Political Economy of Reform

This paper reviews the literature relevant to understanding political constraints to economic reforms. Reform refers to changes in government policies or institutional rules because status quo policies and institutions are not working well to achieve the goals of economic well-being and development. Further, reforms refer to the alternative policies and institutions that are available that would most likely perform better than the status quo. The main question examined in the political economy of reform literature has been why reforms are not undertaken when they are needed for the good of society. The succinct answer from the first generation of research is that conflict of interest between organized socio-political groups is responsible for some groups being able to stall reforms so that they can extract greater private rents from status quo policies. The next generation of research is tackling a more fundamental question: why does conflict of interest persist; or, why do some interest groups exert influence against reforms if there are indeed large gains to be had for society? These are questions about norms and preferences in society for public goods. The next step is to examine where norms and preferences for public goods come from, and which institutional arrangements are more conducive to solve the public goods problem of pursuing reforms. After reviewing the available and future directions for research, the paper concludes with what all of this means for policy makers who are interested in understanding the factors behind successful reforms.

This paper reviews the literature relevant to understanding political constraints to economic reforms. Reform refers to changes in government policies or institutional rules because status quo policies and institutions are not working well to achieve the goals of economic well-being and development. Further, reforms refer to the alternative policies and institutions that are available that would most likely perform better than the status quo. The main question examined in the political economy of reform literature has been why reforms are not undertaken when they are needed for the good of society. The succinct answer from the first generation of research is that conflict of interest between organized socio-political groups is responsible for some groups being able to stall reforms so that they can extract greater private rents from status quo policies. The next generation of research is tackling a more fundamental question: why does conflict of interest persist; or, why do some interest groups exert influence against reforms if there are indeed large gains to be had for society? These are questions about norms and preferences in society for public goods. The next step is to examine where norms and preferences for public goods come from, and which institutional arrangements are more conducive to solve the public goods problem of pursuing reforms. After reviewing the available and future directions for research, the paper concludes with what all of this means for policy makers who are interested in understanding the factors behind successful reforms.

Introduction
"Reform" from the perspective of economics refers to changes in government policies or institutional rules because status quo policies and institutions are not working well. Further, reforms refer to alternative policies and institutions that are available that would most likely perform better than the status quo to achieve economic goals. The main question examined in the political economy of reform literature has been why reforms are not undertaken when they are needed for the good of society.
The succinct answer from the first generation of research is that conflict of interest between organized socio-political groups is responsible for some groups being able to stall reforms so that they can get-away with extracting greater private rents from status quo policies. The next generation of research is tackling a more fundamental question: Why does conflict of interest persist, or, put another way, why do some interest groups exert so much influence against reforms if there are indeed large gains to be had for society? The spread of political participation by people at large, both through formal electoral institutions and informal ones of exerting pressure through protests or the threat thereof, begs the question of why it has not led to the articulation of public interest for reforms and push-back of narrow special interests.
The next generation of research also differs from the first in dealing with areas of policy-making and institution design on which there is far less expert consensus and credible evidence on what are the alternatives that would do better, even though there is both consensus and evidence that the status quo is yielding poor outcomes. The first wave of the literature focused on macroeconomic stabilization and fiscal adjustment reforms, policy areas in which consensus is ultimately rooted in binding budget or resource constraints-sustaining macroeconomic and fiscal excesses is simply infeasible. A second wave closely followed, propelled by the collapse of the Soviet Union and the communist countries of Eastern Europe. This second wave examined reforms of economic institutions from being state-led to market-led. While the pursuit of marketoriented reforms can be regarded as having a greater ideological component, relative to adjustment policies to adhere to budget constraints, the advocacy for those reforms was ultimately based on substantial and durable evidence about better economic performance when allocations are largely made by competitive markets, versus the unsuccessful experience of central planning by government. 2 In contrast to these first and second waves, which reached their heights in the 1990s, the current generation of research is tackling problems of public sector reforms where there is little evidence on what constitutes optimal alternatives to the status quo. 3 For developing countries, the 2 Rajan and Zingales (2003) provide a view of the importance of competitive markets, while critiquing anticompetitive practices, guised as pro-market, in countries where economic institutions are generally based on markets and limited government control. 3 Similar to the lack of evidence on what constitutes public sector reforms, Rodrik (1996) argued earlier that structural reforms for economic liberalization are also different from macroeconomic and fiscal stabilization in having less established technical theory and evidence about what constitutes reforms. What further distinguishes the current stream of work on public sector reforms is that there is much wider consensus about the status quo being a problem-for example, corruption-but nevertheless lack of clear evidence on how to solve the problem beyond applying some basic economic principles to strengthen incentives to refrain from corruption. That is, there 3 questions are about the reform of public sector institutions to tackle corruption, and weak incentives and lack of motivation among public officials and frontline public service providers. For example, one type of reform pursued by international development organizations has been the establishment of independent anti-corruption agencies in poor countries, emulating the success of Singapore's Corrupt Practices Investigation Bureau founded in 1952. But these reforms, which transplanted a set of formal institutions which appeared to have worked in another country, have not been considered successful in reducing corruption. 4 Reforms of public sector institutions to better address problems of public goods and market failures are the current wave in this literature, with a very different political economy analytical framework than in the past. The new political economy framework has less to do with analyzing who are the "winners" and "losers" from reform, and how to solve problems of credible commitment, and more to do with examining norms of behavior and preferences for public goods in society. Societies can get stuck in bad outcomes, even though everyone agrees there is a need for change, because of rational beliefs that others are nevertheless pursuing their selfinterest so we should too. This is a problem of non-cooperative norms. On the other hand, people can have deep disagreements about the nature of problems that society needs to solve. Ideological beliefs in society about the nature of problems-such as, for example, the factors behind the apparent rise in strength and frequency of natural disasters-shapes attitudes towards the role of government and of public policy. This is the problem of preferences for public goods, which can be resistant to change through technical evidence about the costs and benefits of different public policies. This paper pulls out these three distinct strands of explanations of political constraints to undertaking reforms from a large and complex web of literature: one, explanations centered around problems of credible commitment; two, around problems of norms, or beliefs about how others are behaving; and three, around preferences for public goods. It shows how all three strands-the technology of commitments, political norms, and preferences--weave the political markets through which society decides upon reforms to tackle problems of public goods. Further, it shows how these questions about the efficacy of political markets to tackle public good problems are equally pertinent for rich or economically developed countries. For example, reforms of health care and education markets, to tackle rising costs in one and persistent inequality in the other, can be viewed as public good problems that remain unsolved due to partisan gridlock and ideological polarization.
There are large open questions in our understanding of the functioning of these political markets, indicating fertile paths for future growth of the literature on the political economy of reforms. After reviewing the available and future directions for research, the paper concludes with what is little evidence on what formal public institutions should be adopted that would sustainably reduce corruption by strengthening incentives against it. 4 Critical reviews of the experience with anti-corruption agencies are provided by Meagher (2005), Doig et al (2006), Recanatini (2011), Heeks andMathison (2012), andMungiu-Pippidi (2015). all of this means for policy-makers interested in understanding the factors behind successful reforms.

Problem of credible commitment
The literature in the 1990s dealt with the political economy of reforms such as, in the words of some of the leading authors, "macroeconomic stabilization, trade liberalization, privatization, deregulation, and related market-oriented measures" (Sturzenegger and Tommasi, 1998, page 3). These reforms typically entailed short-term losses distributed unevenly across groups in society before the benefits from reforms could be realized. For areas of reform such as fiscal adjustment, that were widely regarded as necessary to adhere to budget and resource constraints, political constraints to reforms took the form of costly bargaining between rival groups over the sharing of the costs from reforms. Alesina and Drazen (1991) provided the definitive explanation for how such bargaining, described in their paper as a "war of attrition", could account for costly delays in undertaking the adjustments that all agents knew were eventually necessary. Alesina et al (2006) review the empirical evidence on how this model explains the timing of fiscal adjustment reforms, finding that stabilizations are more likely to occur during economic crises, when the costs of further delays become too high, and when a government has sufficient political strength to overcome opposition.
At the heart of the models of delayed reforms, when all agents know that delays are socially costly or produce Pareto-inefficient outcomes, is a struggle for political power among rival groups. These groups cannot agree to a sharing of costs, and undertake reforms in a timely fashion, because doing so would entail revealing information about relative political power. Acemoglu (2003) shows more generally how conflict of interest between groups, such as arising from unequal distribution of economic costs of reforms, is difficult to resolve in efficient ways because groups cannot commit to not using their political power in future in their favor. Problems of credible commitment are thus ultimately problems about the future exercise of political power.
The question then arises why groups cannot design institutions to solve this commitment problem, if they know that lack of such commitment creates social costs or inefficient outcomes from which all suffer? If reforms are true reforms, and generate a net gain for society rather than being a "zero-sum" distributive game, where the benefits for some groups comes at an equal cost for other groups, then institutions should, in a normative sense, allow "winners" to compensate "losers". Indeed, some types of institutional reforms in the literature have been explained as the result of society addressing the problem of credible commitment. For example, delegation of monetary policy to a politically independent central bank whose leaders are selected on the basis of technical expertise to pursue technically sound policies (Rogoff, 1985;Keefer and Stasavage, 2003). Political reforms to institute popular elections as the way leaders are selected and sanctioned have also been explained as a solution to the problem of credible redistribution and elite rent-sharing (Acemoglu and Robinson, 2000;Bidner et al, 2014).
The degree of inequality in the distribution of costs of reforms, or of benefits from the status quo, matters significantly in explaining whether reforms are pursued or not. When political 5 institutions and status quo policies are such that large rents are captured by a few small interest groups, those interest groups have little incentives to pursue reforms. Rather, those interest groups have strong incentives to prevent reforms because reforms are most likely to redistribute gains more widely, reducing the net gains accruing to these groups. Hoff and Stiglitz (2008) have this intuition at the heart of their model to explain why societies find it difficult to transition to the rule of law-those that stripped assets illegally during the lawless state would have too much to lose under the rule of law. Similarly, Rajan (2009) models the persistence of underdevelopment as arising from an initial inequality in distribution of endowments such that the constituent groups have no incentives to pursue reforms.
The literature on institutional reforms thus suggests that the commitment problem is linked to initial conditions of inequality in endowments, both economic endowments and those of political power, such as control over instruments of violence and coercion. When the inequality, of power or of benefits from status quo policies and institutions, is too large, then elites select Paretoinefficient institutions to hold-on to their rents and powers in the future. Put simply, it is difficult to commit in future to make everyone better off without making the elites worse off, given the initial inequality in endowments. Reforms under these formidable initial conditions have been modeled as happening either through the technology of revolution to overthrow the elites, or through a real change in the composition of elites and their preferences. Acemoglu and Robinson (2000) analyze how institutional reforms can come about through popular mobilization or the threat thereof to overthrow the elites. On the other hand, Lizzeri and Persico (2004) and North et al (2007) argue that significant shifts among elites in their demand for more public goods, for which they are willing to give up their private rents, account for deep institutional reform in the history of nations.
An initial level of inequality in endowments is not necessary to examine political constraints to reforms. Fernandez and Rodrik (1991) model more generally how uncertainty over the distributive implications of the costs and benefits from different economic reforms can explain persistence of status quo policies. Credible commitment problems emerge whenever there is uncertainty with regard to the distribution of costs and benefits, and explain rational political resistance to reforms as a distributive struggle. But once again the question is why groups cannot find institutional solutions to the commitment problem if there really are significant gains in net to be had from reforms.
Lack of solid technical evidence, and ideological resistance to available evidence, on the value of reforms is part of the political economy explanation for why reforms are resisted. Majumdar and Mukand (2004) explicitly model the uncertainty involved in policy-making choices, and propose that the real political economy question is the following: "what factors influence a government's incentive to engage in policy experimentation and learn from the information so generated?" (page 1207). This paper will show in the followings sections how political norms and preferences for public goods are the fundamental determinants for whether governments have incentives to pursue such learning and make policy decisions on the basis of the best available evidence. Alesina and Tabellini (1989) and Cukierman et al (1992) explain resistance or delays in reforms as ultimately due to the desire of groups to hold office in the future to implement their 6 ideologically preferred policies. These are problems of preferences and norms, rather than commitment problems per se. In the classic "war of attrition" model of Alesina and Drazen (1991), if all groups bore the burden of reforms equally, there would be no inefficient outcomes of delays in reforms. One way in which the authors interpret the expected asymmetry in the burden of reforms is as representing the "degree of cohesion" in society. The following sections review the emerging literature on how cohesion in society can be interpreted variously as norms for cooperative behavior and as preferences for broad public goods, the lack of either of which can explain political obstacles to reforms.

Problem of norms
The research on political economy that came after the waves of the 1990s on macroeconomic reforms, focused, in contrast, on more micro problems of agency within the public sector in developing countries. Dixit (1996Dixit ( , 2003 summarized the problem in the language of principalagent theory as a problem of common agency-when multiple competing powerful interests attempt to control, as principals, the actions of government agencies. At the heart of the problem of common agency is that political contracts are not "justiciable"--you simply cannot ask a court to enforce them. 5 This lack of enforceability of political contracts by an independent third party is also the argument in Acemoglu (2003) on why a Political Coase Theorem is infeasible. If a contract among rival interest groups is not justiciable, it has to be self-enforcing, and therein lies a Prisoner's Dilemma. Each group can extract private benefit by reneging -offering the common agent a little bit extra to better serve their group interests at the expense of the other principals. Resolving this dilemma between groups, which arises ultimately from lack of legal enforcement by a third party, then involves some system of norms, and accompanying informal sanctions, to make credible their commitments to adhere to the agreement.
To concretize this discussion, let's take the case of anti-corruption reforms in developing countries. Dixit (Forthcoming) argues that persistent and systemic corruption should be understood as a Prisoner's Dilemma in a game theoretic framework. He applies the framework specifically to the interaction between different business interests as "givers" of bribes. Although it would be beneficial for business interests as a whole to reduce bribe-giving, they are instead stuck at high corruption levels because individuals believe that engaging in corruption is the best they can do given how others are behaving. He then discusses how collective action will be required to reduce corruption, establishing a new norm of refraining from giving bribes.
Why cannot countries solve this problem by instituting anti-corruption agencies (ACAs), as in the case of Singapore, which would monitor and prosecute the "askers" of bribes among politicians and bureaucrats in government? A number of case studies have examined the experience with anti-corruption reforms in poor countries, including the establishment of ACAs, and found that they fail because of weak incentives among powerful political leaders to reduce corruption (Meagher, 2005;Doig et al, 2006;Recanatini, 2011;Heeks and Mathison, 2012;Mungiu-Pippidi, 2015). Why would there be weak political incentives to reduce corruption, 7 given the substantial efficiency and equity costs of corruption that has been documented by research (Olken and Pande, 2012, provide a review)?
Before turning to what the literature has to say on the question of incentives of powerful leaders, it is worth discussing other evidence in the literature on how even when leaders try to institute anti-corruption reforms on the basis of economic principles 6 , these reforms often unravel. A growing body of evidence shows that despite evidence of short-run impact of some anticorruption initiatives, there is also evidence suggesting that the long-run impact may be very different. Olken and Pande (2012) review the evidence, finding a consistent pattern across countries of public officials adjusting to the new incentive regime and finding new ways to circumvent, or even sabotage the reforms, 7 such that corruption eventually rises again or appears in new forms. Furthermore, these authors conclude their review by indicating that we have little evidence to evaluate whether a swathe of anti-corruption reforms pursued by international development partners, such as the setting-up of independent anti-corruption agencies or the legislation of freedom of information, has impact in sustainably reducing corruption. The case studies referred to earlier provide qualitative and suggestive evidence that these reforms, popular with development partners, have been undermined and politically manipulated, with little reason to expect significant impact on corruption. In sum, what the evidence shows is that there is a problem with the status quo-of widespread corruption-which needs reforms, but what those reforms are that would sustainably reduce corruption, is far from clear.
Stable norms of non-cooperation in society at large can explain why reform efforts by powerful leaders to incentivize reductions in corruption often fail. World Bank (2016) and (2017) review the evidence on how reform efforts by leaders to solve principal-agent problems within government agencies are stalled, repealed, or produce unintended consequences of displacing corruption to other avenues.
To take one example: bureaucrats leading India's National Rural Health Mission identified absenteeism by frontline public health workers as a significant problem. 8 The bureaucrats responded with a technological innovation to enforce attendance: a biometric monitoring system to digitally capture the thumb print of each staff member at the start and end of the working day. Researchers invited to evaluate the impact of this initiative found a significant increase in the attendance and improvements in health outcomes. However, the results suggested that other forms of malfeasance, such as diversion of patients to private practice, and reductions in benefits to which patients are entitled, may have been substituted for absenteeism. Furthermore, the 6 There is evidence from high corruption contexts that applying some basic economic principles to solve principalagent problems, such as by increasing monitoring and enforcement, does have impact. For example, Di Tella and Schargrodsky (2003) find that various initiatives to strengthen incentives to refrain from corruption in procurement in public hospitals in Argentina have impact in reducing corruption consistent with the predictions of standard economic models of bribing a la Becker and Stigler. 7 Banerjee et al (2008) document the sabotage of time-stamp machines introduced in India to monitor attendance by public health workers. 8. This example is drawn from Dhaliwal and Hanna (2013) and pertains to the state of Karnataka in India. Rampant absenteeism by providers in the public sector has been documented in World Bank (2004), deemed as "quiet corruption" in World Bank (2010), and estimated to place a significant fiscal burden on states (Muralidharan et al, 2017). 8 researchers encountered low motivation for taking up the reform and implementing it effectively. The researchers noted that locally elected politicians, whom they thought would have better incentives than the bureaucrats to monitor the provision of health services (because the politicians could be voted out of office if services were not provided), did not think absenteeism or service delivery were problems.
Theoretical advances in research suggest that part of the failure of leaders to successfully implement reforms can be attributed to lack of legitimacy, which is a concept closely linked to norms within complex organizations. Akerlof (2015Akerlof ( , 2017 offers a model of legitimacy-based authority, that is, the level of orders that a principal can give that will be carried out by the agent because following orders is a "duty." Akerlof shows how the model is operationally equivalent to agents monitoring each other to ensure compliance with orders when leaders have greater legitimacy. This mechanism in the model can be interpreted as the existence of norms among peers to perform these duties. If legitimacy is lacking, however, the agents will reject that there is a duty, or a norm, to comply with it. However, the origins of legitimacy are not modeled in Akerlof (2015). Bolstering legitimacy is treated as an investment decision of leaders.
Other work shows how legitimacy for cooperative behavior can arise from democratic processes. Dal Bó, Foster, and Putterman (2010) offer an experimental study that enables a view of legitimacy and legitimate behavior as what arises when people choose their own rules of the game. In their experiment, subjects can vote for a policy (or institutional) reform that will transform a prisoners' dilemma game into a coordination game. Subjects may then find themselves in the new game either as a result of their own votes (an endogenous, "organic," reform), or as a result of an arbitrary, random decision of the computer (an exogenous reform). Subjects coordinate in the good equilibrium significantly more often when they access the coordination game because of their own decision rather than the computer's, even if the information about the vote is held constant, suggesting a buy-in or legitimating effect, of the reform bestowed by the democratic process.
This work by Dal Bó, Foster, and Putterman (2010) suggests that homegrown policies and institutions will be more effective at improving behavior or norms than transplanting reforms from outside. It thus joins other work critiquing external donor efforts to build capacity in developing countries and attempt to replicate or transplant successful rich country institutions in very different contexts (World Bank, 2017;Pritchett, Woolcock, and Andrews 2013;Andrews, Pritchett, and Woolcock 2013;IDS 2010).
But the question raised in Akerlof (2015) remains-where does legitimacy of leaders come from? The research of Dal Bó, Foster, and Putterman (2010) is not able to speak to whether democratically elected leaders will have greater legitimacy for establishing anti-corruption as a duty. Other work on the role of leaders in signaling a shift in norms further underlines the importance of the question of legitimacy of leaders to pursue reforms. Older work has examined how leaders can exemplify certain types of behavior. Hermalin (1998) provides an early analysis of how a leader may promote efforts by others through his or her own effort, thereby signaling the worth of a task. In more recent work, the behavior of leaders is modeled as providing a signal to citizens of shifting political norms under which citizens feel indignant about bad performance 9 and act to hold leaders accountable (Bidner and Francois 2013). Acemoglu and Jackson (2015) also offer a theory in which the behavior of leaders (actors whose actions are visible to all) affects norms (citizen beliefs about mutual behavior). In these papers, leaders can take actions that subsequently lead to citizens' adopting cooperative social or political norms.
The explanation for the persistence of corruption in poor countries has been that powerful political leaders extract private rents from public office and therefore it is not in their interests to reform institutions to reduce corruption (Treisman, 2007;World Bank, 2017). This explanation can be linked to the above discussion on legitimacy of leaders (or lack thereof) to account for why even when powerful leaders try to reduce some forms of corruption further down the chain, at the frontlines of public service delivery, their efforts frequently fail. Presumably, because leaders are corrupt, or at least viewed by society as corrupt, they lack legitimacy to implement reforms to reduce other forms of petty corruption at the frontlines. Yet, if corruption is indeed socially costly, incentives of powerful leaders to extract private rents is not a sufficient explanation. The question of why powerful leaders would maintain Pareto inefficient means of extracting private rents, remains. The answer provided by earlier cited theory also remains-that leaders cannot find a justiciable alternative to ensure their future rents if they were to reform institutions.
Whichever way we look at it-incentives of powerful leaders, or failure of reforming leadersthe problem of norms is relevant to understanding how societies move from a high-corruption to low-corruption equilibrium (Mungiu-Pippidi, 2015). Although it would be beneficial for society as a whole to reduce corruption, society is instead stuck at high corruption levels because individuals believe that engaging in corruption is the best they can do given how others are behaving. For example, bureaucrats ask for bribes in order to provide public services, and citizens pay these bribes because they believe that most others engage in bribery; if you refuse to pay the bribe, you will get nothing, or worse, you may suffer retribution. This characterization of a Prisoner's Dilemma underpinning failures of reforms to tackle corruption is different from that of the macroeconomic literature of the 1990s in that the actors caught in the dilemma are far more numerous, dispersed, unorganized, and "ordinary" than the bargaining between organized interest groups examined in the prior literature.
Reforms to reduce the agency problems in government are questions at the research frontier. While significant advances are being made on the proximate interventions, based on economic principles, that would address specific agency problems and their inherent information asymmetries, the complex problem of norms requires much more research. Political norms, and preferences, which will be taken up in the next section, determine whether the leaders who wield policy-making powers to take-up better proximate interventions have appropriate incentives and legitimacy to pursue those reforms.
Taken together, the available body of work suggests a pathway from political markets to norms in the public sector (World Bank, 2016). If "clean" politics can bolster the legitimacy of leaders for delivering policies that benefit society as a whole, or public goods, and the reverse is true when politics is "dirty," cleaning up politics can be an avenue for larger changes in how bureaucratic institutions function. Bolstered legitimacy of political leaders for delivering public goods and leaders' roles in signaling a shift in behavioral norms, might translate into improvements in norms of performance in bureaucracies. In environments in which politics is becoming cleaner, public officials and frontline service providers may be more inclined to monitor each other to comply with rules rather than bend them for private gain, both because they view leaders as setting an example, and because of what they learn from the behavior of leaders about prevailing norms in society.
The fundamental problem in the political economy of reform that is suggested by this research frontier is that politics is frequently dirty. Political markets that function in unhealthy ways, such as through vote buying and ethnic favoritism, explain weak incentives and legitimacy of leaders to reform, as well as adverse selection of worse types of leaders who are not motivated to pursue reforms (World Bank, 2016). Despite a strong tradition in economics to eschew any explanation that relies on myopic or unenlightened preferences among agents, the newest crop of research that is relevant for understanding political constraints to reforms examines political preferences among citizens at large that can explain why the wrong types of leaders emerge, and why even well-intentioned reformers lack the legitimacy to succeed, or are removed from office when they try.

Problem of preferences
This section begins with showing how work on the political economy of reforms connects to a literature on the correlation between ethnic heterogeneity and economic performance, and how this literature on ethnicity illuminates a broader principle about political preferences going beyond ethnicity in poor countries to the role of ideology in rich countries as well. Citizens' preferences for emphasizing private benefits over public goods when evaluating leaders, either delivered by co-ethnic leaders, or through vote buying at the time of elections, appears to be a significant factor explaining persistence of clientelist politics in democracies in the poor world and its negative consequences for political incentives to pursue reforms.
The basic intuition here-that citizens vote in ways that do not fully account for the negative consequences in equilibrium-is quite similar to another literature on special interest capture in rich countries Helpman, 1996, 2001). The political science literature on expressive voting in advanced economies similarly discusses the powerful role of ideology and charisma in shaping how citizens vote, because citizens derive some direct value (utility, in economists' parlance) from the act of voting for their ideologically preferred party (Romer, 1996, provides a succinct review). Laboratory experiments conducted with college students in the United States, and hence, highly educated subjects residing in a rich country, provide evidence that citizens are cognitively constrained to demand policies that confer short-term benefits at the expense of long-term or "equilibrium" costs (Dal Bó, Dal Bó, and Eyster, 2013).
Ethnic divisions are regarded by some scholars as the primitive or fundamental factor driving all others in explaining why some regions of the world are rich and others are persistently poor (Ashraf and Galor 2015). Substantial evidence also exists of an association between ethnic fragmentation and low investment in public goods, which suggests a connection with political constraints to achieving reforms for the greater public good (Habyarimana et al. 2009;Miguel and Gugerty 2005). More research is also emerging on the mechanisms through which identity politics might reduce incentives of policy-makers to pursue reforms or public goods. Banerjee and Pande (2007) provide micro evidence from India in support of the hypothesis that voter valuation of ethnic identity contributes to weak incentives among leaders. They document that in the state of Uttar Pradesh in India, political parties are organized to appeal to caste identity and regularly field candidates against whom there are allegations of corruption and criminality. The authors explain this through a model in which greater corruption arises in ethnic parties because citizens value honesty in political leaders less than reliable redistribution based on shared ethnic identity.
Further evidence is emerging that is consistent with ethnic identity playing a similar role as group identity in very general models of political bargaining between groups to extract groupspecific benefits from public policies, which was at the heart of the 1990s literature on political constraints to reforms. Alesina, Michalopoulos, and Papaioannou (forthcoming) find that it is not ethnic fragmentation per se that matters in explaining variation in economic development, but rather its interaction with economic inequality. Economic inequality between ethnic groups, largely arising from differences in geographic endowments across their historic homelands, is associated with lower real GDP per capita. The typical measure of ethno-linguistic fragmentation used in the literature loses its significance when the authors' measure of economic inequality between ethnic groups is included. This is consistent with the view that redistributive conflict between groups can contribute to political failures to select and implement appropriate public policies that promote economic development.
The political science literature has examined the role of ethnic identity in sustaining clientelist political mobilization in developing countries, in which power is won on the basis of providing targeted private benefits in exchange for political support (Kitschelt and Wilkinson, 2008). While there is a lot of work examining the details of clientelist political transactions, and its political outcomes, there is relatively little work on implications of clientelist politics for the incentives of leaders who wield powers over policy-making. Although there are some theories about the negative consequences of clientelist politics for policy selection in equilibrium (Baland and Robinson, 2006;Bardhan and Mookherjee, 2012;Anderson et al, 2015), it is also conceivable that vote buying could be a form of political responsiveness to the demands or needs of voters. Gift-giving at election times and service delivery in between elections may be positively correlated, simultaneously driven by underlying conditions of demanding citizens and political contestation. Or, electoral strategies may have no bearing on public policy choices and government performance once politicians gain office. The received theory is insufficient to draw firm conclusions.
However, a few studies are beginning to provide direct empirical evidence that political leaders trade-off the provision of private benefits against that of providing broader public goods-that is, in equilibrium, clientelist politics yields under-investment in broad public goods (Khemani, 2015;Anderson et al, 2015). A survey was designed to be able to correlate vote buying with the delivery of public health services, taking advantage of an opportunity in, and an appropriate institutional context offered by the country of the Philippines (Khemani, 2015). The data reveal that where political contestation using vote buying strategies is more widespread, politicians reduce investments in those service delivery instruments-in primary health-that are particularly likely to be broadly pro-poor in delivering benefits. Anderson, Francois, and Kotwal (2015) show that weakening enfranchisement through clientelist relations between land-owning elite families and agricultural households in Indian villages is associated with lower public investments in social safety nets and antipoverty programs. In related results, Baland and Robinson (2006) find historical evidence that the introduction of the secret ballot in Chile was associated with a decline in votes for right-wing parties and a drop in the price of land (both of which are consistent with a reduction in the power of clientelist relations).
Yet, myopia among ordinary citizens, with limited cognitive capacity to internalize the equilibrium consequences of their political demands for private benefits, cannot explain why the elites-both economic and political-who have access to expert advice and technical evidence on equilibrium impact of different policies, do not pursue reforms. Indeed, one possible institutional solution to dealing with myopia among ordinary citizens is to select authoritarian institutions of decision-making. Yet, there is no systematic evidence that countries with more authoritarian institutions perform better than those with more democratic institutions (Acemoglu et al, 2014;World Bank, 2016). Resurgent evidence of elite capture in the richest economy of the world also poses the question of why reforms are not pursued when elites are in-charge (Gilens and Page, 2014). One answer, logically, seems to be that certain reforms are simply not Pareto efficient given an initial distribution of endowments-they would require the elites to give-up some private rents in net, because the starting point is one in which the elites capture enormous advantages from the status quo.
One account, therefore, of how reforms happen is: elites pursue reforms when they are persuaded that those reforms would indeed provide more public goods that elites value. An increase in elite valuation of public goods has been argued to be at the heart of deep institutional reforms in the history of currently rich countries, which persuaded elites to reform institutions such that government bureaucracies would have the capacity to deliver those public goods (Lizzeri and Persico, 2004;Glaeser and Goldin, 2006).
In contexts with strong democratic institutions in which political leaders have to be responsive to demands from citizens at large, some of the political economy of reform literature has directly engaged the question of populist demands that are not sufficiently informed by technical evidence on the costs of those demands. Boeri et al (2002) examine the case of pension reforms in Europe, finding that popular opinion is against reforms even though people are generally aware that the status quo is unsustainable. They speculate that the underlying problem could be selfishness of preferences among citizens which leads them to push the costs of pension systems on to future generations. In further work in this area of pensions reform in Europe, Boeri and Tabellini (2012), find that specially targeted information about the reforms can help to increase popular support for it, while exposure to typical media coverage does not.
The role of information in persuading citizens appears to be equally important in poor countries, and even for issues such as ethnic identity which could be regarded as fixed preferences that are resistant to change. The same authors who argued that ethnic preferences can explain persistence 13 of corruption show in other work in the same political context that voter preferences regarding ethnic identity are malleable and responsive to information about the severity of allegations of transgression (Banerjee et al. 2014;Banerjee et al. 2009). This is consistent with a large literature finding evidence of responsiveness of voters in poor countries to the availability of information, despite theoretical predictions in a long-standing political science literature that voters in these contexts are unlikely to be swayed by information (World Bank, 2016;Keefer and Khemani, 2014).
Yet, the way forward for reforms is far from clear. The solution to the political economy of reforms does not appear to be as simple as giving people more information about the costs of the status quo and the benefits of reform. The problem of communication to persuade changes in beliefs and preferences requires as much serious and scientific study as has been afforded to the body of work on economic reforms and political constraints to those reforms. The problem of information and communication is in fact inter-twined with the issues of credibility and legitimacy of leaders that has been part of the economics literature, and it requires the tools of the economics discipline to tackle.
For example, Glaeser and Sunstein (2013) find that communication by leaders matters in ways that go well beyond the underlying information being communicated. They provide the following example from the United States to show how whether citizens are persuaded or not depends upon their perceptions about the agent communicating the information: "When liberals and conservatives are asked for their private views about a generous welfare policy and a more stringent one, they react in the predictable ways, with liberals favoring the former and conservatives the latter. But things change dramatically when they are informed of the distribution of views within the House of Representatives. More specifically, conservatives end up disapproving of the more stringent policy, and favor the generous one, when they are told that 90 percent of House Republicans favor the generous policy. Liberals show the same willingness to abandon their private opinions, and thus end up favoring the stringent policy, when told that this is the position of 90 percent of House Democrats. Notably, the effect of learning about party views is as strong among those who are knowledgeable about welfare policy as it is among people who were not. Also notably, both conservatives and liberals believe that their judgments are driven largely by the merits, and not by what they learn about the views of their preferred party-but in that belief, they are wrong." (Glaeser and Sunstein 2013, 23) The body of research in the United States on how political polarization impedes reforms also examines how even the existence of free and balanced media markets is unable to overcome the problem. For example, Glaeser and Sunstein (2013) show that in the presence of preexisting beliefs that polarize citizens, media-provided information is not only unlikely to shift beliefs but might in fact cause previous beliefs to become even more entrenched. The same information can activate completely different memories of personal experiences and associated convictions, thus producing polarized responses to that information. Campante and Hojman (2013) provide evidence from the US that the intersection of media and political markets influences ideological polarization among politicians and citizens. They find that the introduction of broadcast television decreased the ideological polarization of the U.S. Congress. Information provided by television acted as an important force for bringing members of congress toward the political center. They also find similar effects for the expansion of radio in the 1930s. Their framework can be extended to suggest how different types of media play a role in the steady increase in polarization in the United States since the 1970s. This increase in polarization has coincided with significant structural changes in the American media environment-the rise of talk radio, the expansion of cable television, and the growth of the Internet, all of which are associated with content differentiation and market segmentation that may have contributed to increased polarization.
The insights from the United States can be extended to developing countries' experience with a particularly difficult area of reforms-of energy price subsidies. A massive IMF report on the problem of reforming subsidies concludes that informing citizens about the cost of subsidies is not sufficient to solve the problem (Clements et al, 2013). This is the Gordian knot of the political economy of reforms: lack of information about the value of reforms impedes them, and yet citizens are resistant to change their beliefs and preferences when confronted by evidence. Widespread beliefs about the corruptibility of officials make idiosyncratic policy choices-such as subsidy reforms-seem suspect, thereby providing honest politicians with incentives to pander to populist demands (World Bank, 2016, reviews the literature). Credibility and legitimacy of leaders is key (as in the model of Akerlof, 2017), but this begs the question of where credibility and legitimacy come from.
Questions about credibility and legitimacy are really questions about beliefs and preferences and requires economics to confront the problem of endogenous formation of these. The economics literature has given little attention to where preferences for public goods come from. It has waded into analyzing political choices by extending the assumption that individuals will pursue selfish interests and demand their privately preferred goods from the public sector, just as they do in private markets. But recent advances in economics are connecting with sociology, anthropology, political science and social psychology to examine the role of "culture", or how beliefs and values shape preferences and behavior, and thereby, economic outcomes (Alesina and Giuliano, 2015, provide a review). The next step in the study of the political economy of reform is to examine where preferences for public goods come from, and which institutional arrangements are more conducive for society to solve the public goods problem of pursuing reforms.

Conclusion: What does this mean for policy-makers who want to know which factors make reforms more likely to be successful?
An earlier article reviewing the political economy of reforms literature discussed the answers provided to this question in Williamson (1994), based on the contributions of high-ranking technocrats of the lessons from their experience with reforms. The reader is referred to Table 9 in Rodrik (1996) for the proposed list of factors that make reforms more likely to be successful. However, Rodrik (1996) concludes that none of these factors were found to be either necessary or sufficient for successful reforms. One of the dominant opinions that emerged among the leading technocrats engaging in reforms that were gathered for this exercise is that reforms should be undertaken by stealth, obscuring the policy actions from society and citizens-the opposite conclusion to the current support, two decades on, for transparency as an inescapable factor in winning reforms (Clements et al, 2013). A stock-taking of available research on transparency suggests that its role is far more complicated than suggested, with many open questions on how transparency, or shared information, and deliberation around information, impacts beliefs and preferences (World Bank, 2016). With these caveats of large gaps in our knowledge, and therefore a fertile research agenda on how beliefs and preferences about the nature of public goods are fundamental to understanding reforms, here is a new list of factors for reforming policy-makers.
1. Reforms are more likely to be successful if they are truly reforms-they would bring large benefits over a costly status quo. Counting any type of policy or institutional change as a "reform" is not consistent with how economics research examines reforms.
2. Reforms are more likely to be successful if they are supported by extensive technical evidence, that is credibly non-ideological and non-partisan, on how costly the status quo is and how the reform would bring substantial benefits. A global public good for reformers everywhere is the presence of independent think-tanks and research agencies who establish their independence by making use of technical expertise and building professional reputation.
If conditions 1 and 2 are met, and the obstacle to reform is losses concentrated on a few, the reforming policymakers should be able to convince the concentrated "losers" from reform that they will share the benefits, and cut a deal with them, so they do not block the reform. However, one problem that may present itself, despite conditions 1 and 2 being met, is the future rents these groups risk giving-up if they agree to the reforms. The costs for them may be simply too large for the reforming policy-maker to compensate in practice. If that is the case, then: 3. Reforms of political institutions are needed to support all other types of policy reforms, when the losers from reforms are so powerful that they will block reforms despite conditions 1 and 2. Short-term negotiations over reforms have to be accompanied with longer term institutional change to reduce the power of special interests to capture the state. This requires better understanding where the power of special interests comes from.
Part of the power of groups to block reforms, despite the evidence that the reforms would benefit an overwhelming majority, is likely to come from their ability to persuade the "winners"society or citizens at large-that the reforms are not in their interest. Or, to distract citizens into valuing other aspects of policy-making (such as social and religious policies), systematically mobilizing citizens on those non-economic policy dimensions, while deliberately pursuing economic policies to the detriment of those citizens.
4. If conditions 1 and 2 are met, and the obstacle to reform is populist demand against it, or the powerful "losers" spreading mis-information, then reformers have to communicate the evidence better-communication is not a soft option. Understanding the role of information and communication in shifting beliefs, norm, and preferences is at the cutting edge of political economy research and requires rigorous scientific analysis, using the tools of economics to analyze market interactions. Simply addressing credibility problems, such as, for example, attempting to remove a price subsidy by first committing to a lump-sum transfer to compensate citizens for the loss of the subsidy, will not be durable. The next time a new price shock hits, citizens will be out in the streets again, demanding new subsidies. This whac-a-mole problem of reforming subsidies has been documented for several countries (Clements et al, 2013).
Even with scientific investment, ideological polarization may make communication impossible. Kahan et al (2011) and Kahan (2012) provide evidence from laboratory experiments in the United States that those who score highest on cognitive abilities are also the most likely to adopt ideological beliefs in defiance of technical evidence. The authors explain this as "motivated reasoning," adopted by people to conform to the values of the cultural groups to which they belong. Some of their findings in the specific area of environmental policy suggests that public divisions over climate change are not a reflection of the public's lack of knowledge or misunderstanding of science; rather, divisions stem from conflicting interests-between an individual's interests in adopting common beliefs held by one's community versus a collective interest in making decisions to promote common welfare. Holding a viewpoint that goes against one's surrounding cultural community can exact a cost. When this cost is high, it is rational for individuals to adopt the reasoning prevailing in the group. For such polarized societies the work of reforms requires long-term building of common purpose and identity-taking us into territory on which there is little work in economics (Singh, 2015). Even so, economists can provide valuable analysis of the impact of policies, such as education curriculum reform, on outcomes of beliefs and preferences (Cantoni et al, 2015).
In addition to examining communications options more scientifically, reformers may also want to examine other areas of policy (that are not currently part of the reforms package they are pursuing) that may be responsible for the problems of credibility and legitimacy. Citizens may not support reforms in one area-such as repealing subsidies-despite understanding its costs, if they believe there are no other benefits they can expect from the state and its prevailing institutions. Populism in some areas of policy may be connected to experience with elite capture in other policy areas. Other policy actions may be needed for the government to build its credibility-such as reforms in areas where there is evidence of elite capture; or, delegation of some areas of policy-making to independent agencies and professional bureaucracies that are protected from elite interference. That is, this is a problem of credibility when a government is only pushing for one reform, and not willing to do others that would take-away its own rents or grip on power.