting for the quality of education, health status and demographic factors (Bloom et al., 2004). For example, empirical analyses based on panel data for the BRICS countries, Turkey, the MENA (Middle East and North Africa) region and some countries on the African continent have shown that human capital has a significant impact on economic growth (Bal et al., 2014; Bashir et al., 2024; Fahim & Rhanami, 2018). In these studies, human capital, typically measured by education level, health indicators or personnel qualifications, is positively and statistically significantly correlated with economic growth. These results suggest the importance of recognizing human capital as a crucial component of economic policy and strategic planning. At the same time, these studies emphasize the need to consider institutional and demographic factors that vary across contexts. In an empirical study using panel data for 99 countries, Zhu (2013) finds that the effect of human capital on economic growth varies significantly across countries’ income levels. The results show that in high-income countries, human capital has a strong and positive impact on economic growth; increases in education, health and other human capital indicators significantly raise economic productivity. However, this effect was much weaker or statistically insignificant in low- and middle-income countries. Zhu (2013) attributes this difference to variations in institutional quality, the effectiveness of the education system, and the extent to which human capital is integrated into practical economic activities. This means that human capital can serve as a significant driver of growth only if supported by effective systems, strong institutions and high-quality educational infrastructure (Namozov et al., 2024; Zhu, 2013). These results indicate that human capital alone is not enough; it must be effectively utilized within the economic system. Therefore, policymakers should implement measures to enhance human capital alongside institutional reforms and improvements in educational quality.

Sheidaei and Tash (2014) empirically examine the complex, non-linear relationship between human capital and technological innovation across 104 countries, including 79 developing countries. The results demonstrate that human capital is crucial in enhancing productivity and serves as a strategic mechanism for generating, adapting and adopting innovations. However, this effect is not linear: if human capital is too low, there is insufficient capacity to support innovation. Conversely, if human capital is too high, its effectiveness in stimulating economic growth and innovation declines, likely due to imbalances in labour-market matching or inconsistencies in the quality and direction of education. Sheidaei et al. support this finding by arguing that the positive impact of human capital on innovation capacity is optimal, implying that policymakers should develop human capital quantitatively, qualitatively and contextually (Sheidaei & Tash, 2014). Sharma (2016) investigates the relationship between economic growth and financial development. This study analyzed data from 66 countries to explore the interplay between human capital and the development of the financial sector. The findings reveal that the influence of human capital on economic growth is closely linked to the maturity of the financial system. In nations with well-developed financial frameworks, human capital serves as a significant catalyst for economic growth, as investments in education, healthcare and innovation are effectively facilitated by efficient financial mechanisms. Conversely, in countries with underdeveloped financial infrastructure, the positive effect of human capital on economic growth was statistically insignificant. Sharma (2016) elucidates this situation by highlighting inadequate mobilization of financial resources and the absence of mechanisms to promote the long-term benefits of human capital investments. The author underscores that the economic efficiency of human capital should be assessed both independently and in relation to the institutions and financial systems that underpin it. This perspective reaffirms the necessity for a multifaceted approach in economic policymaking. There is an interesting result reported by Nasreen et al. (2015). She examined the impact of factors such as institutional quality, human capital and economic freedom on economic growth in early developing countries, using complex panel and cross-country analyses. The study results demonstrate a strong, stable and statistically significant positive relationship between human capital and economic growth, particularly in countries that invest actively in the education and health sectors. The authors argue that human capital does not drive growth independently; it reaches its full economic potential only when accompanied by good institutions and economic freedom. Using fixed-effects (FE) and random-effects (RE) models, it is possible to analyze how human capital changes over time, rather than across countries, within a country, particularly in the MENA region. This approach isolates and accounts for fixed and random factors (Ada, 2014).

In the econometric model, the system GMM method has been used to address endogeneity. For instance, in research focused on the D-8 countries, this approach has been instrumental in identifying the lagged impacts of human capital. The system GMM model has proven particularly adept at capturing high-level dynamics, especially within small panel samples (Yousfi & Arfa, 2023). Conversely, the dynamic panel modelling approach (common long-run, varying short-run) has been predominantly applied to OECD countries, indicating that the long-term effects of human capital remain stable across these nations, whereas short-term effects vary. This methodology is recommended as a vital tool for assessing long-term economic performance (Arnold et al., 2007).

In recent years, research on the empirical relationship between human capital and economic growth has expanded worldwide. Analyses utilizing panel data indicate that key components of human capital—such as educational quality, health status and demographic factors—significantly contribute to long-term economic growth (Kim & Park, 2021; Oketch & Oduor, 2022; Zhang & Ahmed, 2020). Furthermore, methodologies like system GMM, FE and panel autoregressive distributed lag (ARDL) have revealed both lagged and dynamic effects of human capital (Lee & Nguyen, 2019; Torres & Delgado, 2023). Notably, the studies mentioned above also report bidirectional relationships with growth (Granger causality), and factors such as political stability and institutional quality are important mediators of these effects. Qualitative and strategic investments in human capital typically enhance labour productivity and lay a solid foundation for technological innovation, innovative capacity and sustainable economic growth. These methodological approaches have enabled the exploration of the complex relationship between human capital and economic growth from various perspectives. Therefore, the methodology chosen for empirical studies should be carefully considered, taking into account the sample, data quality and research goals.

While these methodologies have significantly enhanced our understanding of the role of human capital in economic growth, their application in Central Asia—specifically in Uzbekistan, Kazakhstan, Turkmenistan, Kyrgyzstan and Tajikistan—remains inadequately explored. These nations differ notably from developed countries in their socio-economic, demographic and institutional characteristics. Consequently, it is essential to adapt existing methodological approaches to align with the regional context, particularly by developing econometric models that incorporate factors unique to Central Asia. In Central Asian countries, there is a notable lack of empirical research exploring the link between human capital and economic growth (e.g., Yormirzoev, 2023). This underscores a considerable gap in the scientific literature on this topic. Therefore, developing new analytical frameworks and models specifically designed for this region is essential to enhancing our understanding of the economic implications of human capital, thereby providing a solid scientific foundation for informed policymaking.

This article addresses the identified research gaps by examining the influence of human capital on economic growth in the Central Asian nations of Uzbekistan, Kazakhstan, Turkmenistan, Kyrgyzstan and Tajikistan from 2000 to 2023. Utilizing pooled ordinary least squares (pooled OLS), FE and RE methods, we analyze the relationship between economic growth and various components of human capital, including education and health. This methodology facilitates a comparative analysis across the three models, enabling us to pinpoint the most effective approach. Additionally, the examination of results over the 2000–2023 period reveals changes that have been largely overlooked in Central Asia, thereby providing a robust foundation for informed policymaking.

In our study, the impact of human capital on economic growth rates is examined through a comparative analysis of recent empirical studies covering a wide range of countries. At the same time, the differences observed in the role of human capital across regional and institutional settings, as well as the mechanisms that influence growth (e.g., labour productivity, innovative capacity, social stability), are analyzed with particular attention. Based on these analyses, evidence is presented confirming the central role of human capital in macroeconomic growth, and differences between developing and developed countries are identified.

3. Materials and Methods

3.1. Data and Variables

This study examines the key factors that influence economic growth in the five Central Asian countries: Uzbekistan, Kazakhstan, Turkmenistan, Kyrgyzstan and Tajikistan. It utilizes annual panel data spanning from 2000 to 2023, a period following their independence from the Soviet Union in 1992. Specifically, the analysis focuses on the years following these nations’ structural adjustments after 8 years of independence. The rationale for selecting these countries stems from their shared characteristics in socio-economic development stages, demographic profiles, cultural and historical contexts, and experiences during the post-Soviet transition. Moreover, their recent strategies targeting sustainable growth, government intervention levels and regional integration efforts exhibit notable consistency. We have sourced from the World Bank’s World Development Indicators (WDI) database.¹ This source is considered a reliable source on macroeconomic, social and institutional variables for both developing and developed countries. The main variables selected for this research are outlined in Table 1 and encompass essential macroeconomic and human capital indicators that directly or indirectly influence economic growth. The selection process is thoughtfully guided by well-established theoretical foundations, as discussed in Section 2, which reviews a range of prior empirical studies. This dual approach ensures a comprehensive and nuanced analysis of how these elements contribute to driving economic growth. By considering both theoretical insights and reliable data, the process aims to illuminate the complex interactions that fuel economic development.

Table 1.

Description of the Variables.

Variables	Name	Symbol	Unit of Measure	Description
The dependent variable	Real GDP growth rate (%)	GDP_Growth	Per cent (%)	Represents the level of growth or decline in a country’s economy. It determines the country’s economic activity, population well-being and potential for future development.
The independent variables	Education expenditure (% of GDP)	Edu_Exp	Per cent of GDP (%)	The share of state or government funds allocated to the education system relative to total GDP indicates the development of human capital.
	Health expenditure (% of GDP)	Health_Exp	Per cent of GDP (%)	The amount of funds allocated to the health system in relation to GDP. This indicator affects the health and work capacity of the population.
	Life expectancy at birth	Life_Expect	Years	The average life expectancy of a newborn baby. This indicator reflects the level of health, living conditions and medical care.
	Infant mortality rate	Inf_Mort	Per 1,000 live births	How many children out of 1,000 live births die before reaching the age of 1. This depends on the quality of medical care and the effectiveness of the health system.
	Labour productivity	Lab_Prod	USD per person	The volume of real GDP per worker. This indicates the efficiency of the labour force and production capacity.
Control variables	Foreign direct investment, net inflows	FDI_Share	Per cent of GDP (%)	Net foreign direct investment (FDI) to GDP. Used to measure the impact of foreign capital on economic growth.
	Inflation rate	Inflation	Per cent (%) – annual change in consumer price index (CPI)	Annual change in the general price level. Inflation reflects the impact of economic stability and monetary policy.
	Trade openness	Trade_Open	Per cent of GDP (%)	Trade openness: the share of foreign trade (exports + imports) to GDP. This indicates the degree of economic integration into the global market.
	Population growth	Pop_Growth	Per cent (%) – annual population increase	Annual population growth rate. Indicates the rate of expansion of the labour force base.

This study selected the real gross domestic product (GDP) growth rate (GDP_Growth) as the primary outcome variable. This indicator reflects the extent to which a country’s economy has grown over time, expressing the dynamics of economic growth in percentage terms. This variable is widely used in international experience as a criterion for determining the level of economic development.

Five leading indicators of human capital were selected as independent variables. They are:

Public spending on education (% of GDP),

Healthcare spending (% of GDP),

Life expectancy (in years),

Under-5 mortality rate (per 1,000 live births),

Labour productivity (GDP per worker).

These variables are widely used in modern economic research as primary measures of the impact of human capital on economic growth. They show the contribution of human capital to economic activity through health, education and productive capacity.

We have chosen four additional control variables to ensure accuracy. These control variables can directly or indirectly affect economic growth through their effects on macroeconomic stability, external economic relations and demographic factors.

Inflation rate per annum (in per cent),

Degree of openness (ratio of exports and imports to GDP),

Population growth rate per annum (in per cent),

Net foreign direct investment (FDI as a % GDP).

The statistical indicators used in this study—namely the mean, standard deviation, minimum (min) and maximum (max)—are presented in Table 2. These statistics were calculated using Stata 17 software, based on a panel database that includes economic and social indicators collected for five Central Asian countries (Uzbekistan, Kazakhstan, Turkmenistan, Kyrgyzstan and Tajikistan) between 2000 and 2023.

This descriptive analysis stage enables us to gain an initial understanding of the overall distribution, variability and spread of the variables before undertaking empirical modelling. It enables us to identify regional differences for each indicator, outliers and potential imbalances such as heterogeneity. Additionally, presenting statistical indicators in this manner is a crucial first step in identifying informal relationships or structural differences among the variables in the model.

Table 2.

Summary Statistics of the Variables.

Variable	Obs.	Mean	Std Dev	Min	Max
GDP Growth	120	6.407	3.329	–7.1	14.7
Edu Exp	120	4.386	1.345	2.3	7.4
Health Exp	120	5.136	1.491	2.6	8.9
Life Expect	120	68.996	2.413	63.5	74.4
Inf Mort	120	28.62	12.817	7.6	62.2
Lab Prod	120	22,657.45	16,210.572	6,109	62,197
FDI Share	120	4.716	4.224	–4.9	22.5
Inflation	120	9.432	6.836	0.1	60.6
Trade Open	120	81.491	31.381	29.2	175.4
Pop Growth	120	1.699	0.5	0.3	2.6

According to the statistical indicators in Table 2, the real GDP growth rate, which is included as an outcome variable, averages 6.41%, indicating that the region’s countries have an average high economic growth rate. However, the range from –7.1% to 14.7% indicates a sharp fluctuation of this variable. Global crises, political instability or the degree of dependence on natural resources can explain these differences.

Among the human capital variables, education expenditure (relative to GDP) averages 4.39%, and healthcare expenditure is 5.14%, indicating that these countries invest moderately in human capital. Life expectancy is approximately 69 years, and this indicator exhibits a relatively stable, right-skewed distribution. The infant mortality rate averages 28.62, with some countries reporting very high rates (62.2) and others reporting low rates (7.6). This may be due to the quality of the health system and the development of its social infrastructure. Labour productivity (GDP per worker) averages $22,657, but the substantial range between the minimum and maximum values ($6,109–$62,197) indicates significant disparities in the region’s productivity levels.

Among the control variables, the share of FDI averages 4.72%, with regional variability (standard deviation 4.22), indicating that investment was very high in some years. The inflation rate averaged 9.43%, reaching 60.6% in some years, which may reflect years of economic instability. Foreign trade openness is 81.49%, with significant variation in trade volume (ranging from 29.2% to 175.4%). The population growth rate averaged 1.70%, indicating relatively stable demographics.

This descriptive analysis shows significant differences in human capital, economic growth and macroeconomic stability among Central Asian countries. This justifies the use of panel modelling methods (e.g., FE) that account for individual country characteristics in empirical analysis.

3.2. Empirical Model

In this study, a panel regression model was used to estimate the impact of human capital on economic growth. Panel data are constructed for each country (i) and year (t), allowing for the analysis of variable changes across space and time. The panel approach has the advantage of allowing temporal variation and cross-country differences to be accounted for.

The central empirical model equation of the study is expressed as follows:

{G D P_G r o w t h}_{i t} = α + β_{1} {E d u_E x p}_{i t} + β_{2} {H e a l t h_E x p}_{i t} + β_{3} {\times L i f e_E x p e c t}_{i t} + β_{4} \times {I n f_M o r t}_{i t} + β_{5} {\times L a b_P r o d}_{i t} + {γ C o n t r o l_V a r i a b l e s}_{i t} + μ_{i} + λ_{t} + ε_{i t}

Here:

${G D P_G r o w t h}_{i t}$ : GDP growth rate in year t for country i;

${E d u_E x p}_{i t}, {H e a l t h_E x p}_{i t}, {L i f e_E x p e c t}_{i t}, {I n f_M o r t}_{i t}, {L a b_P r o d}_{i t}$ : independent variables related to human capital;

${C o n t r o l_V a r i a b l e s}_{i t}$ : set of control variables (FDI_Share, Inflation, Trade_Open, Pop_Growth);

μ_i: country-specific fixed effects;

$λ_{t}$ : time effects;

$ε_{i t}$ : random errors.

The analysis was conducted using the following step-by-step methodology. First, a pooled OLS regression was performed, followed by FE and RE models. The Hausman test was used to select between the models. According to the test results (χ²(8) = 35.07; p < .05), the FE model was preferable, as it allows for the consideration of individual constant factors.

A crucial step in regression models with panel data is to assess the linear relationship among the independent variables. Specifically, including highly correlated variables simultaneously can lead to inaccurate estimates of regression coefficients, larger standard errors and reduced model reliability. A Pearson correlation matrix was constructed to assess multicollinearity (Table 3).

Table 3.

Correlation Matrix of the Independent Variables.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
(1) Edu_Exp	1.000
(2) Health_Exp	0.630	1.000
(3) Life_Expect	0.565	0.159	1.000
(4) Inf_Mort	–0.436	0.006	–0.812	1.000
(5) Lab_Prod	–0.391	–0.484	0.312	–0.454	1.000
(6) FDI_Share	–0.307	–0.233	–0.297	0.045	0.129	1.000
(7) Inflation	–0.166	–0.066	–0.257	0.298	–0.133	–0.112	1.000
(8) Trade_Open	0.019	0.064	–0.345	0.246	–0.378	0.365	0.292	1.000
(9) Pop_Growth	0.186	0.392	0.361	–0.053	–0.057	–0.089	–0.177	–0.139	1.000

In this study, a strong inverse relationship was observed between the variables Life_Expect (life expectancy) and Inf_Mort (infant mortality rate) (Pearson correlation coefficient r = –.81). This means that including both variables in the model simultaneously may introduce multicollinearity. Therefore, including only one of these two variables in the model was necessary. While Inf_Mort indicates the negative aspects of the health system’s efficiency, such as existing problems in healthcare, Life_Expect is considered the overall measure encompassing healthcare, living conditions, nutrition, environmental factors and other socio-economic factors. Additionally, the Life_Expect indicator is more stable, exhibits lower variance (Table 2) and reflects long-term changes, making it more suitable for macroeconomic analysis. Therefore, in this analysis, the appropriate methodological approach was to retain the Life_Expect variable and exclude the Inf_Mort variable from the model. This approach reduces the risk of multicollinearity and increases the statistical reliability of the model estimates.

A Modified Wald test based on the FE model was performed to assess the presence of heteroscedasticity. The test statistic (χ²(5) = 58.33; p < .001) indicated heteroscedasticity across groups in the panel. Therefore, standard errors robust to heteroscedasticity were used in the model estimation. A Wooldridge test was performed to determine serial autocorrelation. The test did not reject the null hypothesis—the absence of first-order autocorrelation (F(1,4) = 1.513; p = .2861). This means that there is no autocorrelation in the panel data.

Based on the results presented, the final model was estimated using robust standard errors (FE robust) within an FE framework. This methodology enhances the accuracy and reliability of the coefficients, thereby ensuring the model’s statistical soundness. Furthermore, to address multicollinearity, only Life_Expect was retained in the final estimation, whereas Inf_Mort was excluded. Consequently, the final model can be expressed as follows:

{G D P_G r o w t h}_{i t} = α + β_{1} {E d u_E x p}_{i t} + β_{2} {H e a l t h_E x p}_{i t} + β_{3} {\times L i f e_E x p e c t}_{i t} + β_{4} \times {L a b_P r o d}_{i t} + {γ C o n t r o l_V a r i a b l e s}_{i t} + μ_{i} + λ_{t} + ε_{i t}

The chosen methodology enables a thorough analysis of how human capital impacts economic growth over space and time. The final model, which meets the criteria of economic logic, empirical evidence and statistical reliability, provides a solid foundation for interpreting the results presented in the next section.

4. Results and Discussions

4.1. Analyses of the Results

During the analysis, we assessed the impact of human capital on GDP growth rates using a final regression model estimated via the FE approach. This analysis covered 24 years of data across five countries (n = 120). The regression model employed robust standard errors, which accounted for heteroscedasticity and intra-cluster correlation typically found in panel data. The results of the final regression are presented in Table 4.

Table 4.

Results of Fixed-Effects Regression Model (Robust Standard Errors).

GDP_Growth	Coef.	Std Error	t-Value	p Value	[95% Conf. Interval]		Sig.
Edu_Exp	1.012	0.339	2.99	.040	0.072	1.952	**
Health_Exp	–0.430	0.160	–2.68	.055	–0.875	0.015	**
Life_Expect	0.042	0.241	0.17	.870	–0.627	0.711
Lab_Prod	0.000	0.000	–0.59	.586	0.000	0.000
FDI_Share	0.065	0.090	0.72	.510	–0.186	0.316
Inflation	–0.033	0.058	–0.56	.607	–0.195	0.130
Trade_Open	0.042	0.015	2.79	.049	0.000	0.083	**
Pop_Growth	–2.701	1.432	–1.89	.132	–6.676	1.275	***
Constant	2.876	16.771	0.17	.872	–43.688	49.44
Mean dependent var.		6.407		SD dependent var.		3.329
R ²		0.228		Number of obs.		120
F-test		0.000		Prob. > F		0.000
Akaike crit. (AIC)		586.882		Bayesian crit. (BIC)		598.032

Note: ** Signifies a 5% level significance; *** signifies a 10% level significance.

The model’s explanatory power (R²) is 0.228, indicating that it explains approximately 22.8% of the variance in GDP growth rates. The effect of education expenditure (Edu_Exp) on economic growth is statistically significant (p = .04) and positive (coefficient = 1.012). This result suggests that investment in education contributes to sustainable economic growth in the country. This finding aligns with previously studied scientific literature (Banik & Khatun, 2012; Kharel et al., 2025; Liao et al., 2019; Meilisa et al., 2024). The study also reveals that health expenditure (Health_Exp) has a negative impact on economic growth, with a coefficient of –0.43 in the regression model. This result is statistically significant at the 10% level (p = .055). Several factors may explain this phenomenon. First, while healthcare spending often yields long-term social and economic benefits, it may not have an immediate effect on short-term economic growth rates. Second, the negative impact may be attributed to resource misallocation or inefficient use, such as diverting budget funds to administrative costs rather than to health infrastructure, or to systemic issues such as corruption and poor governance. Additionally, increases in healthcare spending are sometimes linked to economic crises, such as pandemics, natural disasters or political instability, indicating that these expenditures do not stimulate growth on their own but rather respond to existing challenges. Consequently, the negative and significant impact of this variable may result from poor performance in the health sector or various contextual factors. The analysis reveals that the variables life expectancy at birth (Life_Expect) and labour productivity (Lab_Prod) are not statistically significant in the model (p > .5). However, Life_Expect was included instead of infant mortality (Inf_Mort) to reduce multicollinearity and maintain a meaningful explanatory role in the model. Trade openness (Trade_Open) is a positive and statistically significant growth factor (coefficient = 0.042; p = .049), indicating that economic growth is higher in countries with high levels of trade integration. This finding supports previous studies (Adeboje et al., 2022; Mutunga, 2023). The results also indicate that some variables had no statistically significant effect on economic growth. Specifically, the p values for indicators such as labour productivity (Lab_Prod), life expectancy (Life_Expect), inflation rate (Inflation) and foreign direct investment share (FDI_Share) exceed 0.1, indicating that their effects are not statistically significant in the model. This may suggest that these variables have an indirect or long-term effect on economic growth. For instance, life expectancy (Life_Expect) and labour productivity (Lab_Prod) mainly influence economic efficiency through the quality of human capital and technological progress. However, these effects typically manifest over time through factors like the infrastructural and institutional environment, which may not be evident in short-term panel analyses. Additionally, inflation and FDI shares are subject to significant variations in their impact across countries due to differing political and economic contexts. Macroeconomic policies, institutional stability and the quality of the investment environment in various countries may weaken or complicate the impact of these variables in the model. Furthermore, FDI flows may respond more to geopolitical factors, resource-based investments or global capital flows than to economic growth. Therefore, it might be more appropriate to regard these variables as indirect determinants of economic growth or to analyze them within long-term structural models. Figure 1 presents the results for these factors, including their confidence intervals, and shows each factor’s coefficient and confidence interval. This graphical representation helps readers understand the direction of influence and the statistical reliability of each variable in the model.

Based on the above empirical results, it can be observed that in Central Asian countries, increasing investment in education and adopting an open economic policy focused on foreign trade are key strategic factors for stimulating economic growth. The positive and statistically significant impact of education spending on improving human capital directly contributes to economic efficiency and sustainable growth. Similarly, the positive effect of trade openness on growth underscores the need to deepen the region’s integration into external markets and enhance its competitiveness. On the other hand, the low statistical significance of variables such as healthcare spending and population growth rate in the model suggests that their effects should be interpreted with caution. Specifically, the negative impact of healthcare spending necessitates further detailed research on the effectiveness of these funds and their influence on short-term economic outcomes. It is also possible that population growth rates have an indirect, rather than direct, effect on growth through demographic pressures, the labour market and productive resources. These results highlight essential policy directions: education and foreign trade reforms should be prioritized, while health and demographic factors should be assessed through long-term and systematic approaches.

4.2. Additional Diagnostic Tests

Several diagnostic tests were performed to evaluate the statistical reliability and accuracy of the panel data regression model. These tests help identify and address key technical issues that may affect the model, such as heteroscedasticity, autocorrelation and multicollinearity. Specifically, the modified Wald test indicated the presence of heteroscedasticity in the panel model (χ² = 85.29; p < .000) (see Table 5). This result suggests that the model’s variance is unstable (i.e., variable) across panel groups (countries), which can reduce the reliability of estimation using simple standard errors. Therefore, robust standard errors were used in the model estimation. This method addresses heteroscedasticity, allowing for a more accurate assessment of the statistical significance of the coefficients. Using robust standard errors is especially important when each country in the panel data has its own internal variance. As a result, a robust version of the FE model was employed in the final estimation, allowing for a more straightforward interpretation of the coefficients and their p values by accounting for individual effects and mitigating the impact of heteroscedasticity.

Table 5.

Diagnostic Test Results for Model Validity.

Test Name	Null Hypothesis (H₀)	Test Statistic	p Value	Conclusion
Modified Wald test	Groupwise homoscedasticity: σ²ᵢ = σ²	χ²(5) = 85.29	.0000	Reject H₀: Heteroscedasticity is present in the model.
Wooldridge test	No first-order autocorrelation: ρ = 0	F(1,4) = 1.509	.2866	Fail to reject H₀: No autocorrelation detected in the model.

We have assessed a time-dependent (autocorrelation) problem in the model using the Wooldridge test. We find that there was no first-order autocorrelation, since the statistical value F = 1.509 and the p value was .2866 (i.e., p > .1) (see Table 5). This indicates the absence of time-dependent problems in the panel data and confirms the robustness of the results.

Additionally, the variance inflation factor (VIF) test was conducted to assess multicollinearity in the model. Among the VIF values calculated, the highest was 4.94 for the Edu_Exp variable, and the average of all values was 2.61. These results fall within the generally accepted critical limits (VIF < 5), indicating no significant linear relationships among the independent variables (see Table 6). Therefore, based on diagnostic tests, it was determined that although the model exhibits heteroscedasticity, there is no evidence of autocorrelation or multicollinearity. Thus, the final FE model, estimated with robust standard errors, is statistically reliable and interpretable.

Table 6.

Multicollinearity Diagnostics Using Variance Inflation Factor (VIF) Statistics.

Variables	VIF	1/VIF
Edu Exp	4.944	0.202
Life Expect	4.315	0.232
Lab Prod	2.835	0.353
Health Exp	2.514	0.398
Pop Growth	1.707	0.586
Trade Open	1.703	0.587
FDI Share	1.569	0.637
Inflation	1.290	0.775
Mean VIF	2.610

In general, the diagnostic test results confirm the reliability and statistical robustness of the final panel regression model. Although the modified Wald test detected the presence of heteroscedasticity in the model, this problem was eliminated by using heteroscedasticity-resistant (robust) standard errors. This approach ensured the accuracy and statistical reliability of the estimated coefficients. Additionally, the Wooldridge test indicated that the model does not exhibit first-order time autocorrelation; the residuals are not serially correlated. This result indicates that the model built on panel data does not have time-dependent errors. The results of the VIF test, which assesses multicollinearity, indicated no significant linear relationships among the independent variables. The VIF values for all variables are below 5, indicating the model’s internal consistency and stability across parameters.

Based on these diagnostic tests, the developed panel model is statistically stable, consistent and interpretable. Therefore, the model results provide a solid empirical basis for analyzing the relationship between human capital indicators and economic growth in Central Asian countries, thereby enabling scientific conclusions.

4.3. Interpretations

The regression results indicate that educational spending has a significant positive effect on economic growth rates. These empirical findings support the model hypotheses by Becker (1964) and Lucas (1988), classic representatives of human capital theory. According to their theories, investments in human capital—particularly education—not only increase labour productivity but also enhance the economy’s overall production potential, leading to long-term, sustainable growth.

These results are significant in the context of developing countries. In particular, the impact of education spending on economic growth is increasingly salient in Central Asian countries, as the quality of education, infrastructure and pedagogical approaches is diversifying. This suggests that education reforms have a direct impact not only on social indicators but also on macroeconomic indicators. The study’s results are consistent with those of previous empirical studies (Arabov et al., 2023; Sheidaei & Tash, 2014; Zhu, 2013), which have examined the role of education in strategic development. These parallel conclusions, on the one hand, strengthen the credibility of the research results and, on the other hand, emphasize the need for policymakers in the region to prioritize education in economic policy.

4.4. Unexpected Negative Trend in Health Spending

The study results showed that health spending has a negative, albeit statistically weak (i.e., low confidence level), impact on economic growth. Several factors can explain this finding. First, the impact of healthcare investments is typically long-term and may not directly affect short- or medium-term economic indicators. Therefore, the relationship between health spending and economic growth may have been weak during the period analyzed.

Second, institutional problems in the health sector, including inefficient budget allocation and systemic inefficiencies, may have harmed economic growth. In such cases, health spending may not yield the expected benefits for human capital development, and even misallocation of resources may adversely affect macroeconomic balance. This conclusion is consistent with the ideas put forward by Nasreen et al. (2015).²

4.5. The Stimulating Role of Trade Openness/Economic Openness

The study results divulge that foreign trade openness has a positive and statistically significant impact on economic growth. This finding is consistent with several previous empirical studies, particularly by Mutunga (2023) and Adeboje et al. (2022). They have confirmed that global economic integration has a positive impact on economic activity, competitiveness and overall growth rates in developing countries. Especially for small and medium-sized open economies, the expansion of foreign trade relations is emerging as an important driver of economic growth. In particular, by developing products that meet international production standards and integrating into global supply chains, the region’s countries can significantly increase their economic potential. It is also possible that strategic approaches based on a country’s openness may be most effective only when implemented alongside institutional reforms and infrastructure modernization.

4.6. Statistically Insignificant Factors: Deeper Interpretations

The study’s findings showed that life expectancy (Life_Expect), labour productivity, the inflation rate and FDI did not have a statistically significant effect on economic growth. These results suggest that these factors influence economic growth indirectly, through lagged effects or depending on specific conditions. For instance, although the Life_Expect indicator reflects the overall health and longevity of the population, it does not directly affect production levels or growth rates. Although improvements in health have a positive effect on the long-term productivity of labour resources, this process takes time to be reflected in economic indicators. Therefore, the effect of this variable may not be statistically significant in the short- or medium-term models. Similarly, FDI is highly dependent on numerous internal and external factors, including global capital flows, investor confidence, the institutional environment and political stability. As noted by authors such as Sharma (2016) and Banik and Bhaumik (2006), the impact of FDI on economic growth can vary significantly across countries, depending on the extent to which these investments are integrated into the real sector. Also, in some cases, FDI is directed more towards countries with higher economic activity than towards those with higher economic growth; that is, it may appear as a result rather than a cause.

Similar explanations apply to labour productivity and inflation indicators. Labour productivity sometimes significantly affects economic growth only when it interacts with other factors (e.g., technological level, managerial quality, labour force skills). Inflation, in turn, is a complex macroeconomic indicator that can positively or negatively affect growth under different economic conditions, and its effects are not always unidirectional.

When analyzing these factors, it is necessary to consider their complex and interconnected nature. Future research can yield more precise results by accounting for the lagged effects of these factors, as well as mediating variables and structural equation modelling.

5. Conclusion

This study examined the influence of human capital on economic growth in Central Asian countries, using a panel dataset spanning 2000–2023. The analysis employs three different econometric models: pooled OLS, FE and RE. The FE model was determined to be the most suitable for the final analysis based on the results of the Hausman test. The empirical findings lead to several key insights:

Impact of educational expenditures: The study reveals that investments in education significantly and positively influence economic growth. This finding underscores the importance of enhancing human capital through practical, relevant educational systems. It is observed that investing in education not only promotes individual development but also fosters overall economic stability and boosts production efficiency across the region.

Health expenditures and their effects: In stark contrast, expenditures on health reveal a negative impact on economic growth, albeit with marginal statistical significance. This outcome underscores a pressing need to enhance the efficiency of resource allocation within the healthcare sector. It suggests that without comprehensive reforms and a more effective management structure, financial investments in healthcare may not yield the intended economic benefits.

Role of economic openness: Economic openness has emerged as a vital catalyst for economic growth within the region. It suggests that Central Asian countries may actively pursue strategies that enhance their integration into the global economy.

Analysis of additional indicators: Other variables evaluated—such as life expectancy, labour productivity, FDI and inflation—did not demonstrate a direct, statistically significant connection to economic growth. This finding implies that, although these factors are relevant, their influence on growth may be indirect or contingent on other situational variables. Therefore, further exploration is warranted to understand their effects fully. In conclusion, the findings of this study advocate for a strategic approach to human capital policy formulation in Central Asian countries—one that is informed by qualitative factors and tailored to local contexts.³ Prioritizing increased investment in education and implementing modernization reforms could serve as the principal drivers of economic growth in the region. Furthermore, it is essential to craft policies that ensure institutional stability in the health sector, thereby enhancing operational efficiency, promoting the rational use of financial resources and ultimately harnessing the long-term economic potential of human capital.

Footnotes

Acknowledgement

The authors are grateful to the journal’s anonymous referees for their extremely useful suggestions to improve the quality of this paper. The authors would also like to thank Professor Arindam Banik for his helpful comments on an earlier draft of the paper prior to submission to the journal.

Authors’ Contribution

The first author conceptualized the research topic, while the second and third authors reviewed the literature and identified gaps in the literature. The fourth author was responsible for data collection and curation, as well as developing the econometric model in collaboration with the first and fifth authors. Ultimately, the estimation and preparation of the manuscript’s final draft were undertaken by the first and sixth authors.

Declaration of Conflicting Interests

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

Ethical Declaration

The authors abide by all the ethics involved in this academic work and have not submitted it to any other journal.

Funding

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

Notes

ORCID iDs

Akmal Amirov

Ravshan Nurimbetov

Nurali Arabov

Burkhon Boltaev

Habibullo Hasanov

Ikrom Nasimov

References

Ada

(2014). Human capital and economic growth: A panel data analysis with health and education for the MENA region. Advances in Management and Applied Economics , 4(4), 1–6. http://www.scienpress.com/Upload/AMAE/Vol%204_4_6.pdf

Adeboje

O. M.

, Folawewo

, & Adedokun

A. J.

(2022). Trade integration, growth, and employment in West Africa: Implications for African Continental Free Trade Area (AfCFTA) . Research Square.

Altiner

, & Toktaş

(2017). Relationship between human capital and economic growth: An application to developing countries. Eurasian Journal of Economics and Finance , 5(3), 87–98. https://doi.org/10.15604/EJEF.2017.05.03.007

Arabov

, Belotserkovskaya

, Nasimov

, Utemuratova

, Abduraimov

, & Makhmudova

(2023). Improving methods for assessing the effectiveness of tourism infrastructure development. In E3S Web of Conferences (Vol. 452, p. 07012). EDP Sciences.

Arnold

J. M.

, Bassanini

, & Scarpetta

(2007). Solow or Lucas?: Testing growth models using panel data from OECD countries . Research Papers in Economics. https://ideas.repec.org/p/oec/ecoaaa/592-en.html

Aydin

, & Oguzhan

(2023). The effects of green innovation, environmental taxes, and financial development on renewable energy consumption in OECD countries. Energy , 280, 128105. https://doi.org/10.1016/j.energy.2023.128105

Bal

, Algan

, Manga

, & Kandır

(2014). The relationship between human capital and economic growth: Cases of BRICS countries and Turkey. Proceedings of International Conference of Eurasian Economies 2014 , pp. 795–803, Skopje, Macedonia. https://doi.org/10.36880/C05.00923

Banik

, & Bahumik

P. K.

(2006). FDI, skilled labour and the Caribbean’s emerging competitive disadvantage. Global Business Review , 7(2), 174–193.

Banik

, & Khatun

(2012). Role of knowledge in supporting growth across Indian states: A co-integration and causality approach. Modern Economy , 3(3), 345–353.

10.

Barro

R. J.

(2001). Human capital and growth. American Economic Review , 91(2), 12–17. https://doi.org/10.1257/aer.91.2.12

11.

Bashir

M. S.

, Ali

, & Elamin

A. A. H.

(2024). Human capital and economic growth in Sub-Saharan African countries: A panel data analysis. مجلة العلوم الاقتصادية والادارية, 30(144), 414–426. https://doi.org/10.33095/kdgw1963

12.

Becker

G. S.

(1964). Human capital: A theoretical and empirical analysis, with special reference to education . University of Chicago Press.

13.

Bloom

D. E.

, Canning

, & Sevilla

(2004). The effect of health on economic growth: A production function approach. World Development , 32(1), 1–13. https://doi.org/10.1016/j.worlddev.2003.07.002

14.

Cappelli

, Ridolfi

, & Vasta

(2023). Human capital in a historical perspective. In Oxford research encyclopedia of economics and finance. https://doi.org/10.1093/acrefore/9780190625979.013.520

15.

Fahim

, & Rhanami

(2018). Trade liberalization, human capital and economic growth in countries from the MENA region: Empirical analysis of panel data. Information Systems Management and Innovation (ISMI) , 2(1), 32–49. https://doi.org/10.34874/IMIST.PRSM/ISMI/7872

16.

Gemmell

(1995). Endogenous growth, the Solow model, and human capital. Economics of Planning , 28(2), 169–183. https://doi.org/10.1007/BF01263636

17.

Hanushek

E. A.

, & Woessmann

(2008). The role of cognitive skills in economic development. Journal of Economic Literature , 46(3), 607–668. https://doi.org/10.1257/jel.46.3.607

18.

Kharel

, Acharya

, & Upadhyaya

Y. M.

(2025). Exploring the link between education investment and economic growth: Evidence from a study of Nepal. Molung Educational Frontier , 15(01), 332–357. https://doi.org/10.3126/mef.v15i01.74096

19.

Kim

, & Park

(2021). Human capital and economic growth: A panel data analysis with a global perspective. European Journal of Political Economy , 68, 102004. https://doi.org/10.1016/j.ejpoleco.2021.102004

20.

Lee

, & Nguyen

(2019). Human capital, technological progress and economic growth: Evidence from Asia. Applied Economics Letters , 26(15), 1234–1238. https://doi.org/10.1080/13504851.2019.1605261

21.

Liao

, Du

, Wang

, & Yu

(2019). The impact of educational investment on sustainable economic growth in Guangdong, China: A cointegration and causality analysis. Sustainability , 11(3), 766. https://doi.org/10.3390/su11030766

22.

Lucas

R. E.

(1988). On the mechanics of economic development. Journal of Monetary Economics , 22(1), 3–42. https://doi.org/10.1016/0304-3932(88)90168-7

23.

Meilisa

, Elfindri

, Anas

, & Ariyanto

(2024). Short and long-term impact of educational enrollment rates on economic growth: For Indonesia’s sustainable development of human capital. Current Advanced Research on Sharia Finance and Economic Worldwide , 3(2), 235–249. https://doi.org/10.55047/cashflow.v3i2.1029

24.

Mutunga

(2023). Financial integration, trade integration, and economic growth in the East African Community. Scholars Journal of Economics, Business and Management , 10(8), 178–190.

25.

Namozov

, Arabov

, Aduramanov

, Nasimov

, Makhmudova

, & Ibragimov

(2024). Dynamics of transformation and economic resilience in agricultural lands: An anthropogenic perspective. In E3S Web of Conferences (Vol. 541, p. 03003). EDP Sciences.

26.

Nasreen

, Anwar

, & Waqar

M. Q.

(2015). Institutions, investment, and economic growth: a cross-country and panel data study. The Singapore Economic Review , 60(04), 1550061. https://doi.org/10.1142/S0217590815500617

27.

Oketch

, & Oduor

(2022). The impact of human capital on economic growth: Evidence from African countries. Applied Economics , 54(21), 2384–2397. https://doi.org/10.1080/00036846.2022.2054801

28.

Romer

P. M.

(1990). Endogenous technological change. Journal of Political Economy , 98(5, Part 2), S71–S102. https://doi.org/10.1086/261725

29.

Safarov

, Amirov

, Mansurova

, Hassan

T. H.

, Hasanov

, Pereș

A. C.

, Bilalov

, & Turdibekov

(2024). Prospects of agrotourism development in the region. Economies , 12(12), 321. https://doi.org/10.3390/economies12120321

30.

Schultz

T. W.

(1961). Investment in human capital. The American Economic Review , 51(1), 1–17.

31.

Sharma

(2016). Financial development, human capital, and economic growth: A cross-country analysis [Master’s thesis]. https://thekeep.eiu.edu/theses/2430/

32.

Sheidaei

, & Tash

M. S.

(2014). The cumulative effect of human capital on economic growth: Using panel data method. The Romanian Economic Journal , 17(52), 95–115. https://ideas.repec.org/a/rej/journl/v17y2014i52p95-115.html

33.

Smith

(1776). An inquiry into the nature and causes of the wealth of nations . W. Strahan and T. Cadell.

34.

Sulisnaningrum

, Widarni

E. L.

, & Bawono

(2022). Causality relationship between human capital, technological development, and economic growth. Journal of Management, Economics, and Industrial Organization . https://doi.org/10.31039/jomeino2022.6.2.1

35.

Torres

, & Delgado

(2023). Human capital development and economic growth: Empirical evidence from Latin America. World Development , 167, 106285. https://doi.org/10.1016/j.worlddev.2023.106285

36.

Venkateswaran

P. S.

, Sujatha

, Nasimov

, Arabov

, Rajest

S. S.

, & Nomula

V. K.

(2024). A study on the impact of intelligent systems on human resource management. In Optimizing intelligent systems for cross-industry application (pp. 153–174). IGI Global Scientific Publishing.

37.

Xie

, & Yang

(2020). Solow to Becker-Lucas . Social Science Research Network. https://doi.org/10.2139/SSRN.3465369

38.

Yormirzoev

(2023). Human capital and economic growth in Central Asia. Post-Communist Economies , 35(6), 533–545. https://doi.org/10.1080/14631377.2023.2196872

39.

Yousfi

, & Arfa

(2023). Human capital, institutions, and economic growth: Evidence from panel data analysis. International Journal of Economics and Financial Issues , 13(1), 1–10.

40.

Zhang

, & Ahmed

(2020). Education, health, and economic growth in developing countries: A cross-country empirical study. Sustainability , 12(9), 3745. https://doi.org/10.3390/su12093745

41.

Zhu

(2013). Research on human capital spillover and economic growth based on the cross-countries panel data . Commercial Research.