Precise international metrics and assessments may induce governments to alter policies in pursuit of more favorable assessments according to these metrics. In this paper, we explore a secondary effect of Global Performance Indicators (GPIs): Insofar as governments have finite resources and make trade-offs in public goods investments, a GPI that precisely targets the provision of a particular public good may cause governments to substitute away from the provision of other, related, public goods. We argue that both the main effect of the GPI (on the measured public good) and this substitution effect vary systematically based on the domestic political institutions and informational environments of targeted states. Specifically, we contend that both the main and substitution effects of GPIs should be largest for governments that are least accountable (opaque and non-democratic) and should be smallest for those that are most accountable. We illustrate the logic of these arguments using a formal model and test these claims using data on primary and secondary enrollment rates across 114 countries. We find that countries substitute toward primary (which is targeted by the Millennium Development Goals) and away from secondary (which is not), and that these effects are mitigated as accountability rises.

Are democracies more transparent than autocracies? Some would seek to answer this question by definition. The current secretary general of the OECD, Angel Gurria, identifies transparency as a key ingredient “necessary for the functioning of democracies.” Many people agree: transparency has become synonymous with democracy. As Shapiro (2003: 200) suggests, “democratic leaders can never be entirely free from a commitment to truth-telling.” One could argue that if a country lacks transparency, it is simply not a democracy.

But should we assume, by definition, that democracies inherently seek to disseminate more data than autocracies? Doing so would preclude important questions, such as whether electoral competition leads governments to provide honest assessments of the economy or whether it causes governments to hide the truth.

Indeed, their vulnerability to public disapproval through elections may make democratic officials more inclined to obfuscate or withhold information concerning the performance of their policies than their autocratic counterparts – who need worry less about public perceptions. Political scientists and economists, in fact, have offered theories and evidence that the policymaking of democratic governments is shaped by issues of information revelation. Importantly, democratic governments have incentives to obfuscate (see, for example, Beitz 1989; Busch 2000; Stasavage 2004; Kono 2006; Rejali 2007; Mani and Mukand 2007; Culpepper 2010; Harding and Stasavage 2014; Jacobs and King 2016).

Still, governments that survive by winning elections also have incentives to disclose economic data. The previous chapter reveals an association between transparency and investment. If investment improves economic welfare, which, in turn, influences elections, democratic governments may do well to provide transparency. In fact, Chapter 8 shows that the association between transparency and domestic investment appears stronger for democracies than for autocracies. This chapter builds on these results, investigating whether the economic advantages of transparency hold more value for democrats than for autocrats. We examine whether electoral contestation provides incentives for governments to disseminate data.

Note that this inquiry requires distinct conceptions of democracy and transparency, where one term is not directly defined by the other. As discussed in Part I, many definitions of transparency include explicit references to democracy, and vice versa. Any empirical relationship between such variables would, at least partly, constitute a tautology.

Transparency endures as a mantra for many policymakers concerned with improving life in the developing world. Asked how to improve governance in developing countries, the managing director of the International Monetary Fund, Christine Lagarde, has answered emphatically: “Transparency, transparency, transparency!” The international community has lauded efforts of some states, such as Bulgaria, Mexico, and South Korea, for increasing their historic levels of transparency, sometimes even heralding them as economic and political miracles. Other states, such as Argentina, Bangladesh, and Zimbabwe, have lost credibility due to their failure to report credible information about the state's economic activities.

Why are some countries more transparent than others? And what are the consequences – economic and political – of increased transparency? This book seeks to address these questions using an innovative measure of transparency. While we confirm some broadly held views, we also introduce new and surprising findings.

In essence, the book is about the role of the informational environment in political economy. Political action, such as voting or protest, requires coordination across individuals. Coordination, in turn, requires information – not just any information, but shared information. That is, mass political coordination requires information about actors’ higher-order beliefs: their beliefs about what other actors believe. The informational environment matters for mass coordinated political action.

The first puzzle we confront in our work is a measurement problem: how to gauge the transparency of economic information. nWe take advantage of something that usually stands as an obstacle to research: missing data. We observe that missing data are data. We use patterns of reporting data as a proxy for the publicness of credible economic information.

Using this measure, we find that more transparent countries attract higher levels of investment and that transparency stabilizes democracies. A rich informational environment facilitates economic decision-making and it enables voters to discriminate between good and bad leaders. When elections function better, democracies face fewer protests, enjoy greater legitimacy, and become more likely to survive. Not surprisingly, democracies tend to be more transparent than autocracies.

Opacity provides benefits for autocrats. A closed informational environment serves as ameans to obscure rent-seeking and other malfeasance, enabling autocrats to reap augmented financial rewards from their position of power. Nontransparency also hinders the coordination necessary for popular unrest and thus preserves the autocratic regime.

Yet, autocracies vary widely in the extent to which they disclose economic information to their publics. Some autocrats choose to disclose. Given the economic and political rewards of opacity, why would any autocrat choose transparency? We contend that they do so because transparency offers economic and political benefits too – even for autocrats.

As seen in the previous chapters, the dissemination of data encourages investor confidence, which boosts economic performance. If the consequences of transparency were purely economic, the trade-off would be clear – an autocrat would simply set a level of data disclosure where the marginal costs from transparency, owing to reduced opportunities for rent-seeking, match the marginal benefits, flowing from increased investment.

We have stressed throughout the book, however, that the logic of long-run political survival may weigh more heavily in the calculus of an autocrat than shorter-term material gain. To the extent that transparency causes mass-unrest, which can lead to the ouster of an autocrat, dictatorships should eschew the dissemination of economic data.

What if, however, transparency acts as a double-edged sword politically? Autocrats face many risks to their survival beyond mass unrest, and these other risks may respond quite differently to transparency. Indeed, the very prospect of the mass unrest caused by transparency may reduce the risk of other real threats to the survival of an autocrat.

Consider the interests and desires of members of a leader's inner circle: this elite may include potential rivals for the leadership. Usually loyal allies, inner-circle elites sometimes seek to overturn their leader through a coup. Recall the example of South Korea's President Park Chung Hee, who was assassinated by someone within his clique – his own intelligence chief (Adesnik and Kim 2008: 1). Rival members of the elite can successfully take advantage of such opportunities only under certain circumstances. If masses of opposition groups have coordinated, they may seize on a moment of regime disunity to strike out for regime change.

Survival – as the goal of political leaders – has become a canonical assumption in political science. Just as economists assume that individuals seek to maximize the utility associated with consumption (and firms, profits), political scientists often assume that governments seek to survive in power. Whether they do so by winning elections, by paying off an inner circle of elites, or by quelling the masses through fear,we assume that all governments share this common goal of surviving in power. There are, of course, exceptional leaders throughout history who have voluntarily stepped down from power, including Presidents George Washington (1797) and Nelson Mandela (1999) (see Cogley 2013). We treat such exceptional leaders as deviations from the rule.

How does transparency impact the survival of political leaders? We contend that the answer depends crucially on political institutions. As argued by Bueno de Mesquita et al. (2003), political institutions shape the very logic of political survival. Autocrats typically depend on the loyalty of key inner-circle elites to maintain power – mass unrest does not often take down an autocrat because collective action problems typically hinder threats to their rule from below. Indeed, tight control of information helps keep the masses at bay. Under democracy, by contrast, elections enable the masses to express their will. In order to survive in power, democratic governments must win over and maintain the support of the masses by winning regular elections. Autocratic and democratic leaders face different threats and opportunities from changes to the informational environment.

Part II explores the logic of how the informational environment impacts the likelihood of political survival under autocracy and, alternatively, under democracy. We maintain our focus on the dissemination of data on aggregate economic outcomes, specifically looking at how this form of transparency impacts the decisions of average citizens to continue to support or oppose political leaders and regimes. In these three chapters, we emphasize how the informational environment impedes or enables collective action.

The journey of this inquiry now turns to a question that has been with us from the beginning: why do some governments disclose data while others do not?We have assumed throughout this project that governments have a natural inclination toward opacity. They may better accomplish their priorities in office when not subject to the insistent glare of the population. There may also be financial gains from governing under obscurity. The bright light of transparency makes all sorts of corrupt practices more difficult – bribery, extortion, gombeenism, patronage, graft, embezzlement, and so on. What economic and political benefits might transparency hold for leaders to sacrifice governing in the shadows?

For democratic regimes, Part II reveals benefits of transparency in terms of stability. But do these benefits translate into electoral gains that outweigh the potential for rent-seeking under opacity? Will democratic leaders choose transparency when tempted by the financial gains to be had when the population is left in the dark about the economy?

Turning to autocracies, we have a real puzzle: transparency can lead to unrest. So why would an autocrat ever gamble by disclosing more data? What political benefits does transparency hold for autocrats?

Part III addresses these questions in three steps. Chapter 8 examines the economic benefits of transparency focusing narrowly on investment – both domestic and foreign. The next two chapters address the political rationale to disclose – first for democratically elected leaders (Chapter 9) and then for autocratic leaders (Chapter 10).

In Chapter 8, we argue that transparency brings economic returns in the form of increased investment. Even an autocrat fearing mass unrest may be lured by the sirens of the material gain promised by transparency. As Chapter 8 demonstrates, however, these economic returns are systematically lower for autocrats than for democrats, even as Chapter 9 contends that autocratic leaders care less about these economic gains than do democratic leaders. We conclude that economic benefits may only constitute part of the story of why autocrats would disseminate data.

Investment requires a guess about the future. The investor sacrifices consumption today to create wealth tomorrow. Information about what the futuremight hold is necessary to make a rational decision about the trade-off between present and future consumption. Transparency informs this decision. Indeed, throughout this book, we underscore the importance of our notion of transparency – the public disclosure of aggregated information pertaining to citizen welfare – by advancing a simple premise: such disclosure matters for economic decision-makers.

We recognize that data dissemination impacts a state's economic development. Business people – charged with choosing to expand or contract production, hiring or firing employees, and deciding how to allocate portfolios – perform their tasks with a close eye on the performance of the broader economy. They rely on the government to provide data. Public goods problems inhibit the private production of aggregate economic data, so when the government fails to make such information available, it leaves economic decision-makers blind.

A reader might rightly wonder why, out of all potential economic aggregates, we choose to beginwith investment. Our theoretical models focus on the overall performance of the economy, so one might consider looking at the effect of data dissemination on economic growth. As Roodman (2007a) explains, however, growth fluctuates for idiosyncratic reasons and may not represent a good starting point (see also Dreher et al. 2013).We thus take a straightforward approach and break down economic growth into its components: technological change, capital stock growth, and labor force growth. It may not be obvious how data dissemination would directly affect labor force growth, but transparency does matter for technological change and, perhaps most importantly, for growth in capital.

Investment, both foreign and domestic, is critical to economic performance. The rate of investment, in the short term, dictates the growth rate of the economy and, in the long term, the equilibrium level of production (Solow 1957). Certain investment from abroad (foreign direct investment) may additionally involve transfers of technology and skills between countries, with positive spillovers that increase the rate of growth for the economy as a whole (Grossman and Helpman 2001; Jensen 2006).

Collapsing political regimes bring down with them a hail of uncertainty and can wreak havoc on the lives of citizens, not to mention financial markets. Accordingly, the field of comparative politics has devoted much attention to studying regime transitions. With such an abundance of scholarship apportioned to the topic, debates rage about nearly every aspect of the phenomenon: the causes of autocratic failure, the emergence of democracy, the survival and breakdown of democracy, even the very categories of political regime and how to define a regime transition.

This chapter focuses on the empirical effect of transparency on political stability. We seek to test the specific hypotheses generated by our autocratic and democratic models of Chapter 5. Fully engaging all of the debates surrounding political stability lies beyond the purview of our project. Instead, we accommodate, to an extent, alternative perspectives by testing our theory in different empirical contexts, thus avoiding the need to take sides. This chapter shows that when we control for factors prominent in the scholarly literature, our conclusions about the effects of transparency on political stability hold.

The HRV index proxies as an empirical measure of the concept of transparency in our theoretical models. Recall that in Chapter 5, transparency enters our models as the inverse of the standard deviation of the publicly observable signal of government performance witnessed by all citizens, σs. Our concept of transparency thus pertains to information that is (1) publicly observable – and known to be publicly observable by all actors – and (2) allows citizens to draw accurate inferences regarding government performance. The HRV index, based on the reporting/non-reporting of information to the World Bank's World Development Indicators (WDI) data series, mirrors our criteria. The disclosure of economic information to the World Bank reflects the public dissemination of credible data.

Of course, we do not expect many members of the public to seek access to such data directly from the World Bank. Instead, the public more likely obtains such information straight from the government or through domestic news reports informed by government disclosures. What then is the advantage of World Bank data?

Readers may rightly wonder if other factors explain the patterns thus far revealed.We contend, for example, that data dissemination can cause democratization. We have seen that under high levels of data-transparency, autocracies transition to democracies, and we have also seen that democratic rule survives when transparency is high. Yet these processes of democratization correlate with economic development too, and recall from Part I that data dissemination is also correlated with economic development. Modernization theorists might therefore reasonably argue that increasing economic development could explain democratization as well as the general tendency toward data transparency (see, for example, Epstein et al. 2006). Indeed, an empirical connection between GDP per capita and democratic stability has been thoroughly established (Lipset 1959; Przeworski and Limongi 1997; Przeworski et al. 2000; Benhabib, Corvalan, and Spiegel 2011). Perhaps, then, the connection between transparency and democratization is spurious – driven instead by economic development, which causes both phenomena.

Economic liberalization stands as another potential explanation for the patterns we have revealed so far. Maybe data transparency is simply one facet of liberalization and democratization is another. If so, the direction of causality could be reversed or some other overarching cause of the general tendency to liberalize may cause both transparency and democratization.

Alternatively, some readers might simply suggest that time trends explain the patterns. Over time, data dissemination has increased due to rising demand for data and better technology to collect and organize it. Coincident with this trend, autocracy has collapsed around the world, giving rise to more and more stable democracies. Simply because these phenomena trend together over time does not imply a causal connection. The connections of the HRV index to democratization, autocratic instability, mass unrest, and the removal of democratic leaders through regular means may result from historical coincidence.

Other readers may simply wonder if idiosyncratic heterogeneity across individual countries might better explain their experiences of transparency and stability. The experiences of the countries discussed above – South Korea, Mexico, the Philippines, Cote d‘Ivoire, Bulgaria, Bangladesh, Central African Republic, Guinea-Bissau, and Uganda – may each be better explained with reference to the countries’ individual histories, rather than as part of a broader systematic pattern.

Popular enthusiasm for transparency abounds. Many readers might therefore expect the authors of a book on transparency to conclude with an unadulterated endorsement of transparency and its zealous supporters. Indeed, our original measure of transparency, focusing on data disclosure, could even be used as a metric to shame and pressure governments that lag behind in the quest to produce ever-increasing amounts of data on the functioning of their economies.

Instead, however,we caution readers against the wholesale embrace of transparency as a goal in and of itself. Data disclosure offers many benefits – economic and political. It also entails costs. In this chapter, we consider, in a methodical fashion, the benefits and costs of transparency identified by our research. For democracies, we do recommend transparency. For autocracies, however, our conclusions are nuanced.

ECONOMIC BENEFITS

Economically, data disclosure appears to benefit society. Broadly speaking, transparency increases investment. The effect of transparency on foreign direct investment (FDI) holds across both democracies and autocracies. Interestingly, the effect on domestic investment – while evident when we examine both political regimes together – is weaker under authoritarianism. We suggest in Chapter 8 that this weaker effect on domestic gross capital formation results from the privileged position of economic elites under autocratic political regimes. Because of their especially intimate ties to the ruling leadership, wealthy investors under autocracy may have access to economic data even in the absence of full public disclosure. The effect of making economic data public is lessened, therefore, because key economic actors have access to the relevant information either way.

This is not to say that transparency has no effect on investment under autocracy, but rather that the economic benefits of data disclosure appear, overall, more convincing for democracies. Even as we focus on the economic benefits of transparency, differences emerge across political regimes.

TRANSPARENCY AND DEMOCRACY

Of course, the logic of political survival differs across regime types. Unlike their autocratic counterparts, who survive by cultivating a loyal elite constituency, democrats survive in office by winning elections. Pursuing policies that promote the welfare of large segments of society therefore serves to help keep them in power.

The HRV index represents a new measure of data dissemination. But do we really need another measure of transparency? Other measures of different facets of transparency also exist. If different cross-national measures of transparency were highly enough correlated, then the theoretical distinctions presented in Chapters 2 and 3 would be moot. All indexes would serve as equally strong proxies for the availability of information in a polity. Moreover, it would be nearly impossible to use such indexes in empirical tests to discriminate across precise causal mechanisms. Before proceeding to our original theories about transparency and political stability – and testing them with the HRV index – we must pose a question: How do the various measures of transparency – institutional, media freedom, and data dissemination – correlate with each other?

In this chapter,we study the relationships between various measures of transparency. First, we examine the extent to which measures of these facets covary. We then undertake a more systematic comparison between data dissemination and each alternative measure of transparency in turn. We find substantively important distinctions across the measures. The results of our analysis confirm that the appropriate measure of transparency depends on the theoretical context.

THE COVARIANCE OF FACETS OF TRANSPARENCY

Table 4.1 presents a correlation matrix of measures of five different aspects of transparency. To measure the dissemination of aggregate economic data, we use the HRV index.

As a measure of institutional transparency, we use a variable capturing the presence of a freedom of information law (FOIL). Our measure of FOILs is drawn from Berliner (2014), who codes the year in which FOILs go into effect in a cross-section of countries from 1990 to 2008. We recode these data into a binary indicator ﹛0, 1﹜ that takes the value of 1 if a given country has an effective FOIL in a given year.

As measures of the role media plays in promoting transparency, we begin with two variables. First we include a measure of press freedoms as captured by Freedom House's Freedom of the Press index.We scale this index such that 100 indicates the highest possible level of freedom and 0 represents the lowest. This data series runs from 1994 to 2010.