Decentralised Governance. Crafting Effective Democracies Around the WorldBy Jean‐PaulFaguet and SarmisthaPal (Eds.), London: LSE Press, 2023. To Appear in Public Administration Review. ISBN: 978‐1‐909890‐84‐8
本书综述21世纪分权研究的最新发现，聚焦发展中国家地方政府的福利、服务交付和政治问责后果，适合学生、学者及政策制定者参考。
What have we learned from decentralization research in the 21st century? The main goal of this edited volume is to present and organize recent findings in a way that is accessible to students and informative to scholars of decentralization and local government. In doing so, it updates Bardhan and Mookherjee's earlier edited volume, Decentralization and Local Governance in Developing Countries: A Comparative Perspective (2006). Like its predecessor, the chapters in this edited volume focus on local or third tier governments as opposed to regional or second tier governments. Both volumes focus on the consequences of decentralization for welfare, service delivery, and political accountability rather than on the causes of decentralization. Finally, both volumes are concerned with the Global South and are written by leading scholars. These delimitations of scope facilitate the volume's coherence, and there is good reason for the exclusion of countries from the Global North. Existing research shows that decentralization is usually efficient in OECD countries (Hooghe and Marks 2009; Hooghe and Marks 2016). Public goods and services are usually provided at a jurisdictional level that is sufficiently homogeneous—such that policy preferences are sufficiently similar—and yet large enough to realize economies of scale in their provision. By contrast, decentralization is more inefficient in countries from the Global South. One reason is that many of them are less democratic. Another is that central governments typically bypass regional or second tier governments in favor of local or third tier governments for political reasons: It is often less politically risky for incumbents to decentralize at the local level even if they sacrifice the economies of scale that regional governments could provide (Ricart-Huguet and Sellars 2023). Another reason for this lower efficiency of decentralization in the Global South is that governments have undertaken much more political than fiscal decentralization (Chapter 4). After half a century of decentralization reforms, local elections have become ubiquitous but the power of the purse typically remains in the hands of the central government. There are actually some good reasons for a centralized tax structure. For one, it favors efficiency, conditional on sufficient jurisdictional homogeneity. For another, it limits tax competition and thus a potential “race to the bottom” among local governments if they tried to tax mobile assets. Land is an immobile asset and thus land taxes are immune to this problem (short of bribing local officials, one either pays the tax or relocates). Capital, however, is mobile and a decentralized tax structure could be detrimental to public revenue. This is one reason capital gains and corporate taxes are levied by second or first-tier governments. (Capital also moves between countries, of course, and that is why many countries recently agreed to a global minimum tax of 15%.) The less enlightened justification for a centralized tax revenue structure is that it allows the central government to use the power of the purse for partisan reasons. Indeed, political decentralization in the Global South has not been accompanied by fiscal decentralization even in cases where it would have been efficient (Chapter 4). As we know from much research in political science, decentralization can be sequential for strategic political reasons such that political decentralization precedes or follows economic decentralization (Falleti 2005; Boone 2003). In short, strategic and particularistic political motives such as reelection and local political control, rather than tax competition, often explain why central governments limit local fiscal decentralization in much of the Global South. Some decentralization reforms suffer from those motives more than others, making it difficult to determine whether a reform fails for technical or for political reasons (Green 2010; Grossman and Lewis 2014). Partly as a result of these complex political economy incentives, existing literature presents mixed findings on the efficiency and welfare effects of decentralization. For example, some reforms improve local service delivery but others increase local elite capture. However, Faguet underscores that not all empirical studies are of equal quality (Chapter 2). He finds that, on average, “higher-quality evidence indicates that decentralisation increases technical efficiency across a variety of public services, from student test scores to infant mortality rates” (Chapter 2, 26). Another reason for the mixed findings is definitional. Faguet observes that some scholars confuse the terms “deconcentration” and “delegation” with “decentralization proper,” which he defines as “the devolution by central (that is, national) government of specific functions, with all of the administrative, political and economic attributes that this entails, to democratic local (that is, municipal) governments that are independent of the center within a legally delimited geographic and functional domain” (25). This definition of decentralization rightly reserves the term for actual devolution of power, which deconcentration and delegation rarely entail. One quibble with this definition is that it restricts decentralization to democratic and local (third tier) governments, perhaps unnecessarily. “Decentralization proper” also takes place in semi-democratic local governments and in regional, provincial, or state governments (second tier). The other quibble is that decentralizing reforms may not necessarily devolve both political and economic functions. A subnational government may continue to raise and spend the same amount of public revenue but its legislature or local council become democratically elected following a political decentralization reform. To address these two quibbles, I would revise the definition thus: decentralization is the devolution by the central or federal government of specific political functions, economic functions, or both, to one or more lower tiers of government that are autonomous from the center within a legally delimited geographic and functional domain. This amended definition also ensures that purely administrative reforms are excluded, and in my reading the volume is consistent with this amended definition. The edited volume covers much more ground than this book review can. Chapter 2 alone is a tour de force that organizes and evaluates what we already know and what we should know more about in the near future. In the following chapter, Mookherjee reviews studies showing that technology (e-governance) and the removal of local discretion in favor of formula-based transfers reduce corruption and clientelism, two of the recurrent pitfalls of local decentralization programs in non-democracies but also in patronage-based democracies. The second part of the volume consists of country-specific chapters that examine the politics of decentralization. The case of Pakistan reveals the dangers of “devolution under autocracy” (Chapter 5). Malik et al. argue that, under Zia's military regime, party politics gave way to family politics and that this is why political dynasties are prevalent in Pakistan today. In Chapter 6, Martinez-Bravo et al. show that centrally-mandated local elections in China increased local public goods provision in religiously homogeneous villages but much less so in heterogeneous ones. The findings build upon existing evidence from many other settings, ranging from the United States to African countries, that local social heterogeneity limits the gains from local decentralization. The introduction of local elections in Indonesia suggests that homogeneous villages are more likely to select local leaders by consensus while ethnically diverse ones elect them by majority rule (Chapter 7). This is what one would expect because election by consensus is only practical in small and homogeneous communities. Counterintuitively, Mitra and Pal also find that more diverse villages increased their income and development spending following a country-wide fiscal decentralization reform. The authors attribute this result to more political entrepreneurship in diverse villages. Lastly, Mbate finds that copartisanship between parliamentarians and local politicians weakens accountability in Kenya but that this effect is much weaker under more competitive local elections (Chapter 8). The last part of the book concerns more technical aspects of decentralization, specifically how mechanism design might promote public goods such as efficient service delivery and lessen public bads such as corruption. That is the goal of Chapter 9 by Afridi et al., which models the trade-off between centralized monitoring and decentralization or community monitoring. Chapter 10 by Chachu et al. uses detailed data from Ghana to construct composite indices of local government performance, which again should facilitate monitoring, while Chapter 11 by Shonchoy and Wahajj documents the importance of birth certificates in Bangladesh, one of many countries in South Asia and Sub-Saharan Africa where registering births is not a widespread practice. This informational problem should of course be alleviated if decentralization is to address local needs. Finally, Chegwin et al. examine the consequences of a 2001 education reform in Colombia that granted autonomy in education policy to municipalities larger than 100,000 inhabitants (Chapter 12). The authors exploit this discontinuity to show that decentralized municipalities increase education expenditures and education quality because resources become better allocated, pointing to the importance of local information, knowledge, and autonomy. The volume will prove highly informative to students and scholars of political economy, public policy, and public administration, even if they are not scholars of decentralization. It will also be very useful to policy-makers and practitioners. Many of the authors are academics but others are practitioners or both, and the volume is all the better for it. The author declares no conflicts of interest.