Does the Aging of the Chinese Population Have an Impact on Outward Foreign Direct Investment?

Abstract

Given the progression of population aging in China, does the diminishing demographic dividend boost the promotion of investment abroad in the form of outward foreign direct investment (OFDI)? This empirical study focused on the influence of population aging on outward foreign direct investment (OFDI) decisions and its underlying mechanisms. The research found that population aging has a significant positive effect on the level of OFDI. This effect is particularly pronounced in the eastern and central regions, while not statistically significant in the western region. Furthermore, population aging has a significant impact on the factor cost effect and the technological progress effect. The former is characterized by increased labor costs, while the latter is associated with technological advancements. The study confirmed that population aging positively influences OFDI changes through these two mechanisms. The empirical results hold statistically significant after multiple robustness checks. This study holds a significant reference value in advancing the facilitation of high-level opening-up polices and policy coordination to effectively address the challenges posed by population aging.

1. Introduction

Since the 21st century, China has been confronted with the complex and relentless issue of an increasingly aging population. The data acquired through the National Bureau of Statistics (NBS) of China’s Population Sample Survey in 2022 revealed a significant rise in the proportion of individuals aged 65 years and above, reaching a staggering 14.9%. This figure represents a notable increase of 6 percentage points compared to data obtained from the Sixth National Population Census in 2010. Further analysis indicates that the old-age dependency ratio (ODR), which measures the ratio of individuals aged 65 years and above to the working population aged 15–64 years, has also experienced a substantial upsurge, from 11.9% in 2010 to 21.8% in 2022. Of particular concern is the observation that, in 2022, China’s total fertility rate displayed a negative value for the first time. The escalating phenomenon mentioned above highlights the urgency to confront the issue of “becoming old before becoming rich” in developing nations experiencing rapid population aging (Bloom et al., 2011) [1]. Scholars such as Banister et al. (2012) [2] have emphasized the urgency of resolving China’s rapid aging predicament. Recognizing the gravity of the situation, the report of the 20th Party Congress has underscored the imperative of devising and executing a comprehensive national strategy to proactively manage the challenges presented by population aging.

It is evident that the reduction of the “demographic dividend” resulting from population aging will have profound consequences for economic and social progress. One aspect that warrants close attention is the influence on investment, particularly outward foreign direct investment (OFDI), and this impact can be categorized into three aspects. (1) Population aging leads to a reduction in labor supply, causing an imbalance between labor market supply and demand and a rise in labor costs. This implies higher labor costs for enterprises, which can affect their investment plans. In such circumstances, enterprises are more inclined to seek new development opportunities through the OFDI. By establishing production bases in other countries or seeking inexpensive labor resources, enterprises can overcome the obstacles of rising labor costs in their home country and reduce production expenses. (2) Population aging results in changes to the talent structure of the labor market. With an increasing proportion of elderly people, the skill structure and labor market demand will also undergo transformations. This may necessitate enterprises to adjust their talent allocation and skills training to adapt to the new demographic structure. However, with a shortage of labor supply in their home country, enterprises may be more motivated to acquire the required skills and talent resources through investments on OFDI. (3) Population aging also exerts significant influence on technological innovation and development, which, in turn, affects investment decisions. As a means to address the scarcity of labor, enterprises may increase their investment in research and development (R&D), driving technological progress and innovation. These technological advancements can enhance productivity and reduce the demand for labor. Thus, population aging, through the impetus for technological progress, also impacts the investment direction and decisions of enterprises. In summary, population aging has significant implications for investment, particularly OFDI. The rise in labor costs, changes in talent structure, and shifts in demand due to technological advancements all contribute to enterprises being more inclined to choose OFDI as a response to the challenges brought about by population aging. Understanding and addressing these impacts are crucial for promoting sustainable economic development.

In this part, we present a brief overview of China’s OFDI and explore its relationship with population aging.

Firstly, both outward foreign direct investment (OFDI) by domestic enterprises and foreign direct investment (FDI) attracted to China have displayed a concurrent growth pattern in recent years. Notably, China’s open economy is transitioning from a stage of primarily “bringing in” investment to one of increasing prominence in “going out” investment. In 2014, China’s OFDI reached a historic high of USD 123.12 billion, while FDI utilization amounted to USD 119.56 billion. Moreover, for the first time, the scale of China’s outward FDI surpassed the magnitude of FDI attracted to the country, positioning China as a net exporter of FDI. As of 2022, China’s OFDI has consistently ranked among the top three globally for 11 successive years. Despite encountering external challenges, China’s OFDI has achieved stable and progressive growth. Notably, industry-wide OFDI reached CNY 985.37 billion, representing a 5.2% increase (equivalent to USD 146.5 billion, a 0.9% increase). Meanwhile, non-financial direct investment reached CNY 785.94 billion, indicating a 7.2% rise (equivalent to USD 116.85 billion, a 2.8% increase). Within this, local enterprises accounted for 80.4% of the total, with outward investment totaling US$93.92 billion, reflecting a 13.1% increase from the previous year. The eastern region experienced a 10.3% growth in outward investment, encompassing 81.6% of local investment. Leading the way in local outward investment were Guangdong, Zhejiang, and Shanghai, which ranked among the top three regions in terms of outbound investment.

Secondly, the relationship between population aging and changes in OFDI remains uncertain, and it can be discussed from two different perspectives. (1) Examining the capital flow perspective: it is observed that an increase in the proportion of older individuals in the population, who have accumulated wealth over their lifetimes and serve as “capital providers”, may lead to an overall augmentation in societal capital supply. Empirical studies conducted on OECD countries have supported this long-term positive relationship between the aging population and FDI inflows (Mitra and Abedin 2020) [3]. Conversely, an aging demographic structure imposes an increased burden of elderly care on society, potentially resulting in escalated social welfare expenditures and a potential decline in capital outflows. (2) Considering the perspective of labor supply and the relative cost of factors, a deepening aging process tends to bring about a reduction in the size of the labor force population, lower labor force participation rates (LFPRs), an aging labor force demographic structure, rising labor costs, and a relative decline in the cost-of-capital factors. These factors collectively contribute to an increased demand for outward investment. Consequently, the relationship between population aging and outward investment encompasses inherent uncertainty when examined from these two perspectives. In summary, based on the aforementioned factors, it is evident that the relationship between population aging and outward investment is marked by ambiguity.

The following context investigates the impact of population aging on the level of OFDI. It focuses initially on the overall direction of its effects and subsequently examines two specific channels: the labor cost effect and the technological progress effect. Finally, it provides an integrated logical summary of the research being conducted.

Firstly, we illustrate the labor cost effect. Population aging directly influences the demographic structure by diminishing the size of the working population and elevating the old-age dependency ratio (ODR), consequently leading to an increase in labor costs. The existing literature highlights that the labor cost effect brought about by population aging reduces the rate of return on capital, subsequently provoking outflows of international direct investment (Davies and Reed III 2006) [4]. As China transitions into an aging society, the scarcity of labor and the surge in labor costs render the traditional processing trade model unsustainable. Correspondingly, with rising labor costs and a decline in capital costs, firms are more inclined to opt for capital investments overseas (Cai and Stoyanov 2016) [5]. Consequently, population aging in the home country motivates firms to engage in outward FDI, seeking alternative drivers of growth by transforming factor costs.

Secondly, we explore the technological progress effect. Population aging also contributes to the promotion of the R&D investment and serves as a catalyst for technological progress within society as a whole. Population aging prompts a decline in the size of the labor force, which, in turn, fosters the innovation of labor-substituting technologies (Acemoglu 2010) [6], as illustrated by the recent advancements in industrial automation technologies. Acemoglu and Restrepo (2017) [7] conducted an empirical study using cross-country data, revealing that countries experiencing faster population aging also witness a more rapid development of automation technologies. This implies that population aging acts as a driving force for labor-substituting innovations. Hence, the technological-progress effect triggered by population aging may also serve as a significant channel for promoting an increase in OFDI.

Thirdly, the content structure and logical framework of the study are given. The present study analytically examined the direction and mechanisms through which population aging affects outward foreign direct investment (OFDI) at a theoretical level. Subsequently, an empirical analysis was conducted to assess the impact of population aging on changes in OFDI, its corresponding mechanisms, and the marginal effects. Furthermore, utilizing data from the China Statistical Yearbooks and Outward Foreign Direct Investment Statistical Bulletins spanning from 2005 to 2020, empirical tests were conducted using a difference equation to determine the magnitude of population aging’s impact on OFDI variations, its underlying mechanisms, and the marginal effects. Moreover, this study delved into the regional heterogeneity of these effects across provinces in the eastern, central, and western regions, as well as the heterogeneity between coastal and inland areas. Given the potential endogeneity issues in the empirical model, this study employed the dynamic panel System Generalized Method of Moments (GMM) model and two-stage least squares model (2SLS) with instrumental variables to test the empirical results. Robustness tests were conducted by replacing the independent variables, eliminating the interference of free trade zone (FZT) policies, and excluding the influence of Internet development.

In comparison to existing studies, this paper contributes to the literature by focusing on three key aspects. (1) In terms of research perspectives, while previous studies have predominantly examined the effects of population aging on economic growth (Hashimoto and Tabata 2016) [8] or export trade growth (Cai and Stoyanov 2016) [5], this study diverges by investigating the regional difference due to the vast territory of China and significant economic disparities and variations in population structure across different regions. To be more specific, this study takes a regional perspective from China, considering the regional heterogeneity of population aging and the regional heterogeneity of its economic effects. This adds an important supplement to the existing literature’s research perspectives. Considering the vastness of China, regional economic disparities, and demographic differences, this approach offers valuable insights into the dynamics of OFDI induced by population aging. (2) In the realm of theoretical analysis, this study contributes to the existing literature on population aging and technological progress (Acemoglu and Restrepo 2017 and 2021) [7,9] by extending the discussion to include the factor cost effect and the impact of technological advancement. By examining the channel effect through which population aging influences OFDI, this study enriches the understanding of the underlying causes of outward FDI and provides a valuable contribution within the theoretical framework of population and investment. (3) In terms of empirical content, this study utilized data from the National Demographic Yearbook and OFDI Bulletin, covering the period from 2005 to 2020. Employing the panel data, this paper conducts a comprehensive analysis that explores the impact of population aging on OFDI, considering regional heterogeneity and the coastal–inland dimensions. This empirical analysis provides meaningful insights and empirical support for governmental decision-making processes. In addition, this paper considers the trade tensions between China and the United States starting in 2018. It employs a time-varying variable for in-depth analysis, exploring the impact of a worsening external investment climate on the relationship between population aging and OFDI. In summary, this paper contributes to the existing literature by adopting a unique research perspective, enhancing the theoretical analysis by exploring different mechanisms, and providing robust empirical evidence that considers regional heterogeneity and the coastal–inland dimensions.

The structure of this paper is as follows: Section 2 introduces theories of aging population and its economic impacts; Section 3 puts forward three hypotheses according to the illustration of mechanisms of aging population on the changes of OFDI; Section 4, Section 5 and Section 6 present the empirical results and analysis and develop further tests based on the benchmark regressions, including endogeneity tests, robustness tests, heterogeneity tests, and mechanism tests; and in Section 7, a further study considering China–US conflicts since 2018 is given. Finally, we summarize conclusions and recommendations in Section 8.

2. Literature Review

This section begins by introducing the theory of population aging and the rationale for choosing China as the research focus. It then summarizes the relevant economic effects of population aging, with a primary focus on its impact on foreign direct investment (FDI). Building on this theory, this section further summarizes the theoretical impact of population aging on outward foreign direct investment (OFDI) and supplements it with the existing literature on the determinants of OFDI. Finally, this section highlights the limitations of the existing literature and explains how this study addresses these gaps.

We first discuss the relevant theories of population aging and provide the contextual background of China. The Life Cycle Hypothesis (LCH), which was originally proposed by Modigliani and Brumberg (2005) [10], has long served as the conventional framework for economists studying the economic implications of population aging. According to this hypothesis, rational consumers make optimal savings and consumption decisions based on their income levels throughout their lifetime. The life cycle can be divided into three stages: youth, middle age, and old age. During the youth and old-age stages, individuals are net consumers, with consumption exceeding income. In the middle-age stage, individuals are producers, with income exceeding consumption. During this stage, individuals earn income through work, part of which is allocated for consumption and repayment of debts from the youth stage, while the remainder is saved for retirement. However, the economic development trajectory of China has witnessed a noteworthy divergence from this traditional perspective, characterized by rapid aging and higher levels of savings rates (Modigliani and Cao 2004) [11]. China’s aging issue is of particular concern due to three notable reasons. China’s aging issue merits attention for three significant reasons. Firstly, the pace of population aging is rapid and currently in an accelerating phase. Secondly, due to China’s large population base, the size of the elderly population is immense. Thirdly, the characteristic of “aging before becoming prosperous” is noteworthy, making the study of China’s aging and its economic impacts particularly relevant for developing countries and making the study of China’s aging and its economic impacts particularly relevant for developing countries (Bloom et al., 2011) [1].

Relevant research conducted on developed countries has indicated that population aging may have adverse consequences on the LFPR and labor productivity and lead to excessive savings, which can potentially hinder a country’s economic growth. However, it is important to note that there is no universal empirical evidence suggesting a negative impact of population aging on GDP growth (Acemoglu and Restrepo 2021) [9]. In fact, the findings regarding the relationship between population aging and economic growth are not consistent across different research contexts, exhibiting heterogeneity between developed and developing countries (Pham and Vo 2021) [12]. Given these disparities, it is both important and necessary to conduct studies on population aging and its economic effects within the specific context of China. In the context of China, the impact of population aging on the labor market and the savings rate may differ from that of other countries due to the unique characteristics of the Chinese economy. China’s distinctive economic development model, government policy interventions, and demographic dynamics can create both opportunities and challenges in the face of population aging. Consequently, conducting a comprehensive examination of the economic effects of aging in China is crucial for enhancing our understanding of this issue and devising appropriate responses. Moreover, such a study can provide valuable insights and references for other developing countries grappling with similar challenges.

Then, we demonstrate the economic effect of population aging. From a macro-level perspective, the existing research primarily concentrates on understanding the effects of population aging on economic growth, consumption and savings patterns, and human capital (Maestas et al., 2016; Li et al., 2007; VOGEL et al., 2015) [13,14,15]. On the other hand, from a micro-level standpoint, the majority of research examines the influence of household demographics on household consumption, savings, and investment, typically analyzing households as individual units (Huo et al., 2021) [16]. Furthermore, some studies have explored how demographic changes impact export trade. For instance, Cai and Stoyanov (2016) [5] argue that differences in demographics between countries can serve as a new source of comparative advantage in international trade. However, there is a paucity of studies examining the specific impact of population aging on OFDI. Given the worldwide phenomenon of population aging, scholars have increasingly turned their attention to the potential implications of aging on economic development. This study builds upon the theoretical research that examines the relationship between aging and international capital flows.

The existing literature predominantly investigates the impact of population aging on FDI (Burlea-Schiopoiu et al., 2021) [17], which is commonly discussed through three main channels, as outlined below.

Firstly, population ageing affects FDI by influencing savings rates. Population aging influences foreign direct investment (FDI) by impacting savings rates, thus affecting capital accumulation. This, in turn, influences the supply and price of capital, prompting cross-border capital flows in the form of FDI, stock purchases, cross-border loans, and inter-regional investments. Empirical evidence provided by Melanie (2003) [18] indicates that demographic changes play a substantial role in international capital flows, leading to capital transfers between countries. The differences in age composition among regions prompt capital transfers between countries. Countries with limited restrictions on capital inflows and outflows may experience the influence of population aging on capital flows, as reflected in their current account. If capital inflows surpass outflows, a deficit appears in the current account. Conversely, if capital outflows exceed inflows, a surplus is observed. Demographic changes create imbalances between savings and investment within a country, thereby affecting the rate of return on capital across various regions. Higgins and Williamson (1997) [19] find that a higher dependency ratio leads to reduced private and public savings, encouraging an inflow of international capital. Building on this, Herbertsson and Zoega (1999) [20] argue that a higher ratio of elderly dependents specifically decreases savings, thereby triggering an influx of foreign direct investment (FDI). Additionally, Kim and Lee’s study on the G7 countries (2008) [21] reveals that an increased dependency ratio contributes to a current account deficit by reducing the savings rate, subsequently attracting higher levels of capital inflow. These findings suggest a significant link between population dependency, savings dynamics, and capital flows at both national and international levels.

Secondly, population ageing affects FDI by influencing labor costs; population aging has notable effects on the labor force size and average productivity, leading to an increase in labor costs. Regional variations in the extent of aging consequently result in differences in labor costs across regions. This disparity, in turn, prompts firms to redirect low-end labor-intensive production to other countries while maintaining technology- and capital-intensive products domestically (Lipsey and Sjoholm 2004) [22]. The impact of population aging on foreign direct investment (FDI) can be analyzed through the lens of labor costs. According to the Dunning OLI theory (Dunning 2001) [23], the relationship between an aging population and rising labor costs can be understood through the “Location” factor. As the population ages, the labor force tends to shrink, which can lead to a scarcity of skilled labor and an increase in labor costs in certain countries or regions. Population aging reduces the number of individuals actively engaged in production, thus straining the labor supply and discouraging foreign investment in the country. The increase in the level of aging within a country acts as a catalyst for diminished inflows.

Thirdly, population ageing affects FDI by influencing labor force participation rates (LFPR); the effects of population aging are typically long-term and exhibit irreversible characteristics within a short time frame. Aging leads to an increase in the average age of the labor force, accompanied by a decline in the labor force participation rate, consequently reducing the overall labor supply. Some researchers argue that weak labor demand is the main reason for a declining LFPR, but empirical evidence shows that a change in the demographical structure, such as the aging of the population, is the most significant contributor to a decreasing LFPR (Aaronson et al., 2014) [24]. Currently, our country is undergoing a significant transitional period in terms of demographic dynamics, and the size of the labor force remains essential for maintaining steady economic growth. However, as population aging deepens through the years, the tightening labor demand and diminishing labor force will act as restricting factors for China’s economic development. The impact of population aging on the rate of return on capital can be understood through its influence on the labor supply. An increase in the working-age population will result in a relatively higher labor supply compared to the supply of capital, thus elevating the rate of return on capital in regions with sufficient labor supply. Conversely, regions experiencing a shortage of labor relative to the supply of capital will witness a decline in the rate of return on capital, prompting capital outflows from these areas.

According to the aforementioned information, the flow of international direct investment (IDI) is influenced by the host countries’ labor factor supply capacity, which is closely linked to the age structure of their populations, including China. Population aging primarily affects IDI flows through two significant factors: the constrained supply of labor and the age composition of the labor force (McDonald and Kippen 2001) [25]. Societies with younger demographics tend to be more productive, but population aging impairs the demographic dividend and slows down the potential growth rate of the economy, leading to labor supply limitations in China. Consequently, these dynamics inevitably impact the flow of IDI. However, the underlying mechanisms by which population aging affects IDI, as well as the global empirical evidence of its impact, require further investigation. The study of the relationship between population aging and IDI sheds light on changes in IDI patterns from the perspective of population age structure. Davies and Gresik (2003) [26], in their study on the influence of demographic structure on FDI, proposed that changes in demographic structure are a cause of factor price changes in a country, as population structure can affect a nation’s savings, thus, in turn, impacting the prices of capital factors. Additionally, changes in the demographic structure can affect labor supply and, consequently, labor costs, leading to the cross-regional mobility of capital. Davies and Reed III (2006) [4] posited a hypothesis that there are only two countries: one that provides OFDI (the home country) and one that receives OFDI (the host country). These scholars constructed a transmission model to explain the reasons for FDI flows resulting from changes in age structure, considering both capital costs and labor costs. From the perspective of capital costs, the intensified aging in the home country of OFDI leads to a decrease in the proportion of domestic capital used for investment, resulting in a decrease in capital supply and an increase in capital costs. In the context of FDI inflows, capital flows from regions with lower rates of return to those with higher rates of return. Therefore, the result of population aging is capital outflow. From the perspective of labor costs, the increased level of aging leads to a decrease in labor supply and an increase in labor costs. This also leads to a decline in the capital return rate in the home country, resulting in a preference for capital flows to regions with higher returns. Moreover, Melanie (2003) [18] conducted a study using data on population and capital flows from different countries to provide empirical evidence that capital tends to flow from countries with aging populations to countries with younger population structures in order to obtain higher returns.

Numerous studies have examined the factors influencing outward foreign direct investment (OFDI). The external environment of firms, such as exchange rate shocks and the level of financial development in their home countries, have been found to impact firms’ OFDI decisions (di Giovanni 2005; Russ 2007) [27,28]. From an internal perspective, Helpman et al. (2004) [29] argue that firms must weigh the variable costs of producing goods abroad against the fixed costs of investing domestically, implying that changes in relative costs are crucial for firms’ OFDI decisions. Additionally, productivity serves as a determinant of firms’ OFDI choices, as high-productivity firms are more likely to engage in foreign investment (Yeaple 2009) [30]. Furthermore, prior research has categorized the specific driving forces behind firms’ engagement in OFDI. Chinese firms’ motivations for OFDI can be classified into five main categories: resource-seeking, policy-driven, market-seeking, innovation-seeking, and efficiency-seeking motives. The resource-seeking motive aims to secure scarce domestic resources by directly investing in countries that possess abundant resources, thereby providing a more reliable resource base for domestic industries. The policy-driven motive refers to government support through planning, incentives, and favorable policies, which encourage domestic industries to invest in host countries. The market-seeking motive seeks to overcome trade barriers and relocate excess domestic capacity overseas. The innovation-seeking motive focuses on acquiring advanced technologies or production methods through OFDI and benefiting from technology spillovers. The efficiency-seeking motive aims to reduce costs by optimizing resource allocation, such as utilizing low-cost labor in host countries. As population aging is an external influencing factor, it exerts varying effects on different motivations for OFDI. This study concentrated on the direction and channels of population aging’s impact, offering an innovative examination of how population aging serves as a determinant of OFDI from this perspective. This study’s novelty lies in investigating the aging population as a driving force behind OFDI.

In summary, there are three main shortcomings in the existing literature on the relationship between population aging and international investment. Firstly, the focus is predominantly on FDI, mostly from the perspective of population aging in developed countries, with limited research on the impact of population aging in developing countries on OFDI. Additionally, existing empirical studies primarily concentrate on the macroeconomic growth, trade implications, firm-level effects, development patterns, and influence on consumption and savings behaviors of households. This study fills a crucial gap by taking a regional perspective from China, offering an important addition to existing research perspectives. Lastly, while most empirical studies analyze static panel data, this study employed difference equations to examine the relationship between the deepening process of population aging and the extent of OFDI changes. By utilizing dynamic panel data, this study captures the changing trends between the two variables more effectively, providing a more reasonable and persuasive analysis.

3. Theoretical Mechanisms and Research Hypotheses

3.1. The Promotion of Aging Population on the OFDI

Although there is limited research on the impact of population aging on outward foreign direct investment (OFDI) changes, the analysis framework of the impact of aging on inward foreign direct investment (FDI) can be applied to examine its effects. Existing research demonstrates that population aging affects international capital flows through factors such as savings rates, capital–labor ratios, and returns on capital. However, there is no consistent consensus on whether population aging positively stimulates or negatively inhibits international capital flows. Some studies propose that population aging promotes international capital inflows. Neoclassical growth dynamic equilibrium models indicate that population aging in OECD countries fosters international capital inflows (Domeij and Flodén 2006) [31]. Other studies suggest that population aging in the home country leads to international capital outflows. Research findings show that population aging reduces the rate of return on capital due to increased capital–labor ratios, resulting in greater capital outflows from countries with aging populations and increased capital inflows into countries with younger populations (Krueger and Ludwig 2007; Ito and Tabata 2009) [32,33]. Capital flows tend to move from areas with rapidly aging populations to areas with relatively slower aging populations. Population aging primarily affects international capital flows through the capital–labor ratio and the current account balance, with higher levels of population aging correlating with increased capital outflows.

Therefore, population structure has been proven to be a significant factor influencing economic factors such as economic growth and international capital flows. This study took population structure as the core explanatory variable to investigate its impact on changes in OFDI. This research enriches the study of determinants of OFDI and provides empirical evidence regarding the influencing factors of international capital flows. Based on this, the following proposition is presented:

Hypothesis 1 (H1).

The deepening of population aging exerts a promotional effect on the level of outward foreign direct investment (OFDI).

3.2. Factor Cost Effect

Demographic change, particularly population aging, represents a direct and discernible phenomenon that triggers significant consequences. One notable effect is the decline in the working-age population, which disrupts the balance between labor supply and demand within the labor market. An extensive amount of the literature has scrutinized the impact of demographic change on factor prices, with scholars such as Poterba (2001) [34]; Krueger and Ludwig (2007) [32]; and Ludwig, Schelkle, and Vogel (2012) [35] contributing to this understanding. Specifically, when labor supply fails to meet demand, labor costs experience an increase. Simultaneously, population aging prompts a rise in the capital–labor ratio within the factor market, resulting in a relative decline in the price of capital factors. This decline facilitates firms’ ability to augment their capital inputs, including the pursuit of outward foreign direct investment as a means of exploring novel avenues for development. Consequently, the implications of population aging extend beyond economic development patterns and have a substantial impact on corporate decision-making processes.

Population aging presents a dual effect on the economic landscape. On one hand, it generates labor shortages, an increase in the old-age dependency ratio, and higher labor costs, which pose challenges to sustaining an economic development model dependent on the “demographic dividend”. On the other hand, population aging results in rising labor costs and a relative decrease in the cost of capital inputs, thus facilitating firms’ decision to engage in outward direct investment. Notably, the study conducted by Fan et al. (2018) [36] provides pertinent insights in this context. Their investigation focuses on the minimum wage system as a perspective for analysis and reveals that an increase in the minimum wage raises labor costs at firms’ operating locations, thereby increasing the likelihood of firms engaging in outward FDI. Additionally, the study suggests that approximately 32.3 percent of China’s growth in OFDI between 2001 and 2012 can be attributed to the rise in minimum wages.

In conclusion, the demographic changes associated with population aging have had notable ramifications on the dynamics of labor market supply and demand, as well as factor prices. Specifically, the increase in labor costs and the concurrent decline in the prices of capital factors have significant implications for firms’ decision-making processes, rendering them more inclined toward opting for OFDI. The rise in minimum wages further amplifies this trend. Consequently, the following hypothesis is posited:

Hypothesis 2 (H2).

Population aging induces a rise in labor factor costs and a relative decrease in capital factor costs, thus facilitating enterprises’ inclination toward choosing OFDI, as evidenced by the factor cost effect.

3.3. Technological Progress Perspective

In addition to its impact on factor costs, population aging can also exert a substantial influence on technological progress, specifically by inducing labor-substituting technological advancements. Previous studies have explored this relationship, with findings indicating that demographic changes elevate levels of innovation within the pharmaceutical industry, as observed by Acemoglu and Linn (2004) [37] and echoed in the research conducted by Costinot et al. (2019) [38]. Recent investigations have expanded beyond the pharmaceutical industry, revealing that demographic changes attributed to aging can stimulate broader societal technological advancements. Empirical evidence supporting this notion can be found in the early research of Acemoglu (2010) [6], where labor scarcity is demonstrated to impact technological progress. His study classifies technological progress into two categories: labor-saving and labor-complementary. It reveals that labor scarcity fosters labor-saving technological progress, whereby technology assumes a highly labor-saving nature within a rational environment.

According to subsequent research by Acemoglu and Restrepo (2017 and 2021) [7,9], population aging stimulates technological innovations in industrial automation, thus disproving the notion that aging leads to economic stagnation. Empirical evidence from a study by Graetz and Michaels (2018) [39] emphasizes the significant contribution of labor-substituting technologies in this regard. Consequently, within a certain context, aging prompts enterprises to augment their investment in research and development (R&D), fostering technological progress and facilitating capital accumulation. This, in turn, promotes an upsurge in outward foreign direct investment (OFDI) by firms, driven by the “push mechanism”. Additionally, the “technology spillover” effect of OFDI further advances technological progress. Consequently, the following hypothesis is proposed:

Hypothesis 3 (H3).

Population aging exerts a positive influence on capital accumulation through increased R&D investment and the generation of technological progress. This, in turn, facilitates an upswing in OFDI by firms, as evidenced by the technological progress effect.

4. Empirical Studies

4.1. Empirical Model

In order to assess the impact of population ageing on provincial OFDI, this study set up the following econometric model:

Δofdi_p = α + βΔAging_p + γX + τ_p + ε_p

(1)

Due to the lagged effects of demographic changes on OFDI, a certain reaction time is typically required. Following the methodology employed by Acemoglu and Restrepo (2021) [9] in studying the relationship between demographic structure and trade structures, this study employed a panel data approach to examine the aforementioned impact, using a long-difference model to assess the effects of demographic changes on OFDI. The explanatory variable Δ represents the change in the difference between the t period and t − 1 period after taking the logarithm of the OFDI of province p from 2005 to 2020. The core explanatory variable Δ represents the change in the proportion of elderly people in the population (population over 65 years old) to the total population from 2005 to 2020 in province p. Control variables are denoted collectively as X, and the regression results remain robust after controlling for per capita GDP, industrial structure, foreign trade dependence, share of fiscal expenditure, road network density, deposit and loan balances as a percentage of GDP, and other policy factors in the same period. The study also recognizes the challenges in accounting for all factors influencing provincial-level changes in OFDI levels. To mitigate these limitations, province fixed effects are introduced to account for unobservable provincial-level factors and to minimize the influence of provincial-level policy and macro-factors on the empirical findings. Additionally, year effects are controlled for in the analysis.

4.2. Data Resources

The data utilized in this study were derived from two primary sources. The first source encompasses demographic and economic indicators for Chinese provinces, autonomous regions, and municipalities directly under the central government, covering the period from 2005 to 2020. This data were obtained from the China Statistical Yearbook and the website of the National Bureau of Statistics (NBS). The second source consists of annual statistics from an annual report of outward foreign direct investment (OFDI) of Chinese provinces. These statistics are sourced from the China External Economy Database of the EPS Data Platform, which, in turn, derives its data from the China OFDI Statistical Bulletin.

The study utilized a panel data approach, focusing on China’s provincial-level data, spanning the years 2005 to 2020, to examine the influence of demographic factors on China’s outward foreign direct investment (OFDI). The Statistical Bulletin of China’s Outward FDI first began publishing flow data in 2003 and stock data in 2004. However, substantial outward FDI activities by Chinese enterprises commenced in 2005. Hence, the study selected data from the period ranging from 2005 to 2020 for analysis. It is worth noting that, due to inherent reasons, flow data may exhibit “negative values”, thus making them unsuitable for logarithm transformation. Therefore, the focus of this paper is placed on utilizing OFDI stock data, which provide a more reasonable basis for analysis.

4.3. Core Indicator Construction

The population ageing indicator is the core explanatory variable, which is constructed from national demographic data by province:

$${\mathsf{\Delta }\mathrm{A}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{p}}=\frac{{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}}^{{65}^{+}}}{{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}}^{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}}-\frac{{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}-1}^{{65}^{+}}}{{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}-1}^{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}}$$

(2)

where ${\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}}^{{65}^{+}}$ represents the number of aged population in province p (over 65 years old), and ${\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{p},\mathrm{t}}^{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}$ represents the population overview of province p. The ratio of the two is the proportion of aged people in the population of province p. The difference between the two periods represents the extent of change in the aging of the region, which is the primary explanatory variable in this study. Furthermore, an additional measure of aging is utilized, namely the ratio of the elderly population to the working-age population in province p. This ratio, also known as the old-age dependency ratio (ODR), measures the burden imposed on society in supporting the elderly population. Specifically, it calculates the ratio of the number of individuals aged 65 or older to the number of individuals aged 15–64 in province p.

The main explanatory variable in this study is the outward foreign direct investment (OFDI) indicator at the provincial level. The rate of change of the OFDI for each province in the country is constructed as the OFDI indicator, serving as an important component in the analysis:

$${\mathsf{\Delta }\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{p}}={\mathrm{l}\mathrm{n}\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{p},\mathrm{t}}-{\mathrm{l}\mathrm{n}\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{p},\mathrm{t}-1}$$

(3)

In this text, the province’s annual stock data on outward investment are used. Firstly, the data are adjusted for inflation by deflating them, using the Consumer Price Index (CPI) to capture real terms. Subsequently, the logarithm of the deflated outward investment data is obtained, denoted as lnofdi_p,t. Additionally, the lagged data from the previous period, lnofdi_p,t−1, are incorporated. Similarly, the difference between the levels of outward investment in the two periods is computed to understand the change in the level of outward investment within the province. This change serves as an important measurement in assessing the dynamics of outward investment within the province.

4.4. Control Variables and the Description

In the following content, control variables and the description are given. In addition, descriptive statistics of the major variables are stated in

Table 1. Descriptive statistics of the major variables.

	N	Mean	SD	Min	Median	Max
OFDI	496	0.343	0.344	−0.565	0.28	1.536
aging	496	0.306	0.726	−2.144	0.281	2.316
minwage	496	6.993	0.459	5.914	7.09	7.741
rdi	496	1.48	1.102	0.193	1.179	5.985
pgdp	496	10.488	0.652	8.966	10.572	11.851
stru	496	1.104	0.597	0.545	0.924	4.165
open	496	0.282	0.324	0.013	0.142	1.608
gov	496	0.256	0.19	0.09	0.21	1.289
human	496	8.728	1.198	4.548	8.765	12.288

.

• Pgdp: the log value of GDP per capita in the province;
• Stu: the industrial structure of the province (value-added in the tertiary sector/value-added in the secondary sector);
• Open: the province’s foreign trade dependence (total imports and exports/total GDP);
• Gov: the government intervention of the province (fiscal expenditure/total GDP);
• human: the quality of human capital of the province (average schooling years per capita);
• Freetrade: a dummy variable that takes 1 when the province is a Pilot Free Trade Zone in the current year, and 0 otherwise;
• Inter: Internet coverage in the province, the ratio of Internet-broadband-access users to the total number of users in the province.

Based on the preliminary descriptive statistics of the paper’s core indicators, several noteworthy patterns can be observed: (1) The overall aging process in China is progressively advancing and exhibits substantial regional variation. Figure 1a presents the population aging trend in China, as measured by the proportion of aged individuals (OLD). (2) The level of OFDI experienced an upward trajectory during the analyzed period and demonstrates a similar increasing pattern to that of the proportion of aged individuals in the population (denoted as OLD%). The logarithm of the OFDI stock from 2005 to 2020 is depicted in Figure 1b, revealing a noTable 48% increase in 2020 compared to 2005. Notably, the growth trends displayed in Figure 1b indicate a synchrony between the indicators of population aging and OFDI levels between 2005 and 2020. This suggests a plausible correlation between the two variables, warranting further comprehensive exploration in the subsequent sections.

5. Empirical Studies

In this section, the results and analyses of the paper’s benchmark regressions are first given, followed by an analysis of the empirical results.

5.1. Benchmark Regression

The results of the benchmark regression are shown in

Table 2. Benchmark regression results.

	(1)	(2)
	OFDI	OFDI
aging	0.064 ***	0.066 ***
	(3.156)	(2.973)
pgdp		−0.103
		(-0.775)
stru		−0.133
		(−1.672)
open		−0.332 **
		(−2.094)
gov		1.346 **
		(2.617)
human		0.015
		(0.123)
_cons	0.594 ***	1.469
	(7.133)	(0.928)
Prov/Year	Yes	Yes
N	496	496
r2	0.207	0.242

Note: t statistics in parentheses; ** p < 0.05, and *** p < 0.01.

.

Column (1) of Table 2 shows the regression results without the inclusion of the control variables group, and Column (2) shows the regression results after the inclusion of the control variables group. Column (1) shows the results of the baseline regression, with province effects and year effects fixed, but without the inclusion of the group of control variables, and the results show that changes in the aging rate of the population have a significantly positive impact on changes in the level of OFDI. The results of Column (1) show that the coefficient of the interaction that we focused on is always significantly positive at the level of 1% regardless of whether control variables are added. Population aging, by changing the demographic structure, affecting labor costs, and promoting technological progress, has a significant positive impact on the level of OFDI level by changing the population structure, affecting labor costs and promoting technological progress. Due to the large differences in population structure and economic development level among provinces, Column (2) controls for the control variables of GDP per capita, industrial structure, foreign trade dependence, share of fiscal expenditures, road network density, and share of deposit and loan balances. Importantly, even after controlling for these variables and fixing the province and year effects, the regression results remain robust, affirming the persistently positive and significant relationship between population aging and the level of OFDI. Therefore, the empirical evidence has confirmed the validity of Hypothesis 1.

5.2. Endogeneity Tests

Although demographic change is considered to be exogenous in nature, the regression model used in the previous section includes province fixed effects and year fixed effects to mitigate the issue of omitted-variable bias resulting from province-level factors. This approach helps control for potential interference and provides a more accurate estimation. However, there still exists the potential problem of reverse causality within the model. For instance, provinces may actively pursue economic development transformations and seek opportunities for outward foreign direct investment (OFDI) to facilitate a technological spillover. This may lead to a decreased demand for low-skilled production workers and subsequently contribute to an increase in urban aging. In this scenario, the decline in demand for low-skilled production workers prompts labor mobility toward urban areas, consequently contributing to the overall increase in urban aging. It is vital to acknowledge that the model may still face challenges related to reverse causality, primarily driven by regional initiatives to stimulate economic growth and pursue OFDI opportunities, which can impact demographic dynamics and urban aging.

In this study, a difference operation is employed to transform the original variables into a measure of change over two consecutive periods. Specifically, the difference between the variable in the current period and the variable in the previous period is utilized to capture the extent of change. This approach helps eliminate the effects of trends, seasonality, and cyclicality in time-series data, thereby mitigating the potential issue of a spurious correlation arising from common trends between the dependent and independent variables. Additionally, it mitigates the issue of possible co-trending between the dependent and independent variables. To construct dynamic panel data, the core explanatory variables and the explanatory variables are modeled through difference equations. Notably, the changes in the aging rate and the logarithmic changes in the level of OFDI serve as the transformed data. Moreover, considering the improbable direct influence of changes in the OFDI level on changes in the aging rate of the population, the endogeneity concern inherent in the model is adequately addressed. In order to enhance academic rigor, this paper incorporates two methods to test the potential endogeneity issue within the model.

5.2.1. System GMM Method

In this study, the estimation method known as the System Generalized Method of Moments (SYS-GMM) was utilized, building upon the framework introduced by Arellano and Bover (1995) [40] and Blundell and Bond (2000) [41]. The SYS-GMM approach combines first-order difference equations and level equations, employing the level variable as an instrumental variable for its first-order difference lag term. This sophisticated method addresses the issue of weak instrumental variables and exhibits greater effectiveness compared to the differential GMM estimation. The test results of the SYS-GMM estimation are presented in

Table 3. SYS-GMM.

	OFDI
L.ofdi	−0.057
	(−0.419)
aging	0.060 ***
	(2.976)
pgdp	0.288 **
	(2.339)
stru	0.068
	(1.258)
open	−0.235 **
	(−2.358)
gov	−0.362
	(−1.527)
human	−0.121 **
	(−2.449)
_cons	−1.369
	(−1.263)
Year	Yes
N	465
AR(1)	0.013
AR(2)	0.444
Sargan	0.509

Note: t statistics in parentheses; ** p < 0.05, and *** p < 0.01.

, providing valuable insights into the empirical analysis conducted in this study.

The findings obtained from the Systematic Generalized Method of Moments (GMM) estimation align with those derived from the baseline regression and maintain their significance, indicating the validity of the GMM test. The p-values of the AR(1) and AR(2) tests are less than 0.1 and greater than 0.1, respectively, suggesting the absence of auto-correlation in the perturbation terms. Furthermore, the p-value associated with the Sargan–Bassman test statistic exceeds 0.05, demonstrating that all instrumental variables are considered exogenous, thus enhancing the validity of the results. It is worth noting that the coefficients resulting from the test outcomes exhibit negative values. This can be attributed to the consideration of changes in the level of OFDI, specifically the dynamics of the logarithm of the OFDI stock. Similar to the principle of diminishing marginal utility, the overall trend of change in this variable is upward; however, the rate of change diminishes over time.

5.2.2. Instrumental Variable Two-Stage Least Squares Method

Based on the hypothesis proposed by Semykina and Wooldridge (2010) [42], when the residuals of the model exhibit minimal time-series correlation and are predominantly determined by disturbances in the current period, the issue of endogeneity can be addressed by replacing the unlagged core explanatory variables with lagged core explanatory variables from the previous period. This approach allows for the mitigation of endogeneity, preserving the integrity of the core conclusions presented in the paper. It is clear from this that the core conclusions of the paper have not changed. The results of the test are shown in

Table 4. IV-2SLS.

	Step 1	Step 2
	Aging	OFDI
L.aging	−0.256 ***
	(−6.378)
aging		0.196 **
		(2.224)
pgdp	0.004	−0.099
	(0.015)	(−0.580)
stru	0.336 *	−0.179
	(2.026)	(−1.622)
open	0.350	−0.432 ***
	(0.827)	(−2.629)
gov	−0.324	1.572 ***
	(−0.466)	(3.017)
human	−0.343 **	0.044
	(−2.050)	(0.491)
Prov/Year	Yes	Yes
N	465	465
r2	0.352	0.167
Kleibergen-Paap rk Wald F		17.618 > 16.38

Note: t statistics in parentheses; * p < 0.1, ** p < 0.05, and *** p < 0.01.

.

In this study, instrumental variables (IVs) were constructed by utilizing first-order lagged variables of the dependent variable. This approach is motivated by three primary considerations. (1) It is important to note that the possibility of endogeneity between variables X and Y in this paper is quite small. The paper effectively addresses the issue of endogeneity in the variable selection and model construction process. This serves as a key rationale for employing differential modeling techniques. (2) Selecting appropriate instrumental variables for demographic aging proves to be challenging, particularly within the context of dynamic panel data. Static panel data instrumental variables are not suitable in this scenario, thus further complicating the selection process. (3) Employing one-period lagged independent variables as instrumental variables is justified. The results of the one-stage regression, presented in Column (1), indicate that the instrumental variables satisfy the correlation hypothesis. The subsequent regression results in Column (2) display positive and significant coefficients, reinforcing the hypothesis and providing further evidence of the instrumental variables’ effectiveness. Therefore, these considerations collectively inform the decision to employ first-order lagged variables of the dependent variable as instrumental variables in this study.

5.3. Robustness Tests

Despite the inclusion of province fixed effects and year fixed effects in the benchmark regression model to address the omitted variable problem, it is acknowledged that certain economic policies during the same period may still influence the empirical results. In order to assess the robustness of the findings, this study conducted robustness tests across three dimensions, the results of which are presented in

Table 5. Robustness tests.

	(1)	(2)	(3)
	OFDI	OFDI	OFDI
odr	0.039 **
	(2.614)
aging		0.066 ***	0.065 ***
		(3.038)	(3.039)
pgdp	−0.103	−0.100	−0.138
	(−0.774)	(−0.738)	(−0.991)
stru	−0.133	−0.131	−0.113
	(−1.658)	(−1.569)	(−1.306)
open	−0.320 **	−0.278 *	−0.340 *
	(−2.056)	(−1.848)	(−2.012)
gov	1.344 **	1.434 ***	1.470 ***
	(2.687)	(2.951)	(2.845)
human	0.021	0.029	0.046
	(0.175)	(0.237)	(0.383)
freetrade		0.096 ***	0.109 ***
		(2.988)	(3.165)
inter			0.647
			(1.559)
_cons	1.415	1.297	1.362
	(0.887)	(0.815)	(0.870)
Prov/Year	Yes	Yes	Yes
N	496	496	496
r2	0.241	0.246	0.250

Note: t statistics in parentheses; * p < 0.1, ** p < 0.05, and *** p < 0.01.

.

• The replacement of independent variables is examined. The core explanatory variable in this study, the level of OFDI, is substituted with the old-age dependency ratio (ODR). The ODR represents the proportion of the elderly population relative to the working-age population, serving as an indicator for the age structure and burden of ageing in a region. The regression results after this replacement are displayed in Column (1) of Table 5, revealing that the coefficients remain positive and significant. This implies that population aging continues to have a positive effect on the promotion of OFDI levels.
• Excluding the impact of the Pilot Free Trade Zone Policy. The Pilot Free Trade Zone (PFTZ) policy is a contemporaneous policy that was implemented during the same period. The establishment of PFTZs began in 2013, and as of 2019, 19 provinces in China have established these zones. PFTZs are designated areas designed to facilitate the free entry and exit of foreign commodities without tariffs, aiming to promote trade and investment facilitation. Extensive research has confirmed the positive effects of PFTZs on China’s import and export behaviors, as well as attracting foreign investment. Therefore, this paper endeavors to account for the impact of this policy. To address this, a dummy variable is constructed, taking the value of 1 when a province has an established PFTZ in the current year, and 0 otherwise. This dummy variable is then added to the existing set of control variables in the baseline regression. By controlling for the PFTZ policy, the regression results, which are presented in Column (2) of Table 5, demonstrate robustness, thus mitigating the interference caused by the Pilot Free Trade Zone policy.
• In order to ensure the robustness of the regression model regarding the transformation of export trade, it is necessary to consider the potential impact of Internet development. As we are in the information age, the Internet has proven to be a significant contributor to international trade and export expansion, both at the national and enterprise levels (Freund and Weinhold 2004; Lin 2014) [43,44]. Export and OFDI are the primary paths for enterprise internationalization, and there exists a close relationship between export experience and OFDI. The existing literature suggests that the development and utilization of the Internet facilitates the exchange, search, and replication of information, thus enhancing the innovation capabilities of enterprises. Moreover, Internet development promotes international trade by reducing search and communication costs for importers and exporters, while also increasing the probability and scale of export activities by enterprises. In recent years, China has exhibited strong momentum in Internet development, with exponential growth in Internet broadband access ports since 2005, increasing more than tenfold by 2015. Therefore, it becomes necessary to control for the potential impact of Internet development on the transformation of export trade in the regression model. This study employs the broadband coverage rate of a province, which reflects the ratio of Internet-broadband-access subscribers to the total number of subscribers in the province, as a measure. The regression results, incorporating the proportion of Internet-broadband-access subscribers at the provincial level, are presented in Column (3) of Table 5, revealing a positive and significant coefficient on population aging. This ensures that the potential influence of Internet development on export trade transformation is effectively accounted for in the analysis.

Additionally, to strengthen the credibility of the empirical results, further models were employed, and the results are shown in

Table 6. Robustness tests (FE and RE).

	FE	RE
	OFDIC	OFDIC
age	0.135 **	0.140 **
	(2.234)	(2.413)
pgdp	1.440 **	1.187 ***
	(2.256)	(2.845)
stru	−0.250	−0.323
	(−0.487)	(−0.913)
open	−0.303	0.185
	(−0.564)	(0.323)
gov	0.455	−0.939 *
	(0.375)	(−1.731)
human	0.698 ***	0.616 ***
	(3.053)	(3.131)
_cons	−9.221	−6.065
	(−1.445)	(−1.582)
Prov/Year	Yes	Yes
N	496	496
r2	0.923	0.813

Note: t statistics in parentheses; * p < 0.1, ** p < 0.05, and *** p < 0.01.

. We adopted level variables to reexamine the relationship between the aging population and the level of OFDI. As the result stated in Table 6, both the fixed-effects model (FEM) and random effects model (REM) gave significant results. As the proportion of elderly people in the population increases, the level of OFDI in the region also rises, indicating a significant positive correlation between the two. The empirical results from the fixed-effects (FEs) and random-effects (REs) models align with the baseline regression findings, providing evidence of the robustness of the results in this study. Therefore, Hypothesis 1 (H1), that the level of population aging is positively related to the level of OFDI, as proposed in this study, is once again validated. This suggests that, despite limited labor supply and rising labor costs in the context of population aging, the presence of technological advancements and international capital circulation have made OFDI an investment decision for many companies to expand their overseas presence.

5.4. Heterogeneity Analysis

5.4.1. Analysis of Heterogeneity at the Regional Level

To better understand the heterogeneous impact of population aging and OFDI across different regions in China, this study conducted an analysis of regional variations. Given the disparities in economic development and aging levels among regions, the full sample was divided into three subsamples, representing the east, central, and west regions of China. This division allows for an examination of the relationship between changes in the aging population and the level of OFDI within each respective region. The findings are presented in

Table 7. Regional heterogeneity.

	Eastern	Central	Western
	OFDI	OFDI	OFDI
aging	0.054 **	0.138 **	0.055
	(2.913)	(2.535)	(0.800)
pgdp	−0.054	0.021	0.262
	(−0.181)	(0.079)	(1.044)
stru	−0.108	0.089	−0.204
	(−0.490)	(0.956)	(−1.762)
open	−0.187	0.607	0.650 **
	(−0.703)	(0.371)	(2.340)
gov	2.179	−0.115	2.027 ***
	(1.604)	(−0.131)	(5.563)
human	−0.221	0.227	0.092
	(−1.561)	(1.601)	(0.464)
_cons	2.800	−1.514	−2.688
	(1.027)	(−0.589)	(−1.004)
Prov/Year	Yes	Yes	Yes
N	176	128	192
r2	0.333	0.384	0.310

Note: t statistics in parentheses; ** p < 0.05, and *** p < 0.01.

.

Relative to the western region, the central and eastern provinces exhibit higher levels of economic development, more advanced capital markets, and better enterprise financing environments. However, these regions also experience a faster rate of population aging. From 2005 to 2020, the average elderly population dependency ratio in the central and eastern regions was higher than that of the western region, indicating a more pronounced trend of aging. Columns 1 to 3 in Table 7 present the regression results at the regional level. The findings reveal that the deepening of population aging has a significant effect on the level of OFDI in the east and central regions of China. In contrast, the effect is not found to be significant in the west region. The results of the regional regression analysis demonstrate a substantial variation in the relationship between aging and levels of outward foreign direct investment (OFDI) across different regions. Specifically, the eastern and central regions, characterized by earlier and more pronounced aging, have presented greater opportunities for international investment, leading to a more proactive engagement in OFDI to sustain economic vitality. In contrast, the western region, experiencing a relatively slower aging process, has exhibited a smaller scale of OFDI due to factors such as the investment environment and resource endowments. These findings provide further support for Hypothesis 1 of this study and contribute to a more comprehensive discussion on regional consistency.

5.4.2. Coastal–Inland Heterogeneity

The coastal–inland heterogeneity of the impact of population ageing on firms’ OFDI was analyzed. The results are shown in

Table 8. Coastal–inland heterogeneity.

	Costal	Inland
	OFDI	OFDI
aging	0.077 ***	0.065
	(5.176)	(1.663)
pgdp	−0.104	0.145
	(−0.445)	(0.939)
stru	−0.231	−0.023
	(−1.147)	(−0.232)
open	−0.277	−0.251
	(−1.234)	(−0.495)
gov	2.436	1.779 ***
	(1.400)	(2.890)
human	−0.230 *	0.114
	(−1.925)	(0.753)
_cons	3.503	−1.844
	(1.486)	(−0.865)
Prov/Year	Yes	Yes
N	176	320
r2	0.366	0.259

Note: t statistics in parentheses; * p < 0.1, and *** p < 0.01.

.

Due to varying levels of economic development across different regions in our country, particularly between coastal and inland areas, there are significant differences in their respective development situations. A sample regression was conducted based on whether provinces are coastal or inland, which revealed that the aging population has a significant positive influence on OFDI in coastal areas but not in inland areas. This may be attributed to the coastal regions having a comparative advantage in terms of technological and human capital.

Despite the gradual deepening of China’s population aging, which reduces the demographic dividend, the country has accumulated a substantial amount of capital through persistent trade surpluses. Additionally, under the “learn from doing” effect, the human capital and technological accumulation of the elderly population benefit the introduction, absorption, and enhancement of advanced foreign technology, as well as the development of local technological capabilities. This enhances China’s comparative advantage in capital-intensive industries with a high knowledge and technology content. In the context of population aging, labor costs are rapidly increasing, while the costs of capital and technological factors are gradually decreasing, accelerating the substitution of labor. Therefore, when facing population aging, coastal cities are better equipped to offset the negative impact through avenues such as technological progress, thus improving productivity and smoothly seizing opportunities for outbound direct investment.

6. Mechanism Tests

The results of the tests for the factor cost effect and the technological progress effect are shown in

Table 9. Mechanism tests.

	(1)	(2)	(3)	(4)
	minwage	OFDI	rdi	OFDI
aging	0.020 **	0.056 **	0.052 ***	0.054 **
	(2.744)	(2.517)	(3.237)	(2.400)
minwage		0.508 *
		(2.042)
rdi				0.222 ***
				(3.568)
pgdp	0.102 *	−0.155	−0.427 **	−0.008
	(1.997)	(−1.210)	(−2.224)	(−0.066)
stru	0.059	−0.163 *	−0.223 *	−0.083
	(1.548)	(−1.956)	(−1.758)	(−0.950)
open	0.095	−0.380 **	−0.866 ***	−0.140
	(1.535)	(−2.229)	(−3.859)	(−0.848)
gov	0.247	1.221 **	−1.397 ***	1.656 ***
	(1.488)	(2.254)	(−2.972)	(3.717)
human	0.016	0.007	−0.025	0.020
	(0.490)	(0.056)	(−0.242)	(0.164)
_cons	4.990 ***	−1.067	6.208 ***	0.093
	(8.529)	(−0.510)	(2.942)	(0.067)
Prov/Year	Yes	Yes	Yes	Yes
N	496	496	496	496
r2	0.959	0.260	0.593	0.262

Note: t statistics in parentheses; * p < 0.1, ** p < 0.05, and *** p < 0.01.

. The logarithm of the minimum wage level, denoted as “minwage”, is selected as a proxy indicator for factor cost effects. The logarithm of the minimum wage level, denoted as “minwage”, is selected as a proxy indicator for factor cost effects.

6.1. Factor Cost Mechanism Test

Based on the literature review and theoretical analysis presented earlier, the changes in factor costs serve as an important channel through which population aging affects OFDI. On one hand, population aging leads to a shortage of labor force and subsequent increase in labor costs, which poses challenges for enterprises that have relied on the “demographic dividend” as their development model. As a result, these enterprises seek investment opportunities externally to upgrade and transform their development models. On the other hand, as labor costs rise, the relative cost of capital investment decreases compared to labor, making it easier for enterprises to invest in assets to reduce production costs and improve production efficiency. Increased capital investment facilitates the transformation of enterprises from traditional labor-intensive comparative advantage models to more sustainable models emphasizing investment and technological progress. To empirically test these mechanisms, this study utilizes relevant provincial-level data from the annual China Statistical Yearbook over the years.

This study attempts to demonstrate the labor cost channel from the perspective of the minimum wage standard. The minimum wage standard is a compulsory regulation imposed by the government onto employers, aimed at guaranteeing a fundamental level of livelihood for the workforce. It has been officially enforced since 2004, undergoing periodic adjustments at intervals of 1 to 3 years. This dynamic framework effectively captures the incremental rise in regional labor costs over time. In light of this, prior studies have delved into the ramifications of minimum wage hikes on export trade (Gan et al., 2016) [45]. In this study, annual data on minimum-wage standards were used to measure the current labor costs in the province and regressed. Table 9 displays the results, where the first column utilizes the difference in the minimum wage of urban employees as the dependent variable. The coefficient of aging is significantly positive, indicating that regions with a higher degree of aging experience a faster increase in labor costs, thereby promoting local enterprises to engage in OFDI. This result aligns with the expectations set forth in this study and further validates the reliability of the labor cost channel (H2).

6.2. Technological Progress Mechanism Test

Technological progress plays a crucial role in the relationship between population aging and corporate investment decisions, particularly in the context of outbound foreign direct investment (OFDI). While much of the existing literature focuses on developed countries transitioning toward an aging society and its impact on technological innovation and progress (Acemoglu and Restrepo 2021; Costinot et al., 2018) [9,38], limited research exists within the Chinese context. However, it can be theoretically assumed that the effect of population aging on technological progress is still applicable in China. This assumption is based on the premise that the scarcity of labor resulting from aging necessitates the adoption of advanced technologies to bridge the workforce gap and sustain consistent productivity growth.

To approximate the level of technological progress, the share of science and technology expenditures in provincial fiscal budgets is employed as a proxy. A higher investment in science and technology, reflected by a larger share of fiscal expenditures allocated to these areas, indicates a greater emphasis on technological development and a higher likelihood of achieving technological progress. The regression analysis conducted in Columns (3) and (4) of Table 9 demonstrates that population aging significantly increases provincial spending on science and technology. In other words, provinces experiencing greater aging and labor shortages tend to allocate greater resources toward scientific and technological advancements, thereby achieving progress in this domain. Consequently, population aging plays a significant role in shaping the level of OFDI through its influence on technological progress. This finding confirms the validity of the third mechanism proposed in this study, further establishing the relationship between population aging and OFDI. Thus, Hypothesis 3 was verified.

7. Further Study Considering US–China Conflict

China’s excessive reliance on the US market in the long term has substantially weakened its ability to handle international market risks. Amidst the ongoing trade dispute between China and the US, it is crucial to re-evaluate and break the habitual investment pattern with major trading partners, as it holds significant implications for optimizing China’s outward foreign direct investment (OFDI) strategy. Therefore, this study sets 2018 as a boundary and introduces a time dummy variable, where 0 represents pre-2018 and 1 represents post-2018. A multiplicative interaction term is constructed with the core explanatory variable Δ${\mathrm{A}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{p}}$ and included in the regression equation, with the regression results presented in

Table 10. Conflict of 2018.

	OFDI
aging	0.068 ***
	(3.138)
y18	−0.522 *
	(−1.769)
aging_y18	−0.088 ***
	(−3.172)
pgdp	−0.112
	(−0.894)
stru	−0.128 *
	(−1.794)
open	−0.292 *
	(−1.889)
gov	1.338 **
	(2.627)
human	0.003
	(0.024)
_cons	1.633
	(1.060)
Prov/Year	Yes
N	496
r2	0.247

Note: t statistics in parentheses; * p < 0.1, ** p < 0.05, and *** p < 0.01.

. According to the table, the coefficient of the interaction term is significantly negative, indicating that the China–US trade tensions since 2018 have led to a deteriorated investment environment for Chinese OFDI. Consequently, the promoting effect of population aging on OFDI expansion has significantly declined. This signifies that while population aging positively affects OFDI expansion, other external factors can have substantial impacts on this relationship. Therefore, a comprehensive consideration of various factors is necessary when determining China’s OFDI strategies.

8. Conclusions and Recommendations

During the 40-year period of economic reforms and liberalization, the FDI has been a key driver of China’s economic growth. As China’s open economy continues to expand, its development model is shifting toward greater diversity, with a focus on outbound investment rather than solely attracting foreign capital. This transition coincides with the challenges posed by an aging population and rising labor costs, highlighting the need for further examination and analysis. In light of these circumstances, this study aimed to empirically investigate the influence of population aging on changes in OFDI levels in China. To accomplish this, the research constructed dynamic panel data and difference equations by employing data on population aging and provincial-level OFDI.

The study findings indicate that population aging has a positive influence on the level of OFDI, particularly in the east–central provinces, while no significant impact is observed in the western provinces. Additionally, the promotional effect of population aging on OFDI is significant in coastal provinces but insignificant in inland provinces. The impact of population aging on outward foreign direct investment (OFDI) can be explained by two underlying mechanisms. Firstly, population aging leads to an increase in labor costs, resulting in a favorable cost dynamic for capital-intensive industries, thus attracting foreign direct investment. Secondly, population aging acts as a driving force for regions to invest in research and development (R&D), stimulating technological progress and enhancing innovation capabilities, consequently attracting OFDI. The findings of this study hold significant policy implications for guiding China’s seamless transition and high-level openness amidst its demographic changes. The government can enhance its regulation and policy support for population aging, with a focus on boosting technological innovation capabilities and reducing labor costs. These measures would make China more attractive for foreign investment, optimize the economic structure, and facilitate transformative upgrading. Moreover, policies should address regional disparities, as the impact of population aging on OFDI may vary across different regions. It is crucial to devise location-specific policy measures tailored to the distinct conditions of each region.

Based on the findings, this paper offers several specific policy recommendations, as follows.

Firstly, it is recommended that policymakers provide robust policy support to facilitate outward foreign direct investment (OFDI) in light of China’s demographic transition and labor shortage. As population aging leads to increased labor costs, firms are inclined to invest in capital-intensive industries to mitigate labor scarcity. Policy support during the demographic transition would encourage enterprises to engage in OFDI, relieving labor pressures and enhancing China’s global competitiveness. By guiding and supporting enterprises in their OFDI endeavors, the government can drive structural optimization, promote economic transformation, and achieve long-term sustainable development.

Secondly, attention should be directed toward human capital accumulation and the optimization of talent allocation. Historically, China has enjoyed a comparative advantage in export trade driven by the demographic dividend. However, as the inevitable process of population aging unfolds, the demographic dividend will gradually diminish. To foster a high-level of openness to the global economy, relevant government entities must implement corresponding measures. Firstly, the government should augment investments in education, with a focus on developing both basic and higher education, while ensuring the rational allocation of educational resources. This will actively nurture and accumulate human capital, thus providing the economy with a sufficient labor force.

Thirdly, to promote technological progress, it is crucial to incentivize research and development (R&D) and innovation, particularly as China transitions into an aging society. This approach has dual benefits: it conserves labor resources by adopting advanced technologies and enhances total factor productivity by integrating cutting-edge production technologies and innovative talents. Enterprises should increase investment in R&D, develop and implement new technologies, and attract innovative talent. Simultaneously, the government should provide support for R&D and innovation through policies such as tax reductions and the recognition of scientific research achievements. However, it is important to address potential labor force displacement caused by technology adoption. The government should prioritize measures such as unemployment protection and re-employment training to guide workers toward industries with high demand, striking a balance between technological advancement and labor substitution. This approach enhances the quality and competitiveness of the labor force and promotes sustainable economic development.