Research Article
Print
Research Article
Does Working Age Population Matter for the Impact of Foreign Direct Investment and Credit on Productivity Growth? A Cross-Country Empirical Evidence from ASEAN
expand article info Arintoko‡, Herman Sambodo‡, Rakhmat Priyono‡
‡ Universitas Jenderal Soedirman, Purwokerto, Indonesia

Corresponding author: Arintoko ( arintoko@unsoed.ac.id )

© 2026 Arintoko, Herman Sambodo, Rakhmat Priyono.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation: Arintoko, Sambodo H, Priyono R (2026) Does Working Age Population Matter for the Impact of Foreign Direct Investment and Credit on Productivity Growth? A Cross-Country Empirical Evidence from ASEAN. Population and Economics 10(2): 230-248. https://doi.org/10.3897/popecon.10.e155357
Open Access

Abstract

Demographic factors remain crucial issues in economic development. The size and age composition of the population greatly affect a country’s economic growth. This study aims to investigate the effect of the age dependency ratio on labor productivity through the moderating role of foreign direct investment (FDI) and private sector credit. The study applies the Arellano–Bond Generalized Method of Moments (GMM) using a sample of five selected ASEAN countries over the period 2000–2023. The results reveal that the age dependency ratio weakens the positive effect of foreign direct investment on productivity. Moreover, the age dependency ratio plays a more significant role as a moderating variable than as an independent variable in influencing productivity in ASEAN countries. Policies aimed at increasing productivity among the working-age population (15–64 years) should be a main priority in addressing employment challenges, particularly through improving the quality of education and training and expanding employment opportunities via investment driven by FDI and private sector credit. In addition, birth control efforts through family planning programs and the prevention of early marriage should receive greater attention from governments.

Keywords

age dependency ratio, foreign direct investment, labor productivity, Arellano-Bond GMM model, private sector credit

JEL codes: C33, J11, J24, O47

Introduction

Issues related to the relationship between population and economic growth remain widely debated. Population can be viewed either as a driver or as a constraint on economic growth. In high-income countries, low population growth is considered a challenge for sustaining economic growth (Peterson 2017). Conversely, in low-income countries, high population growth can hinder economic growth. Brida et al. (2024) reveal that there are distinct differences in the relationship between population growth and economic growth across mature, transition, and young economies. Mature economies, mostly OECD countries, are characterized by low population growth and high economic growth. In contrast, young economies, such as those in Central Africa, are characterized by high population growth but low economic growth. Meanwhile, transition economies experience initially high population growth during the early stages of development, followed by declining population growth while maintaining above-average economic performance.

Empirical findings on the relationship between population and economic growth have not resolved the longstanding controversy between the two. There is no theoretical consensus on the scope and channels through which population and economic growth influence each other (Cayssials et al. 2024). However, addressing this controversy requires a more nuanced understanding that population issues should not be viewed solely in aggregate terms. Population dynamics can be further examined through age structure. The age structure of the population reflects the composition of working-age and non-working-age groups, including both young and elderly dependents. Wang and Lee (2019) show that age structure has a nonlinear effect on economic growth. Age structure is commonly represented by the age dependency ratio, which consists of child and old-age dependency ratios. The child dependency ratio exhibits a U-shaped effect, while the old-age dependency ratio shows an inverted U-shaped effect on economic growth. Meanwhile, Bidisha et al. (2020) provide empirical evidence that age structure, as measured by the age dependency ratio, has a long-run effect on economic growth measured by GDP per capita. Specifically, the age dependency ratio has a significant negative impact on GDP per capita growth. Therefore, an increase in the working-age population relative to the non-working-age population, reflected in a decline in the age dependency ratio, is associated with higher GDP per capita growth.

The issue of economic growth related to the population’s age structure has been widely studied. However, studies examining economic growth in relation to population age structure generally measure growth in terms of GDP growth or GDP per capita. There are still relatively limited studies that interpret economic growth in terms of productivity growth. GDP growth primarily reflects the expansion of aggregate national income, while GDP per capita growth captures changes in average income from the perspective of income recipients, namely the population. In contrast, output growth inherent in the concept of aggregate labor productivity is more appropriate when associated with the population’s age structure. Moreover, labor productivity, as a proxy for technological progress, has a dominant positive impact on economic growth, as shown in a cross-regional study by Badriah and Arintoko (2024).

The relationship between population age structure and productivity growth has generally been examined with an emphasis on the importance of the working-age population for productivity. Population age structure is commonly represented by the age dependency ratio. A higher age dependency ratio implies a lower proportion of the productive-age population relative to the non-working population, which includes both young and elderly groups. Fu et al. (2020) provide empirical evidence that changes in age structure, reflected in increasing total dependency and child dependency ratios, have a negative impact on labor productivity across provinces in China. Differences in population dividends and levels of urbanization contribute to heterogeneous development patterns across provinces.

ASEAN is one of the most interesting regional areas for examining the relationship between population factors and productivity growth. The ASEAN region contributes the third-largest share of the global population, after India and China. However, there are substantial variations in population characteristics across member countries. Trong et al. (2024) emphasize that demographic variables play an important role in driving per capita output growth.

Labor productivity is closely associated with a country’s competitiveness in supporting economic development (Bangun 2017). Labor productivity in the five selected ASEAN countries, as illustrated in Figure 1, shows an overall upward trend across all countries. Nevertheless, significant differences and variations in productivity levels persist among them. The five selected countries are emerging economies characterized by relatively advanced economic structures and high levels of investment aimed at achieving rapid growth. These countries also represent the five most populous ASEAN members, excluding Myanmar. Considerable variation in labor productivity is observed, with Malaysia exhibiting the highest productivity level and Vietnam the lowest.

Figure 1.

Output per worker in five selected ASEAN countries, 2000–2023. Source: World Bank

Along with increasing labor productivity, the age dependency burden tends to decline, as shown in Figure 2. Variations in the age dependency burden reflect differences in population age structures among the five countries. The Philippines exhibits the highest dependency burden, while Thailand has the lowest. Accordingly, the share of the working-age (productive-age) population in the Philippines is relatively smaller compared to that in Thailand.

Figure 2.

Age dependency ratio in selected ASEAN countries, 2000–2023. Source: World Bank

A key question arises: does the age structure of the population affect labor productivity? More specifically, does the burden of age dependency dampen productivity when foreign direct investment (FDI) and private sector credit increase? FDI and credit are key factors that can stimulate employment and enhance productivity in growth-oriented economies. Accordingly, this study examines whether the age dependency burden moderates the effects of FDI and private sector credit on productivity when both variables exert a positive influence on productivity.

Empirical evidence from several previous studies (including Tőkés 2019; Asada 2020; Yunus and Abdullah 2022) shows that FDI has a positive impact on productivity. Meanwhile, other studies (including Manaresi and Pierri 2024; Pham and Phan 2024) provide evidence that credit also positively affects productivity. Banks remain the primary source of credit for the private sector, as they continue to rely heavily on interest income and maintain high loan-to-deposit ratios (Arintoko et al. 2024). Alongside increases in FDI and private sector credit, ASEAN countries have also experienced rising government debt and persistent budget deficits. Similarly, financial development in the region has been accompanied by growth in broad money.

According to the production function developed by Solow, output per worker is a function of capital per worker (Mankiw 2016). In Solow’s growth theory, labor alone cannot drive growth in output per worker; rather, it must interact with capital. Capital accumulation occurs through investment, which may originate from foreign direct investment and private sector credit. Previous studies on the determinants of labor productivity typically treat the working-age population as an independent factor influencing productivity. Accordingly, the age dependency ratio is often included as an independent variable affecting productivity, as shown in studies by Huang et al. (2019) and Fu et al. (2020).

If the effect of the age dependency ratio is not purely independent, it may operate through interactions with FDI and credit in influencing productivity. This study is therefore motivated to develop a model of productivity determinants that incorporates moderating effects. The age dependency ratio is specified as a moderating variable, given its close association with employment dynamics arising from investment driven by FDI and private sector credit. An increase in the age dependency ratio implies a decline in the share of the working-age population, which may constrain labor productivity even in the presence of rising FDI and private sector credit, alongside increasing government debt and broad money.

The moderating role of the age dependency ratio is examined using the Arellano–Bond GMM framework, with labor productivity as the dependent variable. The explanatory variables include foreign direct investment, private sector credit, government debt, and broad money. In this specification, the age dependency ratio moderates the effects of FDI and credit on aggregate labor productivity.

Literature review

Economic growth cannot be separated from a country’s population structure and population growth. Population structure refers to the composition of the working-age population in terms of age and skills, which has a significant impact on labor productivity. Consequently, changes in population structure affect not only employment levels but also productivity across different population groups. The working-age population typically includes individuals aged 15 to 64 years. This age group constitutes the labor force, which consists of individuals who are either employed or actively seeking employment. The composition of the working-age population, including its age distribution and skill levels, influences aggregate employment, labor productivity, and overall economic growth.

Huang et al. (2019) provide empirical evidence that workforce aging has a complex relationship with economic growth. While the old-age dependency ratio has a negative effect on economic growth, human capital plays a crucial role in total factor productivity (TFP). Thus, both labor force characteristics and population age structure significantly affect productivity through TFP. In contrast, empirical findings by Hernæs et al. (2023) offer different evidence, showing that an increase in the proportion of workers aged 63–67 has a small positive effect on labor productivity in the short run.

The study by Maestas et al. (2023) finds that population aging, reflected in an increasing share of the population aged 60 and above, leads to a decline in GDP per capita growth. This decline arises partly from slower employment growth and more substantially from reduced labor productivity growth. Similarly, Trong et al. (2024) show that old-age dependency reduces GDP per capita growth in ASEAN countries. Meanwhile, productive young workers remain a key resource for economic development and GDP per capita growth. Overall, demographic variables play a significant role in driving GDP per capita growth.

Choudhary et al. (2016) provide empirical evidence that the population’s age composition affects labor productivity growth. A higher age dependency ratio is associated with lower labor productivity. Moreover, age dependency can modify the impact of other factors on productivity. Both child and old-age dependency ratios exert negative effects on labor productivity, although with differing magnitudes.

Population aging shapes age composition and age dependency ratios through an increasing share of elderly individuals relative to the child population. Alarussi and Yen (2023) show that population aging has a negative impact on labor productivity. Within the age dependency ratio framework, the proportions of child and old-age dependents jointly determine the total dependency ratio. An increase in the total age dependency ratio implies a reduction in the share of the working-age population. In addition to high fertility episodes, population aging contributes to rising age dependency. As population aging reduces the working-age population while simultaneously increasing the elderly population, economic growth may decline (Lee and Shin 2019), partly through reduced productivity. Overall, empirical evidence consistently indicates that the age dependency ratio negatively affects productivity. For example, Park and Shin (2023) find that both old-age and young-age dependency ratios have adverse effects on labor productivity.

The age dependency ratio is closely related to foreign direct investment (FDI) in influencing productivity, as this variable directly affects employment. Nguyen et al. (2024) show that increasing FDI has a positive impact on employment creation. Through FDI, foreign firms tend to have greater capacity to generate employment than domestic firms, creating spillover effects from foreign to domestic enterprises (Dao et al. 2023). Saucedo et al. (2020) further find that rising FDI inflows positively affect both low- and high-skilled employment in the manufacturing sector.

With respect to productivity, Dandi et al. (2024) report that FDI positively affects labor productivity in ASEAN countries. Similarly, Tőkés (2019) demonstrates that, at the industry level, FDI has a positive impact on labor productivity. Using micro-level data, Karentina (2019) finds that FDI enhances labor productivity in the long run. According to Vinh (2019), this positive effect of FDI on labor productivity primarily arises from technology transfer.

In this study, the age dependency ratio, which is closely associated with FDI through its effect on employment, is treated as a moderating variable in the relationship between FDI and productivity. Consistent with the Solow growth model, population factors are assumed to interact with capital in determining output per worker. Capital accumulation occurs through investment, one of the main sources of which is FDI inflows. Through this interaction, an increase in the age dependency ratio is expected to weaken the positive effect of FDI on productivity growth.

In addition to FDI, the interaction between the age dependency ratio and private sector credit also influences productivity growth. Regarding employment creation, an increase in bank credit directly affects employment (Ginting and Widyawati 2022). Acosta and Cortés (2022) and Gutierrez et al. (2023) show that positive credit shocks significantly enhance employment creation. Bank credit can stimulate private sector investment, thereby increasing employment opportunities for the working-age population. The age dependency ratio, when acting as a moderator of private sector credit, is expected to reduce the positive effect of credit on productivity.

Beyond the moderating role of the age dependency ratio on FDI and credit, other factors influencing labor productivity growth include government debt and financial sector development, proxied by the money supply. These control variables can affect labor productivity growth when the age dependency ratio moderates the impact of FDI and credit. According to Afonso and Jalles (2013), the government debt ratio can support the growth of total factor productivity, including labor productivity. Higher government debt allows for increased public investment, which can expand employment and enhance labor productivity. Similarly, Haris and Mohammad (2015) find that government debt can lead to more effectively allocated government investment, contributing to productivity growth in both the manufacturing and service sectors.

Money supply is positively associated with labor productivity growth. A tight monetary policy stance, which reduces the money supply, can suppress both productivity and employment (Fornaro and Wolf 2023). Such tightening negatively affects investment and productivity growth. Conversely, a loose monetary policy that expands the money supply can stimulate investment and enhance productivity. Jolaiya (2024) further emphasizes that increases in money supply and private sector credit positively affect domestic investment. Higher domestic investment, in turn, provides opportunities to boost productivity. Similarly, Dang et al. (2020) note that monetary policy can encourage private investment by increasing the money supply. Empirical evidence also indicates that, in addition to public investment, private investment strongly and positively impacts output growth (Turan et al. 2021). Therefore, money supply positively affects private investment and has the potential to enhance labor productivity.

Based on the previous literature, the following hypotheses are proposed:

H1a: FDI has a positive effect on labor productivity

H1b: The age dependency ratio moderates the positive effect of FDI on labor productivity.

H2a: Credit has a positive effect on labor productivity

H2b: The age dependency ratio moderates the positive effect of credit on labor productivity.

H3: Government debt has a positive effect on labor productivity

H4: Broad money has a positive effect on labor productivity

Methods

Data and Variables

In line with the objectives of this study, the model includes the following variables: GDP per worker, foreign direct investment (FDI), private sector credit, government debt, broad money, and the age dependency ratio. GDP per worker is treated as the dependent variable, while FDI, private sector credit, government debt, and broad money are considered explanatory variables. The age dependency ratio is specified as a moderating variable for FDI and credit.

The inclusion of the age dependency ratio as a moderating variable allows for the analysis of how demographic factors influence labor productivity growth through their interaction with FDI and credit, which are the primary sources of investment that increase the capital stock and, in turn, drive employment and output growth. The definitions and measurements of each variable are summarized in Table 1.

Table 1.
Download as
CSV
XLSX

Definitions and measurement of variables

Variable Abbreviation Definition Measurement
GDP per Worker GDPPW GDP per worker is the ratio of real gross domestic product to the number of people employed, and it measures productivity. The GDP used as the numerator is real GDP converted to international dollars on a constant basis in 2021. This GDP uses purchasing power parity (PPP) exchange rates. In the model, GDPPW is expressed in natural logarithms to smooth out relatively large variations in GDP.
Foreign Direct Investment FDI Foreign direct investment referred to in this study is the net investment inflow into the economy. The net investment calculated includes the amount of equity capital, reinvestment, other long-term capital, and short-term capital. Net investment inflows are shown in the balance of payments. FDI is measured in current US dollars and expressed as a percentage of GDP.
Credit CRE The credit referred to in this study is domestic credit by banks provided to the private sector. Bank credit is measured in current US dollars and expressed as a percentage of GDP.
Government Debt GDEBT The central government debt referred to in this study includes domestic and foreign liabilities. Included in domestic and foreign liabilities are currency and money deposits, securities other than stocks, and loans. Central government debt is calculated with the amount of equity and financial derivatives owned by the government as a deduction. Central government debt is measured at the end of the fiscal year and expressed as a percentage of GDP.
Broad Money BM Broad money supply is measured as the sum of currency outside banks plus demand deposits other than those held by the central government, plus time, savings, and foreign currency deposits held by the resident sector other than the central government, plus bank checks and travelers’ checks, and plus other securities such as certificates of deposit and commercial paper. Broad money supply is expressed as a percentage of GDP.
Age Dependency Ratio ADR The age dependency ratio is the ratio of dependents, namely people aged under 15 or over 64, to the working age population, namely those aged 15-64 years. The data measured is the proportion of dependents per 100 working age population.

Notes: All variables are expressed in natural logarithms in the model, except FDI. Source: World Development Indicators, World Bank

All the data are sourced from the World Development Indicators, World Bank, accessed online. The study covers the period 2000–2023, and the analysis uses annual data. The countries included in the study are Indonesia, Malaysia, the Philippines, Thailand, and Vietnam. These five selected countries are emerging ASEAN economies and were chosen because they are the five most populous ASEAN members, excluding Myanmar.

Model

The Arellano–Bond Generalized Method of Moments (GMM) is an econometric technique for estimating dynamic panel data models, particularly those that include lags of the dependent variable as predictors, as proposed by Arellano and Bond (1991). This method produces unbiased, consistent, and efficient estimates in dynamic panel data settings by using appropriate instruments to address serial autocorrelation and potential correlation between endogenous variables and the error term. It is specifically designed to handle endogeneity issues arising from the inclusion of lagged dependent variables as regressors.

In this study, the application of the Arellano–Bond GMM follows the general model specification used by Fan and Peng (2024), with adjustments to incorporate moderating variables. Several previous studies have applied the GMM method with moderating variables, including Santosa et al. (2023), Ullah et al. (2024), Bogdan et al. (2025), and Agustin et al. (2025).

The core models of the Arellano–Bond GMM in this study are presented in Equations (1)–(3). Equation (1) represents the baseline model without moderating variables. Equation (2) extends the baseline by including the age dependency ratio as an independent variable. Equation (3) is the full model, in which the age dependency ratio moderates the effects of FDI and private sector credit on labor productivity.

LGDPPWi,t = α0 + φLGDPPWi,t-1 + α1FDIi,t + α2LCREi,t + α3LGDEBTi,t + α4LBMi,t + u1i,t. (1)

The estimated parameters of equation (1) are as follows:

φ > 0, α1 > 0, α2 > 0, α3 > 0, α4 > 0,

LGDPPWi,t = β0 + ωLGDPPWi,t-1 + β1FDIi,t + β2LCREi,t + β3LGDEBTi,t + β4LBMi,t + β5LADRi,t + u2i,t. (2)

The estimated parameters of equation (2) are as follows:

ω > 0, β1 > 0, β2 > 0, β3 > 0, β4 > 0, β5 < 0,

LGDPPWi,t = γ0 + ϕLGDPPWi,t-1 + γ1(FDI * LADR)i,t + γ2(LCRE * LADR)i,t + γ3LGDEBTi,t + γ4LBMi,t + u3i,t. (3)

The estimated parameters of equation (3) are as follows:

ϕ > 0, α1 > 0, γ1 > 0, α2 > γ2 > 0, γ3 > 0, γ4 > 0.

The estimated coefficients of FDI and private sector credit, when moderated by the age dependency ratio, are lower than the corresponding coefficients in the baseline model.

The validity of the Arellano–Bond GMM model is assessed using two main diagnostic tests (Ashley and Sun 2016; Rebić et al. 2022; Lee et al. 2022). The first is the J-statistic test, which evaluates the validity of the instruments used. In this test, the null hypothesis assumes that the instruments are valid. The second is the Arellano–Bond autocorrelation test, which checks for the absence of second-order autocorrelation in the first-differenced residuals. Here, the null hypothesis states that there is no AR(2) autocorrelation.

In the difference GMM (Arellano–Bond GMM) method, which uses instruments based on lagged differences, first-differencing the variables serves to eliminate time-invariant effects (Odei et al. 2025). Stationarity of the differenced series is essential in this method; therefore, testing for stationarity is necessary to determine whether the series are stationary at levels or after first differencing.

Before testing the model’s validity, common unit root tests are conducted using the Levin, Lin, and Chu t-statistic, while individual unit root tests employ statistics such as Im, Pesaran, and Shin W-stat, ADF-Fisher χ2, and PP-Fisher χ2.

Results

The descriptive characteristics of the variables used in this study are presented in Table 2. Panel data from the five selected countries show the greatest variation in GDP per worker. This is reflected in the large difference between the maximum and minimum values, which corresponds to the highest standard deviation. The difference between the mean and median further indicates that the distribution of GDP per worker is skewed. Among the explanatory variables, private sector credit exhibits the largest variation, as reflected in its highest standard deviation, whereas FDI shows the smallest variation and the lowest standard deviation. The age dependency ratio displays relatively modest variation compared to credit, government debt, and broad money.

Table 2.
Download as
CSV
XLSX

Summary statistics for panel data

Variable Mean Median Maximum Minimum Std. Dev.
GDPPW 28235.33 23067.47 69044.74 8758.623 16046.54
FDI 2.7171 2.6308 9.6630 -2.7574 1.7876
CRE 83.150 84.256 203.5299 18.1557 50.3207
GDEBT 45.7466 43.900 88.7000 24.8000 11.9129
BM 91.4843 97.0641 148.6330 36.0017 36.8833
ADR 50.3374 48.6384 70.8441 38.8046 8.3147

Source: Authors’ calculations

In panel data, variation is generally larger than in time-series data for a single country because it captures both within-country variation over time and between-country variation across the sample.

To provide a more complete description of the data characteristics, Table 3 presents the correlations between GDP per worker and the following variables: FDI, the interaction of FDI with the age dependency ratio, private sector credit, the interaction of credit with the age dependency ratio, government debt, and broad money. Correlation measures the symmetrical relationship between two variables, without designating either as dependent or independent. Therefore, the correlation coefficient does not necessarily correspond to the regression coefficient in magnitude or sign.

Table 3.
Download as
CSV
XLSX

Lagged and lead correlations

Correlation i Lag Lead
With Lag With Lead
LGDPPW,FDI(-i) LGDPPW,FDI(+i) 0 -0.0201 -0.0201
1 0.0014 -0.0632
2 -0.0024 -0.0743
LGDPPW,FDI*LADR(-i) LGDPPW,FDI*LADR(+i) 0 -0.0317 -0.0317
1 -0.0097 -0.0736
2 -0.0125 -0.0830
LGDPPW,LCRE(-i) LGDPPW,LCRE(+i) 0 0.5918 0.5918
1 0.5647 0.5489
2 0.5377 0.5059
LGDPPW,LCRE*LADR(-i) LGDPPW,LCRE*LADR(+i) 0 0.5941 0.5941
1 0.5675 0.5524
2 0.5421 0.5102
LGDPPW,LGDEBT(-i) LGDPPW,LGDEBT(+i) 0 0.2586 0.2586
1 0.2239 0.2680
2 0.1917 0.2817
LGDPPW,LBM(-i) LGDPPW,LBM(+i) 0 0.5878 0.5878
1 0.5642 0.5465
2 0.5406 0.5059
LGDPPW,LADR(-i) LGDPPW,LADR(+i) 0 -0.3736 -0.3736
1 -0.3555 -0.3366
2 -0.3341 -0.3023

Source: Authors’ calculations

Correlations are reported with two lags and two leads to illustrate how relationships change over time. With lags and leads, the magnitude and sign of the correlation coefficient can vary. Strong correlations tend to maintain a consistent sign (positive or negative) across lags and leads, although the magnitude may increase or decrease.

GDP per worker exhibits the strongest correlation with the interaction between credit and the age dependency ratio, with consistency across the lags and leads. The correlation between GDP per worker and credit remains positive, and the coefficient increases when credit is moderated by the age dependency ratio. In contrast, GDP per worker is negatively correlated with the age dependency ratio.

The data stationarity test is an important step in applying the difference GMM method. Stationarity of the series at the first difference is a necessary condition for the difference GMM approach. If the series is not stationary, the instruments may be endogenous, potentially biasing the results. Table 4 presents the results of the unit root tests for all variables.

Table 4.
Download as
CSV
XLSX

Panel unit root test results

Variable Common Unit Root Individual Unit Root
Levin, Lin & Chu t Im, Pesaran and Shin W-stat ADF-Fisher PP-Fisher
LGDPPW -0.8625 1.9050 4.2570 5.0326
FDI -3.6984*** -3.8921*** 36.0183*** 56.3986***
FDI*LADR -3.6678*** -3.9358*** 36.1797*** 57.2733***
LCRE -2.9708** -1.1854 16.4980* 7.6480
LCRE*LADR -2.3146** -0.8245 17.0266 9.4261
LGDEBT -1.8360** -1.1188 14.0530 10.0016
LBM -1.6768** -0.9283 17.2533* 17.2509*
∆LGDPPW -7.4096*** -6.5266*** 55.7565*** 57.3565***
∆FDI -6.3758*** -7.1231*** 63.3752*** 233.771***
∆(FDI*LADR) -6.9772*** -7.6385*** 67.7047*** 233.809***
∆LCRE -5.9005*** -5.3547*** 46.3528*** 50.4251***
∆(LCRE*LADR) -5.9441*** -5.3489*** 46.1861*** 50.4081***
∆LGDEBT -5.4687*** -4.8572*** 41.0420*** 41.0678***
∆LBM -9.1975*** -7.9409*** 68.7891*** 80.5450***

Notes: *** significant at α = 0.01, ** significant at α = 0.05, * significant at α = 0.10. Source: Authors’ calculations

Based on both common and individual panel unit root tests, FDI and its interaction with the age dependency ratio (FDI*ADR) are stationary in levels (I(0)) at the 5 percent significance level. All other variables become stationary after first differencing (I(1)). Therefore, stationarity is achieved at the first difference for all variables, supporting the appropriateness of applying the difference GMM estimator in this study.

The parameter estimation results for the baseline model are presented in Table 5. Estimation of the baseline dynamic panel model using difference GMM indicates that FDI has a significant positive effect on labor productivity, with a p-value of 1 percent. Government debt (GDEBT) also has a positive effect on productivity, whereas broad money (BM) exhibits a negative effect. Private sector credit shows a positive but statistically insignificant effect in this model.

Table 5.
Download as
CSV
XLSX

Estimation results of the baseline model

Variable Coefficient Statistic Value
LGDPPW(-1) 0.9462 9.2700***
FDI 0.0098 6.6767***
LCRE 0.1140 1.0783
LGDEBT 0.1792 4.4102***
LBM -0.2149 -3.3168***
Validity of overidentifying restrictions test J-statistic = 5.9727 p-value = 0.3089
Arellano-Bond serial correlation test AR(1) = -27914 p-value of AR(1) = 0.0052
AR(2) = 1.6210 p-value of AR(2) = 0.1050

Note: *** significant at α = 0.01. Source: Authors’ calculations

The J-statistic is used to test the validity of overidentifying restrictions in the GMM model, which assesses the suitability of the instruments. A p-value above 0.05 (at the 5 percent significance level) indicates that the null hypothesis, which states that the instruments are valid, cannot be rejected. Therefore, the model is well-specified, and the instruments are likely valid and uncorrelated with the error term.

For the AR(1) test, a significant result is expected if the errors of the original model are uncorrelated, indicating a well-specified model, as the first-differenced residuals should exhibit negative first-order serial correlation. For the AR(2) test, a non-significant result is desirable. The failure to reject the null hypothesis in the AR(2) test indicates that there is no second-order serial correlation.

The results of the difference GMM estimation for Model 2, which includes the age dependency ratio as an explanatory variable, are presented in Table 6. FDI and government debt consistently show a significant positive effect on labor productivity, with p-values of 0.01. In contrast, broad money has a significant negative effect on productivity, with a p-value of 0.1. Private sector credit remains positive but statistically insignificant. The age dependency ratio, when included as an independent variable, does not have a significant effect on productivity. These findings suggest that the age dependency ratio does not independently influence labor productivity.

Table 6.
Download as
CSV
XLSX

Estimation results including the age dependency ratio as an independent variable

Variable Coefficient Statistic Value
LGDPPW(-1) 1.1328 17.9454***
FDI 0.0126 7.4193***
LCRE 0.1551 1.2011
LGDEBT 0.2572 2.7182***
LBM -0.5672 -1.8044*
LADR 0.0701 0.3159
Validity of overidentifying restrictions test J-statistic = 2.9011 p-value = 0.5745
Arellano-Bond serial correlation test AR(1) = -1.9682 p-value of AR(1) = 0.0490
AR(2) = 1.0594 p-value of AR(2) = 0.2894

Notes: *** significant at α = 0.01, * significant at α = 0.10. Source: Authors’ calculations

The validity of overidentifying restrictions, as indicated by an insignificant J-statistic, supports the null hypothesis that the instruments are valid. This result confirms that Model 2 is well specified. For the serial correlation tests using the Arellano–Bond procedure, the m-statistic for AR(1) is significant and negative, as expected, while the m-statistic for AR(2) is positive and insignificant. Since the AR(2) test does not reject the null hypothesis, there is no evidence of second-order autocorrelation.

Model 3, the main model in this study, includes the age dependency ratio as a moderating variable for the effects of FDI and private sector credit on labor productivity. The results, presented in Table 7, show that the positive effect of FDI on productivity remains significant. However, the reduction in the coefficient indicates that moderation by the age dependency ratio diminishes the positive effect of FDI on productivity. Private sector credit continues to show a positive but statistically insignificant effect, suggesting that moderation by the age dependency ratio also reduces the estimated impact of credit on productivity.

Table 7.
Download as
CSV
XLSX

Estimation results with the age dependency ratio as a moderating variable

Variable Coefficient Statistic Value
LGDPPW(-1) 1.1837 35.0133***
FDI*LADR 0.0034 7.5754***
LCRE*LADR 0.0462 1.3424
LGDEBT 0.2861 3.0570***
LBM -0.6400 -1.7972*
Validity of overidentifying restrictions test J-statistic = 2.6133 p-value = 0.4552
Arellano-Bond serial correlation test AR(1) = -1.8616 p-value AR(1) = 0.0627
AR(2) = 0.8972 p-value AR(2) = 0.3696

Notes: *** significant at α = 0.01, * significant at α = 0.10. Source: Authors’ calculations

Government debt maintains a significant positive effect on labor productivity, consistent with Models 1 and 2, with a p-value of 1 percent. In contrast, broad money exhibits weak evidence of a negative effect on productivity in Model 3.

Evaluation of the Model 3 specification using the J-statistic and the Arellano–Bond m-statistic indicates that the instruments are valid and that there is no second-order serial correlation. This confirms the model’s validity and supports the finding that the age dependency ratio reduces the positive effect of FDI on labor productivity. The age dependency ratio also moderates the effect of private sector credit, although this effect is not statistically significant.

Discussion

The estimation results indicate that the positive effect of FDI on labor productivity growth diminishes as the age dependency ratio increases. The rise in investment due to FDI inflows has a limited impact on labor productivity when the share of the working-age population (ages 15–64) declines relative to the non-working population (ages below 15 and above 64). These findings extend the empirical literature on the adverse effects of the age dependency burden on labor productivity, as noted by Choudhary et al. (2016) and Park and Shin (2023). This study provides new insights by showing that the age dependency ratio affects labor productivity primarily through moderation, rather than a direct effect. Labor contributes to production largely through its interaction with capital growth rather than independently.

A similar pattern is observed for private sector credit: its positive effect on productivity decreases as the age dependency ratio rises. In other words, a smaller share of the working-age population reduces the productivity-enhancing effect of private-sector credit. However, the empirical results in this study do not provide sufficient evidence to confirm a positive effect of credit on productivity. Excessive credit, as observed in Malaysia, Thailand, and Vietnam, where the credit-to-GDP ratio exceeded 100%, diminishes the positive impact of credit and renders it statistically insignificant. When the credit-to-GDP ratio surpasses a certain threshold, it can even negatively affect output growth (Lay 2020; Quoc and Duy 2025). Excessive credit growth may overheat the economy, triggering inflation that hinders sustainable growth. Misallocation of credit and inefficient financial intermediation may also explain why credit does not significantly enhance productivity. Credit directed more toward consumption than productive investment is unlikely to increase labor productivity.

The adverse effect of the age dependency ratio on productivity, operating through its moderating effect, may be explained by several factors. A higher proportion of the non-productive population (children and the elderly) requires the working-age population to allocate a larger share of their income to meet the needs of dependents, rather than saving or investing. A high dependency ratio can reduce the effective size of the labor force relative to the total population. This situation may result in labor shortages and a lower capacity for overall economic output and innovation if not managed effectively. Additionally, resources allocated to improving education and health for the working-age population may become constrained, potentially limiting the development of a high-quality workforce essential for sustainable economic growth.

This study also finds that government debt positively affects productivity. Government debt can finance productive investment, stimulate output, and enhance labor productivity. In contrast, broad money exerts a negative effect on productivity. A rising broad money-to-GDP ratio, particularly above 100 percent, can fuel inflation and hinder economic growth. High broad money-to-GDP ratios are observed in Malaysia and Thailand, while rapidly increasing ratios are seen in Vietnam and the Philippines.

Conclusion

This study investigates the effect of the age dependency ratio on labor productivity growth and examines its moderating role. The results show that the age dependency ratio significantly reduces the positive impact of FDI on productivity. Overall, the moderating effect of demographic factors, specifically the age dependency ratio, diminishes the effects of both FDI and private sector credit on labor productivity. The age dependency ratio exerts a stronger influence as a moderating variable than as an independent variable.

These findings provide further insight into how population age composition, as reflected in the age dependency ratio, affects productivity beyond the direct effects of FDI and credit. In ASEAN countries with large populations, demographic factors are critical in supporting economic growth. Increasing the productivity of the working-age population (ages 15–64) should be a top priority for addressing employment challenges. The government can enhance the quality and accessibility of education and training and increase employment by promoting investment. Policies aimed at boosting FDI inflows and private sector credit can stimulate investment, thereby raising aggregate labor productivity.

Furthermore, improving healthcare services, social security, and pension systems is essential for strengthening the productive-age population. In the long term, government policies on family planning and the prevention of early marriage should be given greater attention to manage population age structure and support sustainable economic growth.

References

  • Agustin IN, Tan M, Marheni DK (2025) Assessing the moderating role of geopolitical risk in the nexus between ESG performance and stock price crash risk: a GMM approach. Journal of Accounting and Investment 26 (3): 947–68. https://doi.org/10.18196/jai.v26i3.26923
  • Alarussi AS, Yen EZ (2023) The impact of population aging on economic growth in Asian countries. International Journal of Economics and Business Administration 11 (1): 33–53. https://doi.org/10.35808/ijeba/796
  • Arellano M, Bond S (1991) Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies 58 (2): 277–97. https://doi.org/10.2307/2297968
  • Arintoko, Badriah LS, Sambodo H, Kadarwati N, Rahajuni D (2024) The asymmetric effect of interbank rate and financial performance ratios on commercial bank profitability: Empirical evidence of macro-level data from Indonesia. Montenegrin Journal of Economics 20 (3): 39–52. https://doi.org/10.14254/1800-5845/2024.20-3.3
  • Asada H (2020) Effects of foreign direct investment and trade on labor productivity growth in Vietnam. Journal of Risk and Financial Management 13 (9), 204. https://doi.org/10.3390/jrfm13090204
  • Badriah LS, Arintoko (2024) The dynamics model of cross-regional to enhance economic growth: New empirical evidence from Indonesia. Economics – Innovative and Economics Research Journal 12 (3): 379–95. https://doi.org/10.2478/eoik-2024-0046
  • Bidisha SH, Abdullah SM, Siddiqua S, Islam MM (2020) How does dependency ratio affect economic growth in the long run? Evidence from selected Asian countries. Journal of Developing Areas 54 (2): 47–60. https://doi.org/10.1353/jda.2020.0015
  • Bogdan V, Popa D-N, Matica D-E, Belenesi M (2025) Unveiling governance mechanisms: how board characteristics disclosure moderates the gender diversity and corporate performance nexus in Romania. Systems 13 (6): 420. https://doi.org/10.3390/systems13060420
  • Brida JG, Alvarez E, Cayssials G, Mednik M (2024) How does population growth affect economic growth and vice versa? An empirical analysis. Review of Economics and Political Science 9 (3): 265–97. https://doi.org/10.1108/REPS-11-2022-0093
  • Dandi T, Yunisvita, Atiyatna DP (2024) Determinants of labor productivity in ASEAN-6 countries: Using Dynamic Panel SYS-GMM. Eurasian Journal of Economics and Finance 12 (2): 83–92. https://doi.org/10.15604/ejef.2024.12.02.003
  • Fu L, Wang Y, He L (2020) Age composition change and inter-provincial labor productivity: a study from the perspective of population dividend and population urbanization. Journal of Applied Economics 23 (1): 183–98. https://doi.org/10.1080/15140326.2020.1723885
  • Ginting AL, Widyawati RF (2022) Analysis of commercial bank credit distribution on economic growth and employment absorption in Indonesia. OPTIMUM: Jurnal Ekonomi dan Pembangunan 12 (1): 33–40. https://doi.org/10.12928/optimum.v12i1.4493
  • Gutierrez E, Jaume D, Tobal M (2023) Do credit supply shocks affect employment in middle-income countries? American Economic Journal: Economic Policy 15 (4): 1–36. https://doi.org/10.1257/pol.20210354
  • Huang W-H, Lin Y-J, Lee H-F (2019) Impact of population and workforce aging on economic growth: Case study of Taiwan. Sustainability 11 (22): 6301. https://doi.org/10.3390/su11226301
  • Maestas N, Mullen KJ, Powell D (2023) The effect of population aging on economic growth, the labor force, and productivity. American Economic Journal: Macroeconomics 15 (2): 306–32. https://doi.org/10.1257/mac.20190196
  • Manaresi F, Pierri N (2024) The asymmetric effect of credit supply on firm-level productivity growth. Journal of Money, Credit and Banking 56 (4): 677–704. https://doi.org/10.1111/jmcb.13092
  • Mankiw GN (2016) Macroeconomics. 9th edition. Worth Publishers, New York, 641 pp.
  • Quoc BNK, Duy TP (2025) How do credit and financial cycles jointly affect economic output? A state-dependent analysis for Vietnam. Macroeconomics and Finance in Emerging Market Economies, 1–19. https://doi.org/10.1080/17520843.2025.2516303
  • Rebić M, Paunović S, Popović B (2022) Measuring concentration and efficiency in Bosnia and Herzegovina banking sector using dynamic panel model. South East European Journal of Economics and Business 17 (1): 14–29. https://doi.org/10.2478/jeb-2022-0002
  • Santosa PW, Rahayu SI, Simon ZZ, Santoso PW (2023) Moderating role of audit quality and firm size on pretax profit margin and related party transaction: evidence form Indonesia. Business: Theory and Practice 24 (1): 291–300. https://doi.org/10.3846/btp.2023.17946
  • Saucedo E, Oyuna Jr T, Zamora H (2020) The effect of FDI on low and high‑skilled employment and wages in Mexico: a study for the manufacture and service sectors. Journal for Labour Market Research 54: 9. https://doi.org/10.1186/s12651-020-00273-x
  • Tőkés L (2019) The effect of foreign direct investment on firm labor productivity: Does the country of origin of the FDI matter? Society and Economy 41 (2): 227–43. https://doi.org/10.1556/204.2019.002
  • Turan T, Yanikkaya H, Özer HA (2021) Economic growth effects of public and private investment: Evidence from dynamic panel estimation for developed and developing countries. Prague Economic Papers 30 (5): 613–31. https://doi.org/10.18267/j.pep.781
  • Ullah W, Zubir ASM, Ariff AM (2024) Exploring the moderating effect of regulatory quality on the relationship between financial development and economic growth/economic volatility for developed and developing countries. Borsa Istanbul Review 24 (5): 934–44. https://doi.org/10.1016/j.bir.2024.04.015
  • Vinh NT (2019) The impact of foreign direct investment, human capital on labour productivity in Vietnam. International Journal of Economics and Finance 11 (5): 97–102. https://doi.org/10.5539/ijef.v11n5p97
  • Yunus NM, Abdullah N (2022) The effect of foreign direct investment on labour productivity: Evidence from five investor countries in the Malaysian manufacturing industries. Malaysian Management Journal 26: 55–86. https://doi.org/10.32890/mmj2022.26.3

Information about the authors

Arintoko, Doctor in economics, Professor, lecturer at Department of Economics and Development Studies, Faculty of Economics and Business, Universitas Jenderal Soedirman, Purwokerto, Indonesia. E-mail: arintoko@unsoed.ac.id

Herman Sambodo, Master of Agriculture, Associate Professor, lecturer at Department of Economics and Development Studies, Faculty of Economics and Business, Universitas Jenderal Soedirman, Purwokerto, Indonesia. E-mail: herman.sambodo@unsoed.ac.id

Priyono Rakhmat, Master of Economics, Associate Professor, lecturer at Department of Economics and Development Studies, Faculty of Economics and Business, Universitas Jenderal Soedirman, Purwokerto, Indonesia. E-mail: rakhmat.priyono@unsoed.ac.id

login to comment