What is “quality of government” (QoG)? How much does it matter, and why? And—a crucial practical issue for a world in need and upheaval—how can it be improved?

To address these questions, Bo Rothstein deploys philosophy, statistics, psychological experiments, and history. He reviews what political scientists have written, adds a touch of public administration, and occasionally refers to the work of economists.

Citing John Stuart Mill, John Rawls, and Brian Barry's Justice as Impartiality (1995), Rothstein boils down good government to impartiality, which is different from, and conceptually preferable to, democracy, the rule of law, and efficiency/effectiveness. Democracy is rejected because “it is not difficult to find examples where a majority of voters support corrupt politicians, clientelism, and outright discrimination against ethnic minorities” (p. 13). For Rothstein, political legitimacy is less about democracy than about the government's impartiality in implementing policy (p. 92). The impartial administration of public programs, he argues, leads to better implementation, with the result that those programs are better funded, more popular, and more effective—and this in turn leads to more trust and well-being.

Health is one example: “The main finding is that ‘good governance’ is a critical factor in making national health care systems work, and that public spending on health care is inefficient with low QoG” (p. 65). Social welfare is another. Among 18 countries of the Organization for Economic Co-operation and Development, higher QoG is positively related to the size and generosity of the welfare state and to the effect of working-class mobilization on welfare state generosity.

Conversely, bad government causes many social evils: “Since social trust is an important intrinsic value (personal happiness, optimism about the future) and also has a political value (support for fair institutions, minority rights, tolerance, etc.) and an economic value (its positive relation to individual earnings and aggregate economic growth), it may be that dysfunctional government institutions are the worst social ill of all” (p. 162).

How does one empirically assess these causal chains? After philosophically differentiating impartiality from the usual concepts, in his statistical work Rothstein conflates impartiality with the usual measures. It turns out that that many measures of governance are highly correlated. For example, the bivariate correlations among the World Bank's Rule of Law Index, its Government Effectiveness Index, and Transparency International's Corruption Perceptions Index exceed 0.9 (p. 67). In one analysis, Rothstein measures QoG with the mean of these three indicators; in another, he averages the International Country Risk Guides indicators for corruption, law and order, and quality of bureaucracy.

How much do what kinds of improvements in governance matter to what outcomes and under what conditions? This is a complicated question, for many reasons: no agreed-upon theory or model, incomplete and unreliable data, small N, heterogeneity among countries, and reciprocal causation. Alas, the book's statistical forays do not overcome these difficulties; frankly, the methodological discussion does not adequately countenance them (for example, the analysis of the welfare state in Chap. 6).

If the quality of government matters, how can it be improved? The solution is not more social capital in the sense of fostering voluntary associations: “Concepts like ‘community’ and ‘civil society’ have a nice gentle ring, but behind them may be exclusion, discrimination, spread of mistrust, and manufacture of hatred. The idea of a virtuous path from civil society to social trust to democratization and economic development may simply be wrong and may have led to seriously damaging social outcomes” (p. 191). Rothstein's suggestions? First, “[s]imply put, I argue that the low levels of trust and social capital that trouble many countries are caused by too little government action to reduce inequality” (p. 156). Second, “universal education programs, may be under the auspices of international aid organizations” (p. 163). No case studies or cross-country evidence are adduced to support these recommendations.

Rothstein criticizes principal-agent models that ignore the equilibrium qualities of corrupt systems. The law-and-development movement beginning in the 1960s may look like a QoG-through-impartiality movement, but it failed. Among the reasons, Rothstein says that the approach “left practitioners with little or no guidance on what types of reforms should be prioritized” (p. 57). Practitioners did not participate in designing the reforms. Informal institutions were excluded.

Rothstein correctly complains that too many political science studies describe “how the institutional systems in corrupt and noncorrupt countries differ” without addressing “which types of processes are likely to be successful for enacting such reforms” (pp. 105, 104). “If we are to establish a thorough picture of what can become a successful reform process, research should start to identify different agents' roles and interests” (p. 107).

Bravo—now, how to research these excellent points? Case studies? Rothstein contrasts Singapore and Jamaica to show that democracy is not necessary or sufficient for impartial government or “social well-being.” But he eschews studying “the policies against corruption and clientelism launched by Lee Kuan Yew”—about the prioritizing of reforms, the participation of practitioners, the informal institutions, the processes, and “the different agents' roles and interests.” He argues that “[t]o present such an idiosyncratic explanation is probably not very helpful from a policy perspective. Moreover, this type of explanation does not really count in the social sciences…. [O]ne then has to come up with an explanation for why this type of far-sighted political leadership seems so rare in developing countries” (pp. 205–6).

The author scans nineteenth-century Sweden: “The change toward a Weberian civil service came between 1860 and 1875 and was both ‘nonincremental’ and dramatic” (p. 114). And this, he concludes, is what poorly governed countries need today: “revolutionary changes in institutions [author's emphasis]” (p. 119). “Unfortunately, we know very little about how such a transition can be made” (p. 109).

Revolutionary changes in institutions are what Singapore accomplished, however. How? Through important improvements in incentives (positive and negative) and better information—the steps principal-agent models suggest. Moreover, busting a corrupt equilibrium and creating a new institutional culture required “frying big fish,” strong examples from the top underscoring strong statements of value, new institutions with symbolic as well as functional importance (Singapore's Corrupt Practices Investigation Bureau), and openness to international competition.

Other success stories exist about improving the governance of corporations, cities, ministries, and countries. From 1998 to 2005, Colombia improved from the 4th percentile on the Corruption Perceptions Index to the 56th percentile. Recent examples include Georgia and Qatar. Malaysia's Government Transformation Programme has already logged successes. But some scholars are not interested in “the idiosyncrasies” of particular reforms—the steps, the sequencing, the building of political momentum, the changes in information and incentives, and the involvement of business and civil society. Scholars may focus at a more distant remove, trying to explain why a country produces a progress-inducing leader or tosses off an oppressive regime now instead of then—the sort of explanation that would “count in the social sciences.”

Fair enough. But progress in improving the quality of government will also come from the detailed, theory-informed examination of real cases for suggestions about what works under what circumstances. We need to publish this research in the usual ways. And we may need to share it in new ways—not by telling practitioners and citizens what to do but by providing them with case studies, theoretical frameworks, and data about events that enable creative problem solving.