Complementarity or Crowding Out: The Effects of Government-Led Philanthropic Development (2024)

2.1. Government Guidance and the Development of Philanthropy

Public goods constitute a significant domain over which the government exercises its guiding function. The focus of government guidance lies in the behavior and modalities of reintegrating societal resources, as well as establishing and perfecting an environmental and conditional system conducive to the expansion of needs for all societal actors. The government’s financial guidance for the advancement of philanthropy is bifurcated into direct and indirect approaches. The direct approach encompasses special grants to charitable organizations and activities, governmental subsidies, matching funds, service procurement, venture capital for public welfare, support and seed funds, municipal bonds, consumption vouchers, and loan interest subsidies. The indirect approach involves the government’s public expenditure oriented toward society, tax incentives, resource coordination, and public opinion propaganda, which collectively play a leading role in the overall development of philanthropy. In this study, we concentrate on the indirect method of governmental guidance, specifically examining the impact and relationship of governmental investment in livelihood expenditure on the factual status of philanthropy.

The philanthropic sector primarily addresses the diverse and heterogeneous needs for which public goods supply is insufficient or the government is unable to fulfill, exhibiting greater flexibility, trustworthiness, and diversity compared to traditional social assistance systems and market supply. The government failure theory posits that there is a complementary relationship between government and nonprofit organizations, suggesting that the scale of the nonprofit sector in a specific area should be negatively correlated with the size of the government. In contrast, the interdependence theory suggests that government and nonprofit organizations can have a complementary relationship, implying that there should be a positive correlation between the scale of the nonprofit sector and the size of the government in a specific area [8]. Beyond these two direct theories of government–nonprofit interaction, there are also indirect interactions; for example, government activities may indirectly shape the scale of the nonprofit sector by affecting its resource base, such as private donations [9]. Ye Shihua and Gong Xiaochen argue that charitable donations are not entirely the same prosocial behavior; individuals’ preferences for various philanthropic causes are driven not only by individual-level social capital factors but also by the degree of state intervention in different philanthropic fields [10]. In China, institutional trust has been found to be a favorable predictor of charitable donations, with donors’ giving behaviors deeply rooted in the institutional context. The significance of institutional trust varies in the philanthropic sector depending on the degree of state intervention. For instance, poverty alleviation has long been regarded as a political priority in China, with the Chinese government maintaining a continuous dominant position and strong monopoly over NGOs in this field; people’s willingness to donate is primarily guided by public institutions [11,12].

In the development of philanthropy with Chinese characteristics, philanthropy serves as an essential tool for the construction of modern social governance and the connection of diverse societal actors. The government acts as the leader, organizer, and manager of the philanthropic sector, wherein the logic of public welfare is always subordinate to the national logic [13]. While maintaining the operational order of philanthropy, the government also participates in its planning and development, with fiscal policy being the most direct reflection of the relationship between the government and philanthropic organizations. Khannahe and Sandler have noted that the role of government grants in guiding philanthropic donations is manifested in the following ways: (1) government grants can enhance the reputations of charitable organizations; (2) government grants are typically accompanied by governmental oversight of nonprofit entities, which can curb information asymmetry and increase the willingness of potential donors to contribute; (3) government grants produce a “buy-in” effect, where donors adjust the direction of their donations in accordance with the focus of government grants; (4) the presence of a “matching” component in government grants is akin to reducing prices, thereby stimulating private donations; (5) if it is perceived that overhead and fundraising activities are managed efficiently, the monitoring function of government grants can lower perceived prices in the minds of donors, and a decrease in perceived price can prompt donors to increase their giving; and (6) government subsidies can bypass the outcomes of the Standard Neutrality Theorem associated with crowding out, as long as the government sufficiently “partially recovers the amount of the grant from taxes on non-donors” [14].

2.2. Regional Differences in Government-Led Development of Philanthropy

The Chinese philanthropic market has long been plagued by severe information asymmetry, with the coexistence of a “dammed lake” of philanthropy and an insufficient level of assistance received by vulnerable groups, highlighting the structural imbalances in the market. Nan Rui and Zhai Yujia, using sample data from 2000 to 2011, applied the Gini coefficient, logarithmic mean difference, and Theil index to analyze the disparities and trends in local charitable donation levels. They pointed out that the disparity in regional charitable donation levels in China is significantly high and has shown a clear trend of increasing expansion, with intra-regional differences playing a dominant role. The differences in charitable donation levels among the three major regions—east, central, and west—exhibit a “U-shaped” characteristic, with inter-provincial differences being more pronounced than inter-regional differences, yet both show a similar trend of increasing disparity [15]. The information advantage theory posits that geographical proximity presents an advantage in information transmission, which can reduce the cost of information acquisition and trigger investors’ “home bias” and “familiarity bias”. Lu Jiahuan and Dong Qiang, by integrating administrative and census data, examined the scale of the nonprofit sector in 334 counties in China, noting that the numbers of registered nonprofit organizations per ten thousand people in the southeast and central–western regions of China are higher than in other areas. They also pointed out that population heterogeneity, government social expenditure, and government decentralization have positive impacts on the expansion of the scale of nonprofit organizations at the county level [16]. The factors that drive the growth of nonprofit organizations and lead to changes in the scale of nonprofit organizations in different geographical locations have become hot topics of interest among scholars. Zhang Xiaojun and Peng Zhengbo analyzed the impact of the institutional environment and the supply and demand of public services on the development of social organizations based on provincial panel data from 2005 to 2013. The results indicate that both the institutional environment and the supply and demand of public services have significant impacts on the development of social organizations [17]. However, scholars also raise doubts, particularly as to why the development of civil organizations shows obvious regional differences under the same national institutional environment. This doubt implies that merely considering the institutional environment is insufficient to fully understand the regional differences in the development of civil organizations in China. Li Guowu and others believe that civil organizations may be related to the population size of a region. Civil organizations are the products of people’s associative behavior and charitable activities. The larger the population, on the one hand, the more social individuals are willing to engage in volunteer and charitable causes, thereby increasing the possibility of establishing civil organizations; on the other hand, there is a greater social demand for civil association and nonprofit services, and this demand is more consistent and diverse. Therefore, the larger the population size of a province, the greater the number of civil organizations [18].

2.3. The Effect of Government Guidance on the Development of the Philanthropy Sector

The preference effect of government guidance on the development of philanthropy has been studied by scholars mainly in terms of preferences for government spending and the motivations and preferences of individuals to donate; less focus has been placed on how these behaviors are indirectly shaped by macrostructural factors. Brooks finds that government funding of nonprofit activities can produce more dynamic outcomes depending on the level, amount, and frequency of the funding, as well as the level of government funding directed to the subsectors of the nonprofit economy, in terms of both resource crowding out and resource crowding in [19]. Specifically, governmental subsidies to nonprofit organizations involved in arts and cultural projects can, to a certain extent, leverage private philanthropic funds, whereas if the subsidies exceed a certain threshold, nonprofit organizations may become akin to “quasi-public institutions”, potentially leading to a crowding-out effect from government funding.

Scholars have less systematically examined how and why donors choose to support certain charities, while neglecting others. For instance, women are more inclined to support animal charities [20], and older people are more likely to donate to religious, welfare, and health charities [21]. De Wit and Broese matched governmental support with organizational fundraising information in the Netherlands from 2002 to 2014 and found that donations in the social services, health, and nature sectors are more susceptible to being replaced by government support, whereas there is no crowding-out phenomenon in the international development sector [22]. Arjen et al. examined the relationships between governmental expenditure and private donations across 19 countries and tested the crowding-out hypothesis in different welfare sectors to understand how governmental spending leads to the crowding out or crowding in of private donations across sectors. Their study found a stronger crowding-in effect in the environmental sector; meanwhile, in the social services and health sectors, governmental spending had little impact on the number of donors. Their study also explored “lateral crowding in” (charitable substitution), whether an increase in government spending in one sector leads to an increase in individual donations in other sectors [9]. Governmental expenditures in the social services and health sectors are associated with a higher number of donors in the environment, arts, culture, and international aid sectors, indicating that high levels of social welfare spending in core welfare sectors prompted donors to shift toward other areas, which may also be related to the relationship between government and civil society. Lu Jiahuan et al. found a statistically significant positive correlation between the size of government and the size of the nonprofit sector in the social services sector, while a similar positive correlation in the non-social services sector was statistically insignificant [23], a possible explanation for which is the increasing cooperation and funding between government and nonprofit activities. Under these circ*mstances, the complementary relationship between government and nonprofit organization activities outweighs the substitutive relationship and dominates the government–nonprofit relationship in social services. Therefore, overall government activities may slightly influence the activities of social service nonprofit organizations in an area. In contrast, in the non-social services sector, cooperation between government and nonprofit organizations is less common, and the nonprofit sector plays a more significant role in filling gaps.

3.1. Description of Data and Variables

3.2. Model Setting

In this context, i denotes the province, and t represents the time period. y_it signifies the number of charitable donations, the total number of social organizations, the total volunteer service hours, and the number of individuals involved in province i during year t. ${\mathrm{E}\mathrm{C}\mathrm{A}}_{\mathrm{i}\mathrm{t}}^{\prime }$ indicates the level of government guidance in province i during year t. μ_i represents the fixed effect of time, and ε_it is the idiosyncratic error term that varies with both the individual and the time period.

4.1. OLS Regression

The results of regression using cross-sectional data are presented in Table 2. Model (1) was used to examine the impact of governmental guidance on the total number of charitable donations. After controlling for variables such as fiscal capacity, economic development level, and donative capacity, an increase of one percentage point in governmental guidance leads to a 3.5% decrease in total donation revenue, demonstrating a crowding-out effect; one possible reason is that the areas of governmental guidance overlap significantly with the scope of charitable activities, leading donors to believe that governmental activities can substitute for charitable activities. In terms of fiscal capacity, the more general public budget expenditure, the more total charitable donation revenue; conversely, the more general public budget revenue, the less charitable donation income, possibly because, from the macro perspective of the state, public expenditure has a guiding effect on donation income. Additionally, an increase in fiscal revenue may encourage the government to increase the supply of public services, but it may also lead to a decrease in the efficiency of public service provision. According to our results, an increase in government fiscal revenue crowded out charitable donations.

Model (2) was used to investigate the impact of governmental guidance on the number of social organizations. After controlling for relevant variables, for every one percentage point increase in governmental guidance, the number of social organizations decreases by 13.7 percentage points, which is significant at the 1% level, indicating a crowding-out effect. The higher the general public budget expenditure, the greater the number of social organizations; similarly, the more general public budget revenue, the greater the number of social organizations, possibly explained by the idea that, from a macro perspective of the state, government expenditure has a guiding effect on the “quantity” dimension of the philanthropic sector, and the two develop in concert with each other.

Model (3) was used to examine the impact of governmental guidance on the number of volunteer engagements. It was found that an increase in governmental guidance may reduce the number of volunteer engagements, possibly because the service areas of volunteer service encompass poverty alleviation, relief for the poor, support for the elderly, aid for orphans, medical assistance, disaster relief, medical aid, and educational support, among others. A substantial number of social workers in civil affairs also engage in corresponding activities, which may crowd out volunteer engagements. However, the more general public budget revenue there is, the more the number of people participating in volunteer service increases. In 2016, the Central Propaganda Department and other departments jointly issued the “Opinions on Supporting and Developing Volunteer Service Organizations”. Since 2015, government work reports have also proposed supporting the healthy development of volunteer service work, suggesting support for the operation and management of volunteer services through government purchases and other means. Various provinces have also introduced incentive mechanisms for volunteer services, thus motivating, rewarding, and safeguarding the enthusiasm of volunteers through the exchange of points for volunteer time, honor rewards, etc. An increase in government revenue may increase the frequency and scope of government service purchases, thereby increasing the number of volunteer engagements; an increase in general public budget expenditure may reduce the number of volunteers, but this is not statistically significant, which is consistent with the results of Day and Devlin’s study on the impact of Canadian government expenditure on volunteer time donations; that is, government funding affects the decisions of volunteers, not the time spent by volunteers [32].

Model (4) was used to investigate the impact of governmental guidance on volunteer service time. The more governmental guidance there is, the less volunteer service time there tends to be, which may have similar reasons to those discussed for Model (3). In terms of control variables, the more general public budget revenue there is, the more volunteer service time there is; the more general public budget expenditure there is, the less volunteer service time there is, but this is not statistically significant.

4.2. Fixed-Effects Regression

OLS regression may be biased due to omitted variables; hence, we employed a fixed-effects model for re-regression. Compared to the cross-sectional data, panel data models can control for the OLS estimation bias caused by unobservable economic variables; compared to time series models, panel data models expand sample information, reduce multicollinearity among economic variables, enhance the efficiency of the estimator, and accurately reflect the dynamic adjustment of economic variables. In the fixed-effects model presented in Table 3, the regression results of each model are essentially consistent with the OLS model after controlling for time, region, and related control variables. However, in terms of the number of social organizations, the more general public budget revenue, the lower the number of social organizations.

4.3. Robustness Test

To further verify the reliability of the above conclusions, we conducted robustness tests based on the theoretical assumptions mentioned above; overall, these tests do not alter our original research findings, as demonstrated in the following factors: (1) Endogeneity issues. The random effects model addresses endogeneity issues through a mixed maximum likelihood estimation. Upon testing with the random effects model, the main conclusions of this paper still hold true, as shown in Table 4. (2) Exclusion of sample data from Xinjiang and Tibet. Given that the development of the philanthropic sectors in Xinjiang and Tibet is relatively slow, with a lack of funds, issues of legal legitimacy and social recognition, and mostly dominated by government-run entities, the main conclusions of this paper remain valid even after excluding this portion of the sample data, as shown in Table 5.

5.1. Heterogeneity Analysis of Regional Development

In the comparative analysis by region, Model (1) reveals that governmental guidance only affects the total number of donations in the western region, while the dimension of fiscal capacity has no impacts on donations in the eastern, central, or western regions, possibly because, in economically developed areas, where income levels are relatively high, people donate more out of “altruism”, or the public’s attention to the western region in terms of policy is higher. For instance, when the central finance supports the selection of social organizations to participate in social service projects each year, in addition to setting up projects specifically for the western region, the quota indicators for the western region are also higher than those in other regions, thus making it easier to obtain more social donations. In Model (2), governmental guidance only affects the number of social organizations in the eastern region; in Model (3), governmental guidance significantly affects the central and western regions, with the greatest impact on the central region; and in Model (4), governmental guidance only affects volunteer service time in the central region (see Table 6).

5.2. Heterogeneity Analysis of Industry Regions

Due to the lack of data on donations received by social organizations across various industries, the data analysis in this section only includes the impacts of governmental guidance on the number of social organizations in each industry (see Table 7). In the comparison of industry-specific numbers, governmental guidance significantly affects the number of educational, cultural, and health industry social organizations in the eastern region; however, in the central region, there is only statistical significance for the cultural industry. Moreover, in the western region, only the health industry shows statistical significance. The development of social organizations is closely related to regional systems, encompassing factors such as the level of regional economic development, degree of marketization, level of openness to the outside world, the atmosphere of public welfare culture, and differences in government management and social self-governance organizations, which are also the main reasons for the asynchronous growth of social organizations between provinces [33]. Wang Yuzhen et al., through time series data of 31 provinces from 2002 to 2014, concluded that, during the transition period, provinces with higher levels of regional economic development, larger population sizes, and higher degrees of openness to the outside world exhibited more prominent and obvious public governance demands for social organizations. The relationship between the government and social organizations is complementary and the two develop together; however, provinces with faster urbanization processes have not significantly highlighted the governance demands for social organizations, indicating, to some extent, that social issues caused by urbanization are still mainly resolved by the government [34]. From the perspective of the data used in this study, as social organizations evolve and develop, the negative correlation between governmental guidance and the number of social organizations in the eastern region indicates that, while local governments continuously increase public service expenditures, they partially crowd out the number of social organizations. Of course, the reduction in numbers is also related to the policies of vigorously rectifying illegal social organizations and promoting standardized development of social organizations in 2008 and 2014. Compared with the eastern region, the central and western regions have fewer resources for the development of social organizations, with the central region having even fewer resources than the western region. Due to the economic development and frequency of disasters in the western region, it receives more charitable resources than the central region [35]. The positive correlation between government guidance in the cultural industry and the total number of social organizations in the central region may be due to the fact that, compared with the eastern and western regions, the central region is still in a rapid urbanization phase, and the number of cultural social organizations is relatively low. In 2020, the number of cultural social organizations in the eastern region was 13,493, compared with 7856 in the western region and only 5951 in the central region, and the eastern region is still in a phase where the government and social organizations promote each other’s development. The health industry in the western region shows a crowding-in effect; the positive correlation between government guidance in the health industry and the total number of social organizations in the western region may be due to the relatively small number of social health-related organizations and slow development in the western region. In 2020, there were 19,850 social health-related organizations in the eastern region, compared with 8389 in the central region and only 8137 in the western region, thus showing a crowding-in effect.

In terms of general public budget expenditure, there are significant impacts on the numbers of organizations in the science and technology, culture, and sports industries in the eastern region; in the central region, only the science and technology and health industries exhibit statistical significance; and in the western region, only the science and technology industry shows statistical significance. All of these demonstrate a crowding-out effect. General public budget revenue has statistically significant effects on the number of organizations in the science and technology and health industries in the eastern region, with only the health industry being statistically significant in the central region, while there is no impact on the number of industries in the western region. According to the distribution of the data, there is a positive correlation between general public budget revenue and the number of industry organizations, which may represent the fact that, as government revenue increases, the total investment in the science and technology and health industries also increases. Social organizations perceive this signal and follow suit, showing a crowding-in effect.

Appendix A

The eastern region includes Beijing, Tianjin, Hebei, Liaoning, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Hainan; the central region includes Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan; and the western region includes Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang, and Tibet.