Does ethnic diversity affect public goods provision? Evidence from boundary reform of local governments

ABSTRACT

This study explores how ethnic fractionalization affects the provision of public goods using Japanese city data, and attempts to address potential endogeneity by using boundary reforms of local governments. The results show that an increase in ethnic fractionalization has a significantly negative effect on spending on productive public goods (roads and bridges) and positive effects on spending on nonproductive public goods (sanitation and trash collection). However, ethnic fractionalization is unrelated to the provision of other local public goods such as education and welfare.

KEYWORDS:

• Boundary reform
• ethnic diversity
• fractionalization
• local public goods
• segregation

Acknowledgement

I greatly appreciate Simon Lapointe, Xiang Ma, Maria Metzing, Masaaki Suzuki, Hiroki Tanaka, Fan Wang, Geoff Whyte, and participants in the seminars at Doshisha University and VATT and in the 2017 Asian Meeting of the Econometric Society and the 73rd Congress of the International Institute of Public Finance for their helpful and insightful comments. This study was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant Numbers 16K13370; Grant Numbers 19K01697).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ A simple theoretical model explains the relationship between ethnic heterogeneity and the provision of public goods. Consider a country where ethnically different types of individuals reside. If the marginal utility of a public good decreases greatly when the ethnic diversity of the population increases (or if sharing the use of a public good with other ethnic groups reduces the marginal benefit of public good consumption), a benevolent government will choose a lower level of the public good as ethnic variety increases (Alesina and La Ferrara 2005). If public spending is divided into expenditure on productive (or pure) and nonproductive public goods, an increase in ethnic diversity will lead to a lower share of spending on productive goods and a higher share of spending on nonproductive goods (Alesina, Baqir, and Easterly 1999).

² As for the dispute over Alesina, Baqir, and Easterly's (1999) results, Beach and Jones (2017) state that “(h)owever, recent empirical work finds that this relationship is not robust when estimated in a panel (Boustan et al. 2013; Hopkins 2011) or with additional controls (Gisselquist 2014)”; BriBrian, Levy, and Hero (2018) also describe that “(w)hile Alesina, Baqir, and Easterly (1999) report fairly robust results, others (e.g. Hopkins 2011; Boustan et al. 2013) have not corroborated their findings.”.

³ In addition to the effect on cost reduction, the literature examines the impact on population growth (Suzuki and Sakuwa 2016).

⁴ Many studies have attempted to clarify the determinants of mergers, specifically in Japan, with candidates including the common pool problem affecting local bond issues (Blom-Hansen, 2020; Hirota and Yunoue 2017), political factors (Yamada 2016, 2018; Yamada and Arai 2020), and intergovernmental transfers (Weese 2015).

⁵ Details for each municipal budgetary item are reported in Table A1 in Appendix A.

⁶ A detailed explanation of the decision-making process of merger is as follows. When a municipality seeks to merge, its legislature votes for the establishment of a merger consultation committee. Following discussions by this committee, if the relevant municipalities decide against a merger, the committee is disbanded, but if they agree to merge, the committee continues to pursue a merger under the proposed scheme. In many cases, the mayoral view toward the merger strongly influences the municipality’s attitude because the mayor has a strong power in determining municipal policy under Japan’s local public administration system.

⁷ The numbers of municipalities and mergers from 1999 to 2014 are demonstrated in the figure in Appendix B.

⁸ The number and share of foreign residents in Japan from 1970 to 2015 are illustrated in Figure C1 in Appendix C.

⁹ For details regarding foreign residents in Japan, see Appendix D.

¹⁰ Until July 2012, foreign residents were defined as foreigners who had registered, either legally or illegally, with the Japanese government and had lived in Japan for more than 90 days. Since July 2012, the definition of foreign residents has been restricted to those who are legally registered and have lived in Japan for more than 90 days.

¹¹ Weiner (2008) illustrates how homogenous Japan is by demonstrating that the minority groups, regardless of their specific ethnicity, have suffered from differences between them and Japanese people in terms of culture and the persistent prejudice and discrimination that has resulted.

¹² As discussed in Alesina, Baqir, and Easterly (1999), non-excludable, or pure, public goods are “productive” in the sense that they are less likely to be provided for special interest groups. In contrast, public expenditure targeted toward certain groups is deemed to be spending on so-called “nonproductive” or “patronage” public goods.

¹³ The units and sources of all of the public goods variables are presented in Table E1 in Appendix E.

¹⁴ The numbers of mergers by years since merger as of 2010 are provided in Table F1 in Appendix F.

¹⁵ See Appendix G for details on the local election system for merged municipalities.

¹⁶ It should be noted that foreigners in Japan do not have suffrage for national and local elections. Only those of Japanese nationality can vote in Japan. With the considerable surge of foreign residents over decades, Japan’s local governments have engaged in the provision of a wide range of public services for foreigners, including promoting city planning aligning with foreigners’ ideals, offering public lectures to teach Japanese to foreigners, and hosting various events that encourage better understanding of and cooperation with foreigners. Many prefectures and large cities are concerned about raising funds and retaining staff for projects associated with foreign residents (e.g. Kashiwazaki 2003; Japan Center for International Exchange 2018). Therefore, despite not possessing suffrage, it is likely that foreigners do influence the local public services provided by the municipalities in which they live.

¹⁷ The selection model will be explained later.

¹⁸ Specifically, the following three motivations toward merger are mentioned. First, given the introduction of the Uniform Decentralization Law in 2000, which was aimed at promoting decentralization of local governments, and with the advent of urban sprawl, municipalities with small populations and revenues were implicitly encouraged to enter into a merger to maintain their independence. Second, given Japan’s declining birth rate and rapidly aging population, localities that anticipated a reduction in population in near future were expected to seek a merger. Third, given that the Japanese government had experienced significant fiscal deficits since the late 1990s, local governments were requested to promote administrative reforms by entering into mergers.

¹⁹ Fiscal year is used for the calculations of budgetary items. Sasayama city, which was created in 1999 by merger, was excluded from the sample; ordinance-designated cities, which are mega-cities in Japan, basically with a population of more than one million, are also excluded because they are required to provide a much wider range of public services than other cities.

²⁰ To see the association between differences in ethnicity variables and spending shares, I calculated the correlation coefficients between them. It is shown that spending on sanitation and trash collection are positively correlated with the fractionalization index. The correlation matrix is presented in Table H1 in Appendix H.

²¹ The list of merged cities with their merger dates is provided in Table I1 in Appendix I.

²² represents the time difference of $X$ between 2000 and 2010.

²³ Note that the findings obtained are clear, although foreigners do not have suffrage and cannot directly influence local policies through voting.

Additional information

Funding

The work was supported by the JSPS [16K13370,19K01697].