Education is often considered the basis for developing human capital in all developing countries. Recent studies have examined the relationship between education and economic growth in Nigeria. For instance, Chima and Yusuf (2023) used the Autoregressive Distributed Lag (ARDL) model to demonstrate that investment in healthcare and education has a positive impact on the GDP growth rate in the short and long term. Therefore, education plays a vital role in Nigeria’s economic growth.

Similarly, Wang and Zhang (2024) found that public spending on education had a stimulatory effect on economic growth in countries with stable and relatively more developed economies. They used a panel Ordinary Least Squares (OLS) model with data spanning from 1995 to 2015 for China and Malaysia as a case study. The study also investigated the impact of six factors related to government spending on education and found that investment in secondary schools was strongly associated with prominent economic returns. Also, Diakodimitriou et al. (2025) used econometric techniques such as adaptive LASSO and the 3SLS model on data from 1996 to 2019 for Germany, France, Greece, Italy, Spain and Portugal. They showed that education expenditure significantly impacted GDP per capita and identified R&D expenditure as the main channel through which education expenditure influences economic growth. Ziberi et al. (2022) investigated the relationship between increased public spending on education and economic growth in North Macedonia. Using data from 1917 to 2020 and the two-stage least squares instrumental variables method, they found that an increase in public spending on education had a positive impact on economic growth.

However, not all researchers agree that educational expenditure has a positive effect on economic growth. Some studies have suggested that educational spending could have a negative effect on economic development. For example, Abbah et al. in their study on government spending on human capital formation and economic growth in Nigeria found that educational expenditures had a negative impact in the short-term when using a vector error correction model with data from 1992-2021. They noted that the immediate impact of educational spending on growth could not be seen in the short run. Trinh (2025) also studied the impact of education expenditure on the economic growth of Southeast Asian countries. Using Generalized Method of Moments on data from 1998 to 2022, they found negative coefficients for education expenditure and positive coefficients for institutional quality. This suggests that the impact of education expenditure on growth depends on the quality of governance. Degefa and Daba (2025), in their study on the impact of public expenditure on economic growth in Ethiopia used an Autoregressive Distributed Lag Model and data from 1970 to 2021 and found a negative relationship between education expenditure and economic growth in the short run. In the same vein, Abaneme and Aworinde (2025) investigated the effect of government education expenditure on inclusive economic growth in Nigeria using data from 1990 to 2023 and by constructing ARDL model found that education expenditure had a negative impact on growth in both the short and long term.

Despite the substantial body of research on education expenditure and economic growth, the findings remain mixed, particularly in the context of developing countries such as Nigeria. While some studies show a positive relationship, others reveal negative or insignificant effects, often attributing this to poor governance. Therefore, this study aims to examine both the scale of education expenditure and the quality of the institutional frameworks through which it can stimulate economic growth.

Healthcare is another essential component of human capital development, which affects life expectancy, labor productivity, savings, investment and workforce participation in economic activity. A healthy lifestyle is essential for the population to be productive. Access to high-quality healthcare increases the efficiency of the workforce. As is commonly said, a healthy nation is a wealthy nation.

Many researchers have found links between healthcare spending and economic growth in different countries. For instance, Akinbode et al. (2021) investigated the impact of health on economic growth in sub-Saharan Africa using the system-generalized method of moments (SGMM). The study found that health expenditure had a significant impact on economic growth in the countries under review. This led the authors to emphasize the importance of healthcare in the region’s economic development. Hu and Wang (2024) explored the effects of public health expenditure on economic growth in 33 OECD countries between 2001 and 2017. They constructed a nonlinear dynamic panel threshold model and used it to empirically analyze the threshold effect of public health expenditure on economic growth. They found that in conditions of higher household consumption, employee wages and physical capital investment, public health expenditure will make a significant contribution to economic growth. This is largely true for Nigeria as well.

A study carried out by Ikpe et al. (2024), based on data from 27 Sub-Saharan African countries over the period 2005-2021, identified the role of governance quality in the relationship between health expenditure and economic growth. Using the two-step dynamic panel data - system generalized method of moments (DPD-SGMM) estimation technique - the authors established a significant positive relationship between health expenditure and economic growth. Sosvilla-Rivero et al. (2025) examined the impact of public expenditure on economic growth in 28 European Union (EU) countries between 1995 and 2022. They estimated a growth model for public expenditure using the Autoregressive Distributed Lag panel data approach and found that healthcare expenditure had a positive effect on economic growth over the long term. Similarly, in their study of the relationship between healthcare spending and economic growth among the BRICS countries, Dritsaki et al. (2025) used a panel cointegrated approach with data from 2000-2021 and a dynamic panel ARDL model, finding that healthcare spending had a positive and significant impact on long-term economic growth. Furthermore, using panel data spanning from 1980 to 2015, Cui and colleagues found a positive correlation between healthcare expenditures and economic development in the long run. Studies by Ozyilmaz et al. (2022), Govdeli (2023), Sethi et al. (2024), Hu and Wang (2024) also found a positive relationship between health expenditure and economic growth.

Other researchers have found that healthcare expenditure has a negative impact on economic growth. For example, Faruk et al. (2022) examined the relationship between healthcare spending, institutions and economic growth in MENA countries using ordinary least squares (OLS) with panel data from 2000 to 2017. They concluded that increased healthcare spending does not lead to higher economic growth. Similarly, Islam et al. (2023) found that healthcare expenditure exhibited a strong negative relationship with GDP growth in Saudi Arabia, using data from 1990 to 2019 and a vector error correction model.

A careful synthesis of previous research reveals that studies have employed a variety of methodologies to investigate the relationship between human capital and economic growth in different countries. However, none of these studies have considered the role of governance quality in translating education and health expenditure into economic growth in Nigeria. Furthermore, no previous study has employed the novel dynamic ARDL model, which illustrates the short- and long-term effects of independent variables on the dependent variable. This study uses Nigeria as a case study to make a new contribution to our understanding of the relationship between human capital and economic growth. It applies methods such as the dynamic ARDL model and the Bai-Perron and Chow tests to confirm structural breaks.

The functional model for this study is specified as follows:

RGDP_t = β₀ + β₁GEXE_t + β₂GEXH_t + β₃INSQ_t + β₄GCF_t + ε_t (1)

where β₀ is constant term, β₁ – β₄ are coefficients of the independent variables, RGDP is real gross domestic product, a proxy for economic growth, GEXE is government expenditure on education, GEXEH is government expenditure on health, INSQ is institutional quality, GCF is gross capital formation, ε is the error term.

Based on the methodology proposed by Jordan and Phillips (2018) and applied by Shamwil and Malkina (2025), this study develops the modified dynamic model of the following form:

∆lnRGDP_t = α₀ + α₁lnRGDP_{t – 1} + γ₁∆lnGEXE_t + ϕ₁lnGEXE_{t – 1} + γ₂∆lnGEXH_t + + ϕ₂lnGEXH_{t – 1} + γ₃∆INSQ_t + ϕ₃INSQ_{t – 1} + γ₄∆lnGCF_t + ϕ₄lnGCF_{t – 1} + μ_t. (2)

This study suggests that institutional quality can act as a moderating component in the relationship between economic growth, education expenditure and health expenditure. Following the work of Udeagha and Breitenbach (2023), equation (2) is modified by including the interactive term to account for the effect. Thus, the new dynamic simulation is specified as follows:

∆lnRGDP_t = α₀ + α₁lnRGDP_{t – 1} + γ₁∆lnGEXE_t + ϕ₁lnGEXE_{t – 1} + + ψ1 * (lnGEXE_t * INSQ_t) + γ₂∆lnGEXH_t + ϕ₂lnGEXH_{t – 1} + ψ₂ * (lnGEXH_t * INSQ_t) + γ₃∆INSQ_t + ϕ₃INSQ_{t – 1} + + γ₄∆lnGCF_t + ϕ₄lnGCF_{t – 1} + μ_t. (3)

Table 5 summarises the statistics for the variables used in the model. The results show that the LRGDP, with a mean value of 7.59, exhibits a relatively stable growth pattern over the review period, with a low standard deviation of 0.23 confirming moderate fluctuations and a generally smooth growth trajectory. LGEXE and LGEXH record mean values of 4.87 and 4.20 respectively, with comparatively higher standard deviations of 1.19 and 1.44, implying relative volatility in spending on social sectors in Nigeria. LGCF exhibits a higher mean value of 8.86 with standard deviation of 0.94, indicating moderate variability in investment levels over time. INSQ with standard deviation of 1.67 demonstrates substantial variations, indicating noticeable changes in governance and institutional performance throughout the study period.

The skewness coefficients of most variables are negative, indicating slight left-skewness, while gross capital formation is distributed fairly symmetrically. Kurtosis values for all variables are less than 3, implying platykurtic distributions and the absence of outliers. Furthermore, the Jarque-Bera statistics and their associated probabilities suggest that the null hypothesis of normality cannot be rejected for any of the variables except for INSQ. This is consistent with the findings of Chuba and Muse (2025).

Table 1 shows the results of the augmented Dickey–Fuller (ADF) stationarity test for the variables. The results show that all variables are stationary after first differencing, meaning that all variables used in the model are integrated of order one, and therefore the null hypothesis that the variables are non-stationary is rejected.

The Bai-Perron test was used to check for structural breaks in the variables. As can be seen in Table 3, there is confirmation of a structural break in 2011. This break was corrected using the Chow Breakpoint test (see Table 4), which indicates correction for structural breaks at a highly significant level (0.0000).

When working with the autoregressive distributed lag (ARDL) model, it is important to choose the correct lag. This helps to eliminate serial correlation in the residuals, which can lead to biased results. In this study, lag 3 was selected as the appropriate lag for the model based on the minimization of the Akaike information criterion, the Schwarz information criterion and the Hannan-Quinn information criterion.

The bounds test is a vital tool for evaluating cointegration among variables in time series analysis. According to the results presented in Table 6, the calculated F-statistic is 32.31739. This value is higher than the upper critical bound corresponding to a 5% significance level. In other words, there is less than a 5% probability that this result occurred by chance in the absence of a long-run relationship between the variables. Therefore, we reject the null hypothesis of no cointegration and conclude that a statistically significant long-run equilibrium relationship exists between the variables.

The results in Table 7 provide strong evidence of a stable long-run relationship between economic growth and government spending on education and health care in Nigeria. This is confirmed by the error correction term, which is negative and highly significant (-0.565; p = 0.005). The magnitude of the coefficient implies that approximately 56.5% of short-run deviations are corrected within one period, indicating rapid speed of adjustment. This suggests that although short-term shocks affect output, growth dynamics are anchored by long-term human development and institutional factors.

In the long run, government expenditure on health (GEXH) exerts a statistically significant positive effect on economic growth. The first lag of health expenditure (L1lnGEXH) is positive and significant (0.0764263; p = 0.020), indicating that a 1% increase in health spending raises real GDP by about 0.08% in the subsequent period. This supports the view that health expenditure enhances labour productivity and economic performance in the long run. This finding is consistent with that of Atilgan et. al. (2024), Hu and Wang (2024), Ginn (2025), and Ridhwan et al. (2022). However, the interaction term between health expenditure and institutional quality is negative and significant (-0.1938134; p = 0.016). This suggests that, during periods of unfavourable institutional conditions or structural stress, health spending reduces growth. This implies that inefficiencies or fiscal pressures can offset productivity gains of health expenditure under unfavourable institutional conditions. This is in line with the findings of Faruk et al. (2022), Dritsaki et al. (2025), Boundioa and Thiombiano (2025), as well as Alghannam and Alharbi (2026).

The impact of government education expenditure (GEXE) on growth is both dynamic and conditional. While the immediate effect is negative, the second lag of education expenditure (L2lnGEXE) is positive and statistically significant (0.0226635; p = 0.016). This indicates that education spending contributes to growth only after a time lag, consistent with delayed returns associated with human capital accumulation. This is supported by the findings of Contreras et al. (2025) and Sorto-Bueso et al. (2026). Furthermore, the interaction term between education expenditure and institutional quality is positive and significant (0.1707715, p = 0.019), indicating that education spending becomes more growth enhancing when implemented under supportive policy or institutional regimes. This highlights the importance of complementary governance structures in translating education expenditure into economic growth. The result is consistent with that of Abu-Alfoul et al. (2024), and Liko (2024).

Institutional quality (INSQ) also plays a significant role in shaping growth outcomes. The lagged institutional quality variable (L1lnINSQ) is negative and significant (-0.0077867; p = 0.038), indicating that weaknesses in institutional frameworks in the previous period reduce current economic growth. The result suggests that institutional weaknesses have persistent adverse effects, undermining the effectiveness of public spending and slowing the growth process. The study by Karabou (2024) supports this finding.

Figure 1 shows the impulse response graph of the dynamic ARDL model, indicating the response of economic growth to positive and negative 10% shocks in government health expenditure over time. In the short-run, between periods 1 to 10, the response is relatively weak and nearly flat, suggesting limited immediate transmission effects. The result indicates that a 10% increase in government expenditure on health will cause economic growth in the long run. Conversely, a 10% decrease in government expenditure on health will lead to a decline in economic growth in Nigeria over time.

Figure 1. 

The impact of changes in government health expenditure on economic growth in Nigeria, %. Note: The dark blue to light blue lines in Figures 1, 2, 3 represent 75%, 90%, and 95% confidence intervals, respectively, the dots represent the average prediction value. Source: authors’ computation using Stata 17.

Figure 2 shows the dynamic ARDL’s impulse response graph, illustrating the time path of economic growth following a 10% shock to government expenditure on education over a 30-period horizon. According to the findings, a ten percent increase in government spending on education will eventually boost economic growth. However, over time, Nigeria’s economic growth will slow if government spending on education is cut by 10%.

Figure 2. 

The impact of change in government expenditure on education on economic growth in Nigeria, %.

Figure 3 presents the impulse response function obtained from the dynamic ARDL model, indicating how economic growth responds to positive and negative shocks in institutional quality. The results indicate that a one percent increase in institutional quality enhances governance effectiveness, strengthens accountability and reduces resource misallocation, thereby promoting the efficient use of public funds and stimulating economic expansion. However, if institutional quality drops by 10% over time, Nigeria’s economic growth will decelerate.

Figure 3. 

The impact of change in institutional quality on economic growth in Nigeria, %

Table 8 reports diagnostic tests performed to confirm stability and reliability of the model used.

Figure 4 shows that the CUSUM statistic remains within the 5% critical bounds throughout the estimation period. Despite minor fluctuations, the plot does not cross the significance thresholds. This suggests the absence of systematic parameter drift and confirms the stability of estimated coefficients. Similarly, the CUSUMSQ test remains within the 5% critical region. This conformity within the bounds indicates no evidence of heteroskedastic instability. Overall, the tests confirm the dynamic stability of the dynamic ARDL specification.

Figure 4. 

CUSUM and CUSUMSQ tests. Source: Authors’ computation using EVIEWS 12