The Effect of Government Grants on Private Giving to East Asian Nonprofits: Implications for Social Work Managers

For effective financial management, social work managers must clearly grasp the relationship between government grants and private contributions, which is frequently characterized as crowding-out effects. Crowding-out effects have been investigated for various types of nonprofits in the U.S., and the results have been mixed. In spite of its popularity in nonprofit research, the theory has not been applied to nonprofits serving minority communities. This is the first pilot crowding-out study looking at East Asian nonprofit organizations, including Chinese, Korean, and Japanese-American nonprofits in the NY and NJ metropolitan area (n = 410). Through a panel analysis, the current study found a significant crowding-in effect for donations to East Asian-American nonprofits (p < .01). The relationship between government grants and private giving was different for each East Asian-American nonprofit organization. Particularly, donors of Chinese and Japanese-American nonprofit organizations donated more money when their charities received more government grants (p < .05). In contrast, we found crowding-out effects for Korean-American nonprofit organizations, but the result was not significant (p > .05). The estimated crowding-in effects of government grants on private giving by each of the East Asian countries were explicated based on each country’s social, political, and cultural background such as the quality of the charity, transparency, and political trust. Social work managers in ethnic nonprofit organizations should establish different strategies to help shape donor giving patterns according to the effect of government grants.

Due to these two sources of funding making up such a large portion of nonprofit revenues, the relationship between government grants and private giving in human services nonprofit organizations, known as the crowding-out effect, which is an economic theory suggesting when the government gives a grant to a private charitable organization, private giving will be driven down, has been examined through a body of literature in nonprofit studies (Andreoni et al., 2014;Heutel, 2014;Kim & Van Ryzin, 2014;Lu, 2016). Tinkelman (2009) andDe Wit andBekkers (2017) reviewed prior studies related to the crowding-out effect using the meta-analysis method. Tinkelman (2009) identified 46 empirical and non-empirical studies and summarized their findings. De Wit and Bekkers (2017) analyzed 70 studies that matched the crowdingout effect topic and identified 422 findings of crowding-out/in effects from the studies. The crowding-out effects have been investigated using various hypotheses, research methods, datasets, and different types of nonprofit organizations.
Despite a wide array of efforts that have been dedicated to examining the crowdingout effect between government grants and private giving to nonprofits, all of the prior studies on this topic have investigated nonprofit organizations in mainstream society. Thus far, scholarly research has not examined the crowding-out effect for nonprofits serving culturally-specific communities in the U.S.
Asian-American nonprofit organizations have expanded along with the rapid growth of Asian-American populations compared to other minority communities during the last two decades (Hung, 2005;Hung & Ong, 2012). Yet, Asian-American nonprofit studies have not been highlighted by nonprofit researchers. Like other nonprofits, financial management is a significant issue for Asian-American nonprofit organizations. A lack of experience in securing public and private funding resources has a significant impact on Asian-American nonprofits' sustainability (Hung & Ong, 2012).
This pilot study investigated the crowding-out effect targeting Asian-American nonprofit organizations in the U.S. This paper seeks to fill the gaps in the literature by examining the relationship between government funding and private donations to Asian-American nonprofits. Particularly, we focus on East Asian-American nonprofits like the Chinese, Japanese, and Korean-American organizations in the New York and New Jersey Metropolitan area, given that East Asian groups made up the majority of Asian-American nonprofits (Hung, 2005). The aim of this study was to expand our understanding of the nature of fiscal resources, particularly for those rooted in the macro social work perspective and social work managers serving in East Asian-American nonprofit organizations.

Literature Review
Diversifying revenue sources is critical to an organization's fiscal health (Rebetak & Bartosova, 2019). Such fiscal sources are also known as a critical organizational domain in community practice (Hardcastle & Powers, 2004). Rohayait et al. (2016) highlight that nonprofits need to recognize various factors related to financial management aligned with both internal and external environments that impact their funding mechanisms. There are four major mechanisms intertwined with such forces or trends for funding the organization, including government grants, private giving, service fees, and other investments (Rapp & Poertner, 1992). Since a major focus of the current study is the relationship between government grants and private giving, given the fact that government funds and private giving are comprised of major financial sources of nonprofit revenues (NCN, 2010), a detailed discussion of other financial mechanisms is beyond the scope of this study. Therefore, we examined the effect of government grants on private giving, which is typically guided by the theory of crowding-out effects as a major area of fund-raising (Rebetak & Bartosova, 2019).

The Relationship between Government Funding and Private Contributions
As mentioned earlier, crowding-out effects theorize that an increase in one area is associated with a reduction in another area (i.e., increased government grants correlate with a decrease in private giving). The crowding-out effect assumes that the amount of charitable support changes over time as donors become aware of government grants being channeled to the organization (Horne et al., 2005;Lu, 2016). Many studies on crowding-out effects have investigated the relationship between government grants and private donations using different analytical methods, focused on various nonprofit subsectors, and with particular subsamples (De Wit & Bekkers, 2017). (2017), there were mixed results in terms of the estimated coefficients of the relationship between government grants and private giving for nonprofit organizations. Some studies reported a partial crowding-out effect (Andreoni & Payne, 2011;Cuellar, 2004;Dokko, 2009;Hungerman, 2009;Lindsey & Steinberg, 1990), while others found a crowding-in effect, occurring when higher government grants lead to an increase in private giving (Payne, 2001;Smith, 2007). Other studies revealed there was no significant relationship between government grants and private giving (Khanna et al., 1995;Lindsey & Steinberg, 1990). Tinkelman (2009) reported there were 134 crowding-out effects documented in 46 published and unpublished studies on the relationship between government funding and private donations. De Wit and Bekkers (2017) reported 262 crowding-out effects and 160 crowding-in effects in 70 studies. The estimated crowding-out coefficients varied, but approximately $0.17 of private giving decreased per dollar of government grant increasing across all of the prior studies with a 95% confidence interval (De Wit & Bekkers, 2017). Ambiguous correlations between government funding and charitable giving were also reported in other studies (Andreoni et al., 2014;Kim & Van Ryzin, 2014).

According to De Wit and Bekkers
In sum, existing studies indicate the relationship between government grants and donations has been proven inconclusive in three ways: Increased government funding is associated with private donors giving either less money, more money, or there is no statistically significant difference. In spite of such mixed results, what has been predicted more often in the relationship between government grants and private giving is the crowding-out effects at lower coefficient levels in the majority of the studies (Heutel, 2014).
A major reason for examining the crowding-out effect in nonprofit research is not only to identify the role of nonprofits as a service delivery system intertwined with policy through government subsidies, but also to strengthen nonprofit management to increase the number of private donors (Tinkelman, 2009). Particularly, the latter is more important for nonprofits, given that private contributions are considered a primary source of revenues for many nonprofits (Horne et al., 2005). If nonprofits are aware that government grants crowd out private giving, it is more likely that organizations would adapt strategies that motivate and inspire donors to give (Heutel, 2014), rather than focusing on grant winning proposals. On the other hand, if government grants signal nonprofits' quality of services, which brings about crowding-in effects, such background information would be strategically employed by the organization's management (Heutel, 2014). Therefore, investigating the crowding-out effect allows for a clearer understanding of the relationship between the two components that may lead to more stable funding mechanisms for the organizations (Tinkelman, 2009). In spite of such importance, however, the theory of the crowding-out effect has not been applied to nonprofits serving Asian-American communities. This oversight has occurred even though Asian American philanthropy is considered an emerging area in nonprofit research as the Asian-American population and thus the number of organizations have been growing (Hung, 2007;Semple, 2013).

Background of Asian-American Nonprofit Organizations
The Asian population in the U.S. grew 72% between 2000 and 2015, making it one of the fastest growing minority groups in the U.S. Nearly 20 million Asian-Americans made up 6.3% of the total U.S. population in 2015 (López et al., 2017). As the population has grown, Asian-American communities not only have initially developed informal organizations to create self-help networks, but also have provided formal and organized assistance by establishing nonprofit organizations (Hung, 2007). According to Hung, (2005), approximately 2,400 Asian-American nonprofits were formed between the 1980s and the 2000s, which equate to a 350% growth rate. Although Hung's study was published more than a decade ago, it remains a pathfinder that explores Asian-American nonprofits' nature including financial size and is still the most recent one to our knowledge.
Asian-American nonprofits can be categorized based on target populations for services: Pan-Asian Americans, which refers to organizations established to enhance the interests of all Asian-Americans, and ethnic nonprofits, referring to organizations with a focus on a specific ethnicity (Hung, 2005). Among 2,404 Asian-American nonprofits in 2000, the majority of the organizations were identified as ethnic-focused nonprofits, with 82.6% spread out among three East Asian-Americans: Chinese, Japanese, and Korean ethnicities (Hung, 2005). Such results are consistent with the more recent study of Hung and Ong (2012), who reported that more than 75.1% of Asian nonprofit organizations in the U.S. identified that their target populations for services were either Chinese, Korean, or Japanese.
Organizational size as a determining factor for sustainability, often measured by fiscal resources such as annual revenue (Hung & Ong, 2012), is a limitation among Asian-American nonprofits (Hung, 2005). For example, Hung (2005) assessed the general financial size of Asian-American nonprofits to be small and that less than 750 of 2,400 Asian-American nonprofits have filed Form 990. Form 990 is the Internal Revenue Service (IRS) document for gathering information about tax-exempt organizations and providing nonprofit organizations' financial statements to the public. All registered nonprofits in the U.S. are required to disclose their financial information on Form 990 by annually submitting it to the IRS. Among those that filed Form 990, less than half had annual revenues in excess of $25,000, excluding religious institutions (Hung, 2005). In addition, only a small portion of private grants have been awarded to Asian nonprofits in the U.S. The Greenlining Institute reported that only 1.5% of national foundation grants were awarded to Asian-led organizations, while 92.3% of national foundation grants were delivered to White-led nonprofit organizations in 2005 (Gonzalez-Rivera et al., 2008). The prior studies highlight that Asian nonprofits in the U.S. are more susceptible to challenge or failure (Hung & Ong, 2012). In a situation where developing fiscal resources has become a critical managerial issue (Hung & Ong, 2012;Young, 2010;Yung et al., 2008), Asian-American nonprofits must equip themselves to adopt revenue strategies to stably acquire, effectively maintain, and strategically diversify financial resources to ensure organizational survival (Chao, 1999;Froelich, 1999).
In the context of the growth in Asian populations and nonprofit organizations in the U.S., one should note that the growth of the population also signifies an increase in private donations by Asian-Americans (Semple, 2013). For example, according to Tsunoda (2010), Asian donors donated more than $120 million to various American higher educational institutes between 2006 and 2009. Tsunoda (2010) points out some distinctive qualities between Asian-American and non-Asian-American donors, asserting that Asian-American donors have different motivations for giving than Western donors. Expressly, Asian charitable giving tends to be private, personal, and small while Western donation practices are often public, large, and professional (Tsunoda, 2010). These differences may be based on the fact that ethnic Asian populations, especially those who are foreign-born, have a strong sense of cultural and political views influenced by the unique context of their respective home countries (Pew Research Center, 2013), which implies giving patterns to Asian-American nonprofits may vary based on these differences.
Regarding such cultural and political views, an important concept that should be understood is political trust, which is an indicator of political legitimacy and is vital to regime stability, particularly in Asian societies (Wong et al., 2011). Political trust indicates citizens' perceptional confidence in their government; such political institutions are evaluated by people's levels of trust toward their political system based on their normative expectations (Wang, 2016). A critical factor that is essentially and positively associated with the level of political trust toward government in Asian societies is economic performance (Wang, 2016). In some Asian countries, such as China, Singapore, Taiwan, South Korea, and Japan, political trust is closely related to and interdependent with social trust (Lee & Yi, 2018;Wang, 2016), which refers to a generalized trust that people will uphold shared social norms and rules defined by one's society to create societal stability (Vallier, 2019). Political trust influences people to be more likely to trust any government-sponsored support (Bosacianu et al., 2013).

The Current Study
A clear grasp of the nuances associated with enhancing financial resources is especially critical for Asian-American nonprofits due to the small scale of their fiscal budgets. Related to this, applying crowding-out effects to the organizations will provide a possible way to examine how much these organizations diversify their fiscal resources and depend on those resources. The present study scrutinized the relationship between government grants and private donations among the three East Asian-American nonprofit organizations (Chinese, Korean, and Japanese nonprofits) in the New York and New Jersey Metropolitan area, which is home to the second largest Asian population in the U.S after California (Pew Research Center, 2013). We targeted East Asian nonprofit organizations in the New York and New Jersey Metropolitan area not only because prior studies on the profile of nonprofit organizations focused on major urban areas, including New York (Wolpert et al., 2003, as cited in Hung, 2007, but also because more than half of Asian nonprofits are located here (Hung, 2007).
The main reasons why we focus on comparing these three groups are as follows: First, as previously explained, East Asian groups made up most Asian-American nonprofits (Hung, 2005). Second, it should be noted that understanding these ethnic groups is complex since both homogeneous and heterogeneous characteristics are embedded in these ethnic groups. For example, these groups are homogenous, in that they tend to share Confucian traditions and similar immigration histories in the U.S.; the number of immigrants from these national origins grew rapidly since national origin quotas were eliminated by the Immigration Act of 1965 (Okamoto, 2003;Zong & Batalova, 2017). On the other hand, they are also heterogeneous, in that the groups had antagonistic relationships in the past (e.g., Korean and Chinese historical antagonism toward Japan due to the brutal nature and consequences of Japanese imperialism), which still continues to affect each other (Okamoto, 2003). With this in mind, there are intergroup dynamics within each ethnic boundary as well as intragroup dynamics within a larger, panethnic Asian-American group boundary. Independently analyzing the giving pattern of each ethnic group, thus, is necessary to essentialize differences among them. Due to these reasons, our goal was not to generalize the findings, but rather identify possible variations in the relationship between government funding and private giving among three different ethnic communities.
Based on the assumptions of the crowding-out effect theory and prior studies on this topic (Alpizar et al., 2014;Andreoni & Payne, 2003;Andreoni et al., 2014), we formulated the following research hypotheses: Hypothesis 1 (H 1 ): The amount of private giving for East Asian-American nonprofits will vary when the agencies receive government funding.
Hypothesis 2 (H 2 ): The relationship between government funding and private giving will differ according to ethnic group.

Data and Sample
Financial information was collected from Form 990, filed by the East Asian-American nonprofit organizations using the GuideStar database, which provides key financial information on charities' performance in a simple way. From GuideStar, the most recent three years of Form 990 were collected for this study, going back from 2016. However, not all agencies provided the forms from consecutive years (i.e., [2014][2015][2016]. That is, some agencies reported non-consecutive years, thus, the total range of years that were reported in the database varied from 2007 to 2016. We listed the organizations with Chinese, Korean, or Japanese in their names to identify the three East Asian-American nonprofits. Organizations that did not identify their target populations in their names were not included in this study. Since there was no dataset with only Asian-American nonprofits, we believe that this was a simple and efficient method to sort out our target samples. In the results, a total of 1,481 East Asian-American nonprofit organizations were listed. Among the listed organizations, 1,071 were ruled out because the Form 990 was not reported. For the descriptive analysis in this study, a total of 410 East Asian-American nonprofit organizations were analyzed.

Measures
The dependent variable was the total amount of donations reported in each organization's Form 990. The independent variable was calculated with total government grants as reported in the subsection of "Government grants" under the Part VIII Statement of Revenue in Form 990. In calculating total government grants, we excluded the reimbursements for providing services by the organization under government contracts, as suggested by Andreoni and Payne (2011). Andreoni and Payne (2011) affirmed that the reimbursements for service provisions by the government should be excluded because it covers any payments received by the nonprofit for the services provided. These types of payments are reported on a nonprofit's tax return under program service revenue. In our analysis to estimate the crowding-out effect, we used the lagged variable of the government grants as an explanatory variable to correspond to a possible endogeneity. In the econometric model, the lagged variable is used to predict the current value of the government grant based on previous value. Achen (2000) suggested that lagged variables should be controlled because of the bias of exogenous variables in the panel data, especially for those with a small number of time-points. In the literature on crowding-out effects, the bias caused by omitted variables is considered a problematic issue for accurately estimating crowding-out effects. This could be solved by including the lagged government grants, even though the lagged variable model comes with new assumptions (Wit & Bekkers, 2017).
We also included some key control variables in this model to control for endogenous effects that may influence the private giving variable. They were other types of revenue, location, and types of subsector. Other types of revenue included all income categories in the "Other Revenue" section under Part VIII in Form 990. Location was defined as either "New Jersey" or "New York." Type of subsector was identified based on the National Taxonomy of Exempt Entities (NTEE) code as reported on Form 990. Many were categorized as public and societal benefit organizations, such as the Chinese Association for Science and Business, Inc. and the Korean-American Medical Association, Inc. Others were categorized as art & cultural organizations or human service organizations, such as the Korean Community Center, or Japan Performing Arts, Inc. Based on the self-identification of the type of subsectors, all organizations were recategorized as either religious or secular in the current study. Religious organizations were agencies that self-identified as religious organization for their NTEE code in the Guidestar database. Secular organizations were agencies for all other categories.

Data Analysis
To examine the crowding-out effect between government grants and private giving, panel data analysis was conducted. By using the Consumer Price Index (CPI) inflation calculator provided by the U.S. Bureau of Labor Statistics, the collected data was adjusted to 2016 dollar values. Again, the lagged variable for government grants was generated for the regression analysis to control for the spurious correlation between government grants and private giving. Prior to estimating the crowding-out effect using panel analysis for this study, a Hausman test was conducted to determine whether a fixed or random effects model was appropriate (Wooldridge, 2002). The Hausman test is used to differentiate between a fixed effects and random effects model in panel analysis. If the statistical significance level of the Hausman test is greater than .05, it indicates a random effects model is appropriate; if it is less than .05, a fixed effect model is indicated (Greene, 2017). In our data, the Hausman test results indicated that a random effects model was appropriate for our analysis, χ 2 (2) = .50, p > .05. In the random effects model, each intercept is a result of random deviation from some mean intercept.

Handling Outliers
To accurately estimate the crowding-out effects, it was essential to address outliers. To do this, first, we calculated the total amount of revenue by summing up government grants, private contributions, and other investments. Then, we generated outliers from the total amount of revenue using the median absolute deviation (MAD), known as a more robust method for detecting outliers (Leys et al., 2013). The MAD is useful, particularly when the data distribution violates the normality assumption (Adekeye & Azubuike, 2012;Leys et al., 2013). The MAD was computed based on the following formula: , where = Performance value for period i; ̅ = Average value; and n = Number of data Using the formula, we used conservative criteria that any total revenues exceeding 95% (equivalent to three standard deviations from the mean) of the MAD were considered to be outliers (Leys et al., 2013). With the criteria, we found 47 outliers, which were excluded to estimate the crowding-out effects in this analysis (Chinese = 22, Korean = 12, Japanese = 13).

Crowding-out Effects
The crowding-out effects between government grants and private giving for East Asian-American nonprofit organizations are presented in Table 3. As stated previously, we used the random effects model to estimate the crowding-out effects after removing the 47 outliers generated by the MAD method. Given the fact that some variables are non-normal, we estimated the random effects model with robust standard errors. In column (1) we reported the coefficients on private giving for overall East Asian-American organizations using a robust standard error. In columns (2), (3), and (4) we reported the estimated coefficients for each ethnic nonprofit. Table 3, for overall East Asian-American nonprofits, when the organization increased government grants by $1.00 in the previous year, donors were significantly likely to increase their donations by approximately $0.07 in the current year (hypothesis 1; b = .068, p < .01). Regarding hypothesis 2, the estimated crowdingout effects varied by each ethnicity. As presented in Table 3, there were significant crowding-in effects for Chinese and Japanese-American nonprofits. Specifically, private donors for Chinese-American nonprofits significantly increased about $0.07 of their donations in the current year when the charity received government grants in the previous year (b = .074, p < .05). For Japanese-American nonprofits, significant crowding-in effects were also estimated, suggesting that each $1.00 increase in government grants in the previous year was associated with an increase in of $0.09 in private donation in the following year (b = .088, p < .001). For Korean-American nonprofits, however, government grants increasing by $1.00 in the previous year was associated with a decrease in private giving of $0.53 one year later, which was not statistically significant (b = -.529, p > .05).

Discussion
To the best of our knowledge, this pilot study is the first to examine the association between government grants and private giving to East Asian-American nonprofit organizations using publicly available data. Thus, our main purpose was to contribute to the sparse literature on culturally-specific nonprofits and to develop possible suggestions for strengthening their financial management. Our findings indicated government grants for East Asian-American nonprofit organizations crowded-in private donations to the organizations. Despite the estimated crowding-in effect being small, the results are supported by other studies reporting crowding-in effects (Heutel, 2014;Hughes et al., 2014). Our findings are consistent with a hypothesis of authoritarian orientations and political trust in East Asian communities. Providing credibility or prestige to the nonprofit organizations with government partnerships is stronger in Asian countries. According to Chong (2011) and Ma and Yang (2014), overall levels of political trust in governments in Asian countries remain higher than that of non-Asian societies. Indeed, the values of loyalty (to the king or nation) and filial piety (to the parents) are basic social norms of social behavior for East Asians born and raised in Confucian culture and societies. Within Confucian culture, a father in a family has absolute power to rule the family and the relationship between the government and individuals is seen as the extension of the father-son relationship (Ma & Yang, 2014). Within this context, government funding in general may serve as a signal of quality with respect to the East Asian-American nonprofit organizations' abilities to provide services. In other words, government grants may imply a signal of trust for the charity for the donors, thus they donate more to their charity.

Mixed Results: Crowding-in/out Based on Ethnicity
Another finding in this study is the estimated relationship between government funding and private giving being different for each East Asian-American ethnic nonprofit organization. Government grants significantly lead to increased private giving to both Chinese and Japanese-American nonprofits, but an increase of government grants crowded out private giving for Korean-American nonprofit organizations, though this finding was not statistically significant. Although these three ethnic organizations have relatively long histories in the U.S. compared to other Asian groups (Hung & Ong, 2012), we identified few English-language studies focusing on these nonprofit organizations, which limit our ability to assess our findings. Therefore, to understand such differences among Chinese, Japanese, and Korean-American nonprofits, it would be helpful to expand our explanations to the context of nonprofits in each home country. The context relating to the different environments in the home countries allows us to gain insights regarding the different patterns of each ethnic donor's giving behavior in the U.S.
The estimated crowding-in effects with Chinese-American nonprofit organizations may stem from the nature of Chinese donors' motivations. According to Wang (2014), interpersonal relationships and a strong sense of self-reliance are enormously significant for Chinese donors. Chinese individuals who believe they should play an important role in charitable giving rely on strong values of self-reliance. Government is the foundation of self-reliance in China because the government has played a leading role in charity (Han, 2016;Wang, 2014;Yang, 2012). Crowding-in effects are also featured in the Chinese nonprofit context and history. Until 1978, government attitudes toward nonprofit organizations in China were negative, so the Chinese government had prohibited private nonprofit involvement from providing social services (Yuanfeng, 2016). Even though the attitude has changed from distrust to relative trust and from restriction to encouragement, the Chinese government emphasizes the government and nonprofit corporations to improve their legitimacy and trustworthiness (Yuanfeng, 2016).
Within this historical context, Chinese nonprofits are hardly fully private because many of the nonprofits in China are formal branches of state institutions. Government involvement in nonprofit partnerships are, therefore, one of the characteristics of nonprofit organizations in China, due to the emphasis on state dominance (Yuanfeng, 2016;Zhang et al., 2011). Also, the public in China has emphasized transparency and accountability toward nonprofit organizations since charity scandals occurred after the earthquake in the Wenchuan area in 2008 (Yuanfeng, 2016). For the transparency of nonprofits in China, governmental or government-sponsored foundations in China only are able to raise funds from the public according to the Regulation on the Administration of Foundation in 2004 (Zhang et al., 2011). Within the long tradition of government control of all social organizations and historical background of nonprofit organizations in China, Chinese donors recognize the government grants to their charity as signals of quality or signals of transparency, so they increase their donations to the charity.
Our findings also significantly confirmed government grants increased private contributions among Japanese-American nonprofits. Similar to the explanation of Chinese-American NPOs, it is presumed the Japanese-American crowding-in phenomenon might be also derived from their perspective of the government. There may be two possible explanations stemming from two different settings. Understanding Japanese and Japanese-Americans' different views toward nonprofits might be helpful in interpreting our findings.
The first explanation is related to Japan's domestic standpoint. Historically, state power has been also more pervasive than civil society in Japan (Akira, 2006). Thus, Japanese people have largely sought support from the government when they face difficulties, rather than receiving assistance from civil society organizations (Hernon, 2016). In Japan, legitimate nonprofits hardly existed until the Act on Promotion of Specified Nonprofit Activities was passed in 1998 (Akira, 2006;Hernon, 2016). In spite of their growing numbers, nonprofits in Japan remain in the beginning stages (Hernon, 2016). The biggest challenge they face is the lack of trust from the general public toward nonprofit organizations due to the unclearness of the organizations' mission and structure (Hernon, 2016). If nonprofits in Japan fail to show their clear vision and mission on who they are and what they do, it is hard to receive private contributions from the general public (Hernon, 2016). Given this home country context, it is likely that receiving government grants may signify the organization's clear mission and trustworthiness, ultimately leading to a crowding-in effect.
For the second explanation, we approached historical activities committed by Japanese-American nonprofit groups, particularly the Japanese American Citizens League (JACL). Although Japanese who first immigrated to the U.S. before the second World War tended to show extraordinary loyalty to their new home country (JACL, 2011), the internment of Japanese-Americans started during World War II and even continued after the war. The JACL has historically played an important role in seeking redress for the internment of Japanese-Americans (Kitano & Maki, 2003), which finally led President Ronald Reagan to sign the Civil Liberties Act authorizing redress payments to Japanese-Americans in 1988 (JACL, 2011;Kitano & Maki, 2003). A series of historical achievements by both nonprofit organizations and the U.S. government might project a positive image of the cooperation to Japanese-Americans. Borrowing from the Kitano-Maki Proper Alignment model (Kitano & Maki, 2003), receiving government grants signals to Japanese-Americans "the positive alignment of the mainstream society" (p. 418), which may lead them to give more private contributions to nonprofits. However, these explanations should be cautiously interpreted due to the relatively small sample size (n = 69), which may have affected the power of the analyses. Future study needs to replicate this with a larger sample of Japanese non-profits to draw more robust findings.
The impact of government funding was negative for Korean-American nonprofit organizations. Even though it did not statistically support the crowding-out hypothesis, the crowding-out effect may be understood by Korean donors' motivations for charity and attitudes toward non-governmental organizations in Korean society. Korean immigrants are more likely to help people within their ethnic group to maintain their traditional social norms based on Korea's Confucian traditions (Moon et al., 2015). Thus, people in the Korean community in the U.S. who have stronger normative motives are more likely to donate to Korean charitable organizations than to mainstream organizations (Moon et al., 2015). Also, individuals in Korea are not motivated in their donation behavior by government policy changes on charitable giving because government involvement in charitable giving in Korea has not been well-established (Hur, 2006). Therefore, Korean donors may negatively view government involvement in ethnic nonprofit organizations in the U.S. in order to sustain their origin of traditional social norms.
The negative impact of government funding on private giving to Korean-American nonprofit organizations may be related to "too close for comfort" (Kim, 2009, p. 888), describing the relationship between nonprofits and the government in their home country. Nonprofit organizations used to be considered the most trustworthy institution than any other domestic and foreign institutions including government agencies in Korea. However, the trust was weakened when the government and nonprofit relationship grew stronger in Korea. According to Kim (2009), increased government involvement in non-government organizations such as financial and political support led to the decline of NGO's institutional trust in Korea. From 1993 to 2003, the Korean government was devoted to increasing the government partnership with the nonprofit sector. As a result of the engagement, some of the major nonprofit organizations in Korea have received government funding for their activities and services. When nongovernment organizations in Korea have received government funds, serious questions about conflicts of interest and the groups' sincerity were raised. Consequently, nongovernment organizations' institutional trust in Korea ranked as first in 2003 but dropped to fifth place in 2005 since the Korean government drove strong support for the nonprofit sector. Within these contexts, it may be plausible Korean-American donors view government involvement as weakness or lost independence and autonomy, so they decreased their donations to their Korean charities in the U.S.

Limitations
The data used in this analysis are by no means without shortcomings. First, it should be noted that a relatively small percentage of our sample (21%) received any government grants. Approximately three-quarters of the East Asian-American nonprofits in the current study have not received government funding at all. This indicates securing an adequate number of qualified subgroup samples (i.e., nonprofits receiving government grants) is critical to assuring more valid findings.
Second, our findings could not provide detailed information regarding the relationship between different types of government funding and private giving because Form 990 data only displays aggregated amounts funded from local, state, and federal governments. Related to this, Steinberg (1987) and Kim (2014) demonstrate the impact of government grants on private giving may vary depending on the type of government grants. Therefore, as Steinberg (1987) suggested, future studies should use more rigorous data by combining individual aggregated data of the amount each donor has contributed to a charity with individual organization data reflecting each type of government grant through annual or audit reports. Methodologically, this also implies the joint crowding-out effect may provide more rigorous results than the simple crowding-out effect.
Related to using Form 990, other covariates that would potentially be correlated with the relationship could not be adjusted for due to data limitations. For example, agency level data such as staff hours and/or administrative decisions on fund-raising efforts and staff capacity to write compelling grants are important factors that should be included. Future study needs to integrate these factors to accurately and empirically portray the relationship.
Lastly, our method is solely based on a quantitative approach. Given inconsistent findings of the association reported in the literature review, it might imply that a solely quantitative approach would not be an effective research design to measure the relationship and to reflect the respective ethnic perspectives of each organization. Mixed methods will allow for the identification of some critical contextual factors related to fund-raising, but are not addressed in this study, including the effectiveness of agency services and programs, relationship with various levels of government and funders, and current government priorities for funding. In future studies, therefore, using mixed methods to reflect the realities from each ethnic community (or stakeholders such as donors, agency staff) is necessary to ensure a more accurate interpretation of the findings.

Conclusion
Examining the crowding-out effect has contributed to understanding the relationship between government funding and private giving, which is an important aspect of effective financial management for nonprofit managers. Although our findings should be cautiously interpreted due to limitations derived from the data and research design, our main conclusion is that government grants crowd in private giving in East Asian-American nonprofits in the New York and New Jersey Metropolitan area. However, our findings also reveal the complex nature of giving patterns depending on ethnicity when it comes to the mixed results of -in (e.g., Chinese and Japanese-American) and crowd-out (e.g., Korean-American) effects. Such differences may possibly be explained by the different political and cultural environments surrounding nonprofits in the respective home countries. Each ethnic nonprofit organization in the U.S. would need to establish different strategies to utilize such information (i.e., receiving government grants) to help shape giving patterns. Obtaining adequate insight about the association between government grants and private donations will allow Asian-American nonprofits to pursue better planned and more efficient fund-raising efforts.