Policy makers concerned with development and reform increasingly put forward political will as a silver bullet in the fight against weak development performance. As effectively argued by the Director of the Bureau for Development Policy at the United Nations Development Programme (UNDP) in the discussion about how to successfully achieve the Millennium Development Goals: “There is no rocket science in any of the action points. It's a matter of political will.”Footnote 1 According to the UN, “[today] we not only have the financial resources to end extreme poverty once and for all, but we have the technological knowledge and know-how to realize the Goals…. The way forward is marked, it is only the political will to achieve the Goals that is in question.”Footnote 2 In line with this conviction, the World Bank has launched a leadership development programme, and the Ibrahim Foundation has introduced an annual prize that rewards excellence in leadership.Footnote 3
The reasoning of policy makers gains resonance in contemporary, and highly influential, policy-oriented academic discussions on development and reform.Footnote 4 As such, the importance of political will is frequently invoked in popular pieces on development and reform written by academic scholars for newspapers, magazines, and journals directed towards policy makers and an educated readership. For example, Robert Rotberg, Sahr J. Kpundeh, Jeffrey Sachs, Arnold C. Harberger, Marina Ottaway, and Jon T.S. Quah have all highlighted the decisive role of a “highly principled” leadership in the development process.Footnote 5 In a nutshell, according to the contemporary policy discourse on political will, development failure can be blamed on “vicious, wicked, and grossly incompetent leaders”, while development success can instead be ascribed to “responsive and responsible leadership.”Footnote 6
We do not question the important role played by political will in the process of development. Rather, our very motivation here is the conviction—supported by a vast number of academic studies—that the capacity of political elites to mobilize from “above” has a significant impact on the way ordinary citizens behave and, hence, on development outcomes.Footnote 7 Political will—in this context broadly understood as the will of leaders to initiate and sustain reform—is, in other words, rightly put forward by the policy-making community and policy-oriented academic scholars as an important causal factor in the development process.Footnote 8
However, having said this, we still argue that the contemporary policy discourse on political will is based on a severely flawed understanding of leadership behavior. Hence, it is also limited in terms of the ability to bring about positive development change. To begin with, the carriers of the contemporary policy discourse on political will lack the analytical tools necessary for understanding the behavior of leaders across time and space. As a result, they have a propensity to engage in retrospective and circular post-hoc, atheoretical analyses. In particular, they tend to resort to the simplistic argument that leaders who promise reform but fail to implement such did in fact not have any “genuine” political will to begin with, implicitly assuming that all failed reformers are “liars.” Furthermore, the carriers of the contemporary policy discourse on political will tend to adopt an excessively voluntaristic view of leaders, effectively down-playing the contextual influences on their behavior. By overlooking the opportunity and incentive structures leaders face, they hence fail to offer any deeper understanding of why some countries—such as the majority of the sub-Saharan African ones—seem to be constantly plagued with inherently kleptocratic leaders, lacking the political will to conduct reform, while others—such as the majority of countries situated in Western Europe—seem to be as constantly blessed with farsighted leaders. In short, the contemporary policy discourse on political will is based upon a circular and voluntaristic understanding of leadership, making it suffer from severe theoretical shortcomings. As a consequence, it is also severely constrained in terms of its ability to generate effective policy solutions to the pressing problem of bad leadership.
Our overall aim is to contribute to the development of a more analytically thorough understanding of leadership behavior across time and space. As such, we aim to answer the question what factors beyond the personal traits of leaders can explain variation in leadership behavior. At the most basic level, a theoretically more satisfying answer to this question implies a move away from the voluntaristic and circular approach offered by the contemporary policy discourse in favor of an approach that takes context and causality seriously. In particular, the development of a more analytically and empirically valid account of leadership behavior demands that we refrain from using the concept of political will (or the lack thereof) as a “black box” or an “error term”—a default, exogenous explanation to why development and development reform succeed or fail—and instead firmly integrate the concept into a more systematic framework of analysis.
Such a systemic framework, we argue, must contain an account of both the micro and macro foundations of leadership behavior. As such, it should not only contain a theory about the micro-level logic of leadership behavior, but also a theory about what broader, contextual factors influence leadership behavior. In the endeavor to advance such an integrated theory, we believe that the contemporary policy discourse on political will (and hence our understanding of development and development reform) would benefit greatly from engaging in more active dialogue with established theories within the fields of economics and political science. In particular, it would benefit significantly from taking part of the combined insights offered by principal-agent theory and what we refer to as state theory.
While both these theories can to some extent each and on their own theoretically account for variation in leadership behavior reaching beyond the influence of personal traits, they are still limited in the sense that they focus exclusively either on the micro foundations of leadership behavior (i.e., principal-agent theory) or on the macro foundations (i.e., state theory). However, in tandem, we argue, these two theoretical frameworks have the potential to lead us a very long way towards a more analytically integrated and empirically valid account of leadership behavior. More specifically, by introducing the concepts of “moral hazard” and “adverse selection,” the principal-agent framework offers the analytical tools necessary to describe and properly label empirically observed leadership behavior. In particular, it provides theoretically anchored tools for labeling what is now commonly brushed aside and in an ad hoc manner referred to as “kleptocratic liars, lacking the political will to conduct reform.” Moreover, principal-agent theory provides an analytical framework for modeling leadership behavior as a response to the opportunity and incentive structure leaders face. As such, within the framework of principal-agent theory, political will is not understood in voluntaristic terms, but in accordance with the logic of “bounded rationality.”Footnote 9
Yet, while the tools offered by principal-agent theory have the potential to take us a long way towards a more theoretically informed analysis of political will, by focusing almost exclusively on the administrative and non-political aspects of development and reform, it fails to address the broader political and social context within which principal-agent relationships are enacted. This is, we argue, where political science and, more specifically, state theory should enter the picture. With its focus on the ways in which broader social and political forces—and in particular the character of underlying social contracts—shape and constrain the behavior of both “principals” and “agents,” state theory has the potential to be a crucial complement to the more micro-oriented principal-agent theory.
In sum, the analytical framework we develop suggests a new way of analyzing political will that better describes and explains leadership behavior and, thus, development outcomes, than existing approaches. While principal-agent theory offers the tools necessary for analyzing leadership behavior at the micro-level, state theory provides important insights regarding the more systemic factors that shape leadership behavior. We argue that, in tandem, these two perspectives provide an analytical framework for understanding leadership behavior that—with its proper labeling of observed leadership behavior, as well as through its focus on causality and context—is considerably less circular and voluntaristic in nature than the prevailing policy discourse on political will.
We then develop this argument further. In the next section, we discuss the limits of the political will approach as put forward in the contemporary policy debate more in detail. On the basis of this critique, we develop the proposed analytical framework, combining principal-agent theory and state theory. In the subsequent section, a short empirical investigation of the relationship between state legitimacy and leadership behavior—using qualitative data on sub-Saharan Africa, as well as cross-country quantitative data developed by Pierre Englebert and Bruce Gilley—serves to illustrate the explanatory power of the suggested analytical framework for understanding leadership behavior. In particular, the empirical investigation reveals the ways in which the character of the social contracts between citizens, on the one hand, and citizens and the state, on the other hand, play an important role in shaping principal-agent relationships and, hence, the opportunity and incentive structure leaders face. We end with a discussion about the implications of the new framework for future development policy and scholarly understandings of leadership behavior, as well as development success and failure.
Important to note before we proceed with the critique of the contemporary policy discourse on political will is that the call for a more causal and contextual understanding of political will should not in any way be interpreted as an excuse for the many human rights violations and cruel misdeeds conducted by predatory and autocratic leaders. Rather, the aim is to develop a more robust theory of leadership behavior precisely by incorporating institutional factors of the kind that are central to theories of political systems. In our view, until a more analytically thorough approach to political will is provided, development reforms will, at best, be unlikely to make any real difference for the large number of individuals—including the group of people who Paul Collier refers to as the “Bottom Billion”—often suffering under the rule of predatory leaders.Footnote 10 At worst, without a more informed, theoretical understanding of leadership behavior, development efforts risk making the situation poor people face even more pressing.
The Analytical Limits of the Discourse on Political Will
Despite its great influence in policy circles, the contemporary use of political will, we argue, lacks the analytical leverage necessary to further our understanding of development and reform outcomes for at least two reasons.Footnote 11
To begin with, the policy discussion on political will tends to be circular in the sense that it commonly fails to provide any other explanation to why a particular leader fails to implement reform than that he or she lacked any “genuine” intent to conduct reform to begin with. As such, it is at least implicitly assumed that leaders who promise reform but do not succeed to make any real political change are “liars.” This line of thinking is, for example, illustrated by the fact that while a large number of analysts praised the new cadre of bright and young African leaders coming to power after independence, as well as their declared intent to pursue political reform, as soon as those reforms failed to materialize the very same leaders were blamed for lacking the necessary political will. For example, as remote as it may seem today, Mobutu Sese Seko—former president of Democratic Republic of Congo (then Zaire) and today commonly held as one of the most inherently corrupt leaders in the world—was in many analyses in fact once held forward as a “savior” who would lead his country away from chaos.Footnote 12 When seizing power, Mobutu declared his political will to conduct reform by giving a promise that he would remain in office only for five years. Moreover, part of his outspoken mission with his rule was to ensure law and order, political unity, as well as effective administrative services. However, today, these promises are widely dismissed as mere rhetoric, corresponding to the general picture of African leaders as having “grown skilled at enacting the letters of reform while sabotaging its spirit.”Footnote 13
Many more examples of retrospective reasoning can be found in relation to more recent discussions on, in particular, Africa. In the mid-1990s, following the toppling of Mobutu, the international donor community seized upon the notion that a group of “renaissance leaders” had emerged who would apply “African solutions to African problems.”Footnote 14 For example, Isaias Afwerki in Eritrea, Meles Zenawi in Ethiopa, Olusegun Obasanjo in Nigeria, Mwai Kibaki in Kenya, Yoweri Museveni in Uganda, and Paul Kagame in Rwanda have all been described as such renaissance leaders. In the introduction of his best-selling book The End of Poverty, the influential American economist Jeffrey Sachs described these leaders as a “new generation of democratic leaders” pointing the way ahead with their “powerful and visionary leadership.”Footnote 15 African leaders painted a similar picture of their own leadership, maybe most visibly by denouncing the corruption of their predecessors and giving a promise to clean things up as part of the justification for their claim on office. Yet, today, when a clear majority of those leaders' credentials have become severely tarnished, the tune of analysts has changed radically. For example, while hopes that Kenya's president Kibaki's 2002 election would mark a new era in Kenyan politics and a real effort to provide good government were extremely high within the donor community,Footnote 16 in the light of Kibaki's involvement in high-level corruption scandals only shortly after the elections, today few analysts would admit they ever thought that Kibaki ever had any real intention to fight corruption.Footnote 17 Similarly, while many analysts and donors actually believed Uganda's president Museveni's promise to stay in power only for two terms within the framework of his “no-party democracy,”Footnote 18 in the light of the changes he later made in the constitution—allowing him to still be in power—few analysts would now hold that he ever intended to fulfill his promises.
Thus, the contemporary policy discourse on political will has a tendency to engage in post hoc reasoning in the sense that political will is measured by its outcome ex post. That is, the political will argument is tautological in the sense that, since there is no exogenous measure of political will, as well as a lack of any valid theory that can help us predict the chances that rhetoric will turn into reality, the absence of reform is seen as proof of lack of political will, while successful outcomes in turn constitute proof of its existence.Footnote 19 Similarly, to the extent that leaders have promised reform, but failed to implement it, the initial statements of these leaders are commonly disregarded as “mere rhetoric.”Footnote 20 As argued by Derick W. Brinkerhoff, the discussion on political will simply seems to lack the analytical tools necessary for distinguishing between when leaders will or will not take certain actions from when they can or cannot. In the end, we thus agree with Morris Szeftel as he argues that while it is easy to attribute unfulfilled reform promises to the lack of any real intention to actually implement reform in the first place, “unless we assume that everyone in politics is simply a liar,” we should at least be open to the possibility that failed reformers maybe do not necessarily lack the political will to implement reform but rather the ability.Footnote 21
Yet another problem with the contemporary policy discourse on political will is that it is voluntaristic in the sense that it repeatedly tends to end up with the conclusion that some countries are for some reason per se plagued by a larger number of kleptocratic and predatory leaders than others, offering no deeper analysis of why this is the case. This line of reasoning is, for example, evident in Robert Rotberg's writings. According to Rotberg:
Venal leaders are the curse of Africa. If sub-Saharan Africa is “in a mess”, to quote Julius Nyerere, Tanzania's founding president, it is a mess made by its leaders…. where visionary leadership lifted Asia up out of poverty since the 1960s, too many African heads of state in the same period presided over massive declines in African standards of living while carefully enriching themselves and their cronies.Footnote 22
In a series of articles in Foreign Affairs, Rotberg further develops the argument that the underdevelopment of the African continent is much to be blamed on its inherently poor leaders. In fact, according to Rotberg, the underdevelopment of at least 90 percent of sub-Saharan African nations can be attributed to “poor, even malevolent, leadership; predatory kleptocrats, military-installed autocrats, economic illiterates, and puffed-up posturers.”Footnote 23 More specifically, Africa's bad leaders:
… use power as an end in itself, rather than for the public good; they are indifferent to the progress of their citizens (although anxious to receive their adulation); they are unswayed by reason and employ poisonous social or racial ideologies; and they are hypocrites, always shifting blame for their countries' distress.Footnote 24
Blaine Harden—former bureau chief for the Washington Post in sub-Saharan Africa—draws a similar caricature of the typical African leader:
If you took a quarter-century worth of His Excellences the African leader and tossed them into a blender, you would come up with a Big Man who looks like this. His face is on the money. His photograph hangs in every office in his realm. His ministers wear gold pins with tiny photographs of Him of their lapels of their tailored pin-striped suits. He names streets, football stadiums, hospitals, and universities after himself…. He insists on being called “doctor” or “conqueror” or “teacher” or “the big elephant” or “the number-one peasant”, or “the wise old man” or “the national miracle” or “the most popular leader in the world”…. He rigs elections. He emasculates the courts. He cows the press. He stifles academia…. He is—and makes sure that he is known to be—the richest man in the country…. He uses the resources of the state to feed a cult of personality that defines him as incorruptible, all-knowing, physically strong, courageous in battle, sexually potent, and kind to children…. His rule has one overriding goal: to perpetuate his reign as Big Man.Footnote 25
In short, in the contemporary policy discourse, the leaders that inherited power from the colonialists are commonly described as “a sort of little greedy caste, avid and voracious, with the mind of a huckster, only too glad to accept the dividends that the former colonial power hands out.”Footnote 26
The few development successes that in fact exist in Africa and other developing regions of the world are, on the other hand, often directly attributed to “striking examples of effective … leadership in recent decades” or, in the words of Arnold C. Harberger; “a handful of heroes.”Footnote 27 More specifically, the reason why some developing countries have been comparatively successful is that they have simply been blessed with leaders that “stand out because of their strength of character, their adherence to the principles of participatory democracy, and their ability to overcome deep-rooted challenges.”Footnote 28 The developments in Hong Kong and Singapore provide an excellent example of this logic. From the 1950s to the mid-1970s, corrupt rule was nothing short of standard operating procedure in any of these countries. Yet, today the situation is very different; Singapore and Hong Kong stand out as two of the least corrupt countries in the world. This, in turn, is almost exclusively attributed to the deft guidance by the leadership of especially the former Prime Minister Lee Kuan Yew in Singapore and the government-led Independent Commission Against Corruption in Hong Kong.Footnote 29 In the African context, Botswana and Mauritius are the two countries most frequently put forward as examples of where inherently benevolent and farsighted leaders have been decisive for shaping comparatively prosperous societies. According to Rotberg:
It is Botswana's history of visionary leadership, especially in the years following independence, that best explains its success. Sir Seretse Khama, Botswana's founding president, came from a family of Bamangwato chiefs well regarded for their benevolence and integrity…. Modest, unostentatious as a leader, and a genuine believer in popular rule, Khama forged a participatory and law-respecting political culture that has endured under his successors, Sir Ketumile Masire and Festus Mogae.Footnote 30
Similarly, Mauritius' first leader, Sir Seewoosagur Ramgoolam, is generally perceived to have held on to the same leadership codes as Khama. According to this argument, Ramgoolam gave Mauritius a robust democratic beginning, which has been sustained by a series of “wise successors” from different backgrounds and parties.Footnote 31 In fact, if we are to believe this argument, both Khama and Ramgoolam could clearly have followed many of their contemporaries by establishing strong, single-man, kleptocratic rule. But, as argued by Rotberg, they “refused to do so.” That is, often implicit in the descriptions of leaders is the assumption that they could always do differently if they only wanted to. That is, the contemporary policy discussion on political will clearly builds on a voluntaristic view of leaders, making any further analysis of the origins of their behavior redundant. This line of reasoning for example explicitly shines through in Rotberg's claim that unfortunately “too few African leaders have followed the examples of Mandela, Khama, and Ramgoolam.”Footnote 32
We do not question the argument that some regions and countries—such as, for example, sub-Saharan Africa—are in fact vastly overrepresented when it comes to bad leaders, seemingly lacking any political will whatsoever to conduct developmental reform. In fact, the past decades even provide ample and distressing evidence of especially African leaders conducting outright anti-developmental reform in the form of, for example, “politics of the belly,”Footnote 33 “patrimonial politics,”Footnote 34 corrupted or predatory rule,Footnote 35 “criminalization of the state,”Footnote 36 as well as other types of massive human rights violations. Neither do we question the argument that some failed reformers are nothing but simple, inherently kleptocratic “liars” and that at least some successful reformers closely resemble “saints.” For example, it would certainly be a stretch to suggest that South Africa's former president Nelson Mandela and Zimbabwe's president Robert Mugabe are even remotely guided by a shared moral standard. However, this being said, as this section has so far revealed, the contemporary policy discourse on political will still fails to provide any convincing analytical leverage when it comes to our understanding of the behavior of leaders. Thus, we agree with Lori Ann Post, Amber N.W. Raile, and Eric D. Raile who argue that the popularity of explanations emphasizing political will is based on the fact that the concept is “ideal for labeling political failures when the diagnosis is unclear.”Footnote 37 For political will to be useful as an analytical concept, as well as for the design of international interventions and policy prescriptions, we suggest it be firmly integrated into a more systematic analytical framework, combining insights from principal-agent theory and state theory.
A New Framework of Analysis
In historic events, the so-called great men are labels giving names to events, and like labels they have but the smallest connection with the event itself. Every act of theirs, which appears to them an act of their own will, is in a historical sense involuntary and is related to the whole course of history and predestined from eternity.
Leo Tolstoy, War and Peace Footnote 38While the argument we put forward is not entirely consistent with Leo Tolstoy's view that leaders are ultimately little more than historical pawns, predestined by eternity, we still share Tolstoy's critique directed towards the view of leaders as being almost completely unconstrained by their surrounding environment. Moreover, the citation in a nice way illustrates our point that, as applied in the contemporary policy debate on development and reform, the argument that political will and leadership matter for outcomes can only be evaluated in retrospect.
In an attempt to make the discourse on political will less circular and voluntaristic and, hence, more viable as a tool of analysis and policy making, we suggest it should engage in more active dialogue with principal-agent theory combined with what we refer to as state theory. The use of these two theoretical frameworks in tandem has the potential to add the agency and structure necessary for a less circular and voluntaristic analysis of leadership behavior and, thus, of development and reform to the discourse on political will. More specifically, such a synthetic approach can contribute insightful answers to questions such as “How do particular leaders become leaders in the first place?” and “What are the incentives, temptations, and constraints shaping the behavior of leaders?” In short, in tandem, the two perspectives lay the foundations for a more analytical and systematic discussion on how to understand the selection of leaders, as well as the uses of their power. In particular, principal-agent theory provides the tools necessary for analyzing leadership behavior at the micro level, while state theory offers the broader social and political context, within which principal-agent relationships and, subsequently, leadership behavior can be understood.
By highlighting different aspects of leadership behavior and development dynamics, the two theoretical frameworks also contribute significantly to each other. While state theory is strong in terms of theorizing about the more fundamental aspects of development and reform, emphasizing how broader social and political forces affect the opportunity and incentive structure leaders face, it fails to provide a more distinct model of the micro-level logic of leadership behavior. Principal-agent theory, on the other hand, is strong in terms of modeling leadership behavior at the micro level while it fails to recognize the causal importance of the broader social and political context.
Analyzing Political Will from the Perspective of Principal-Agent Theory
The principal-agent model has the potential to contribute significantly to our understanding of leadership behavior in at least two ways. To begin with, it offers the analytical framework for thinking about leadership behavior not as an “error term” or a “black box,” but as a rational response to the opportunity and incentive structure leaders face. In addition, the principal-agent framework provides theoretically anchored tools for labeling what is now commonly brushed aside and in an ad hoc manner referred to as “kleptocratic liars, lacking the political will to conduct reform.”
The principal-agent theory is a rational choice model of human behavior that describes a situation in which a so-called “principal” delegates to a so-called “agent” the authority to perform a certain task. It is based on two key assumptions. The first is that the principal and the agent are assumed to have different—and often conflicting—goals. The second is that agents have more information than the principals, which results in an information asymmetry between the two groups of actors. As a consequence, principals will not know whether the delegated task will actually be performed in line with the principals' interest.Footnote 39 In the end, the implication of the theory is that, insofar as the principal is not able or willing to effectively control the agent, the agent will act in his or her own self-interest and, hence, betray the principal's interest. This is because, in the absence of formal monitoring and screening mechanisms such as, for example, accountability mechanisms, transparency, and an effective regulatory institutional framework, the agent's benefits of breaching the contract are assumed to outweigh the costs.
The asymmetric information between the principal and the agent is, in turn, assumed to have two dimensions. First, there is a pre-contractual informational asymmetry, more specifically called the problem of hidden information or adverse selection. Adverse selection arises as a consequence of the fact that there are different types of agents (meaning that agents differ in their general levels of, for example, ability, effort, and honesty) and that agents have an obvious informational advantage over principals in evaluating their own particular skills. The existence of an informational advantage on behalf of agents is, in turn, expected to attract a disproportionate number of low-quality applicants, aiming to use the informational advantage for their own benefit.Footnote 40 In the end, since principals cannot know the true “type” of the varied candidates in the pool of potential agents—and since there is a large number of “lemons” on the market due to informational asymmetry—there will be greater risk of adverse selection.Footnote 41 Secondly, once the contract is agreed upon, a post-contractual informational asymmetry arises in the form of so-called hidden action or moral hazard. This means that, since the agent's actions are to some extent unobservable by the principal, principals cannot be sure that agents are actually carrying out the delegated task once they are selected.
How, then, more exactly, can the principal-agent theory contribute to a less voluntaristic and circular understanding of observed variation in political will? While early applications of the theory focused almost exclusively on issues of corporate governance,Footnote 42 today—popularized mainly through the work of Susan Rose-Ackerman and Robert Klitgaard—the principal-agent framework is used also for the analysis of bureaucratic and political processes.Footnote 43 If the model is applied to bureaucratic processes, rulers are usually understood to be the principal and the bureaucracy the agent, while, if applied to political processes, citizens are instead depicted as the principal and rulers as the agent.Footnote 44 Taking the point of departure in this reasoning, the principal-agent model can contribute significantly to our understanding of variation in leadership behavior in at least two ways. To begin with, by modeling leadership behavior as a rational response to the prevailing opportunity and incentive structure, the principal-agent theory facilitates an analysis of political will that reaches far beyond a voluntaristic view of leaders. More specifically, if we apply the logic of the principal-agent framework, lack of political will can be conceived of as the result of an information and interest asymmetry between the leader (in this case the “agent”)—and the citizens (in this case the “principal”).Footnote 45 In particular, the theory posits that the behavior of leaders is guided by the opportunity and incentive structure they face, and that this opportunity and incentive structure should, in turn, be expected to vary considerably depending on the degree to which the citizens are willing and able to control the leaders. In societies in which citizens take on the role of being a “principled principal,” leaders should be expected to have fewer opportunities and incentives to act opportunistically compared to societies in which the citizens are not willing or able to take on such a role.
Moreover, the principal-agent framework provides theoretically anchored tools for labeling what is now commonly brushed aside and in an ad hoc manner referred to as “kleptocratic liars, lacking the political will to conduct reform.” More specifically, if we apply this framework of analysis, what is commonly labeled “lack of political will” could in fact be understood as two analytically distinct phenomena—either moral hazard or adverse selection. More precisely, failed reform would from this perspective be understood as a moral hazard problem rather than the outcome of some leaders for some reason being “simple liars.” That is, instead of adopting an idiosyncratic and ad hoc analysis of leaders, positing that some leaders are for some reason per se more prone to breaking their promises than others, the principal-agent framework would rather suggest that all leaders are to some extent corruptible. What differentiates countries in which leaders fulfill their reform promises from countries in which leaders break their reform promises is hence the extent to which leaders have the opportunity and incentive to engage in moral hazard. Moreover, the principal-agent framework would suggest that the reason why some countries seem to be constantly plagued by bad leadership is not necessarily that some countries are plagued by a larger number of bad leaders per se, but that bad leaders manage to come to power due to adverse selection.Footnote 46
In sum, the principal-agent theory would posit that what we observe in the majority of the world's countries is best understood not as lies—or as culturally predestined, inherent kleptocratic tendencies of political candidates—but as opportunity and incentive structures geared towards moral hazard and adverse selection. While to some perhaps a mere technical detail, we hold that the reframing of the discussion on political will, by applying the terminology used by established social scientific theory, opens up the possibility of a more stringent and consistent analysis of the behavior of leaders.
However, having said this, we still argue that the principal-agent framework fails to take us all the way to a more analytically thorough understanding of leadership behavior. This is because, although the principal-agent theory delivers a large number of important insights regarding the micro foundations of leadership behavior, it is still limited in terms of the capacity to account for the broader context within which principal-agent relationships are enacted. This shortcoming is evident if we consider the outcomes of the vast number of reforms conducted in line with the principal-agent theory's operational counterpart “the good governance agenda,” emphasizing the important role of formal monitoring and screening mechanisms.Footnote 47 In fact, over the last couple of decades, the majority of developing countries have undertaken significant institutional reform in line with the principal-agent framework.Footnote 48 Yet the outcomes of these reforms have been far from uniform. As such, while in some countries reform in line with Weberian ideas about the “ideal” type of bureaucracy, emphasizing the importance of formalized, accountable, transparent, regulatory, disciplined, and limited government has been successful, in other countries similar reforms have had limited, if any, positive effects.Footnote 49 Thus, insofar as the aim is to provide a more thorough analytical approach to political will, there seems to be reason to move beyond principal-agent theory's tacit assumptions of development as an exclusively administrative and managerial process, completely stripped of politics. In this endeavor, we argue, state theory has a lot to offer.
State Theory and Leadership Behavior
Quite different from the principal-agent framework of analysis, state theory moves beyond the assumptions of an ahistorical and static relationship between so-called principals and agents, and explicitly recognizes the ways in which principal-agent relationships are enacted in a broader social and political context (even though other terms are commonly used).Footnote 50 As developed by scholars such as Robert H. Bates, Pierre Englebert, Peter Evans, Jeffrey Herbst, Atul Kohli, Margaret Levi, Joel S. Migdal, and Douglass C. North, state theory hence puts forward the idea that development and reform are highly political and conflict-ridden processes.Footnote 51 At the core of state theory is an emphasis on the social dynamics of states. In particular, scholars within this field draw attention to the underlying legitimacy of states, arguing that states have to be defined and analyzed not only in a strict “Weberian” sense (i.e., in terms of their physical base and formal institutional expression) but in terms of “the capacity of the state to command loyalty—the right to rule.”Footnote 52 A state with the capacity to command loyalty is, within this framework, understood as a legitimate state—i.e., a state in which a social contract between the citizens and the state exists, upon which the citizens collectively and (quasi-)voluntarily agree.Footnote 53 That is, a state with the capacity to command loyalty is built upon what we refer to as a shared social contract.Footnote 54 As such, it is widely perceived as an instrument of collective action oriented towards providing public goods.Footnote 55 In this sense, a legitimate state differs radically from, for example, so-called neo-patrimonial state systems since in such systems the general expectation is not that the state should serve the public good, but that the state should serve as an instrument for the self-enrichment of rulers.Footnote 56
Constituting a tacit agreement about “the rules of the game,” social contracts can ultimately be understood as institutions. As such, they constitute incentive systems or, in the words of Douglass C. North, “the humanly devised constraints that shape human interaction,” providing individuals with the cognitive, coordinative, normative, and informational micro-foundations of behavior as they enable, guide, and motivate social actors to follow a specific behavior.Footnote 57 Such institutional constraints thus have far-reaching consequences for the behavior of societal actors. In particular, in line with the tenets of state theory, we should expect leaders to be less likely to satisfy the needs of the broader population—and more likely to resort to patronage, nepotism, corruption, and other patterns of political behavior that are occasionally subsumed under the category of “neopatrimonialism”—the less legitimate is the state over which they rule.Footnote 58
More specifically, in terms of moral hazard, the literature on state theory puts forward at least two potential mechanisms that should lead us to expect that leaders of states lacking a shared social contract will be more likely to resort to bad governance in the form of corruption, clientelism, nepotism, and other forms of factionalism, even if in the beginning of their mandate they promised and seemed to be truly committed to reform.Footnote 59 The first mechanism is a lack of a coherent and well-defined public, which significantly reduces the degree of potential civic monitoring and control of the ruling elite. That is, in states lacking a shared social contract, leaders face what principal-agent theorists would call multiple principals. Facing multiple principals, the argument goes, agents may receive different and contradictory messages and expectations from their principals. This, in turn, risks increasing the informational asymmetries and the difficulties of monitoring. That is, what we may observe is that the principals (i.e., the citizens) fail to co-ordinate their strategies, either because they do not share common experiences and interests, or because they cannot commit to collaborate. The effect is severe weakening of the power of the incentive schemes provided to the agents (i.e., the leaders). In the end, the more the principals' interests diverge, the more room for the agent's discretion, and the less effective monitoring and control of the agents. Together, the principals tend to cancel each other out, so the agent is at best confused and demoralized, or at worst feels he or she can get away with doing nothing to promote the collective good.Footnote 60 In fact, in some instances the lack of a shared social contract not only limits the degree of monitoring and control in the system, but even leads to a situation in which the ruling elites are punished for serving the public good.Footnote 61 In such settings, the state is conceived merely as a potential resource to be appropriated, or an instrument to be used in the domination of other groups.Footnote 62 The result, as Michela Wrong describes, is a system in which competing elites alternately take power only to serve their own group's interests according to a logic of that “it's our turn to eat.”Footnote 63 That is, politics invariably turns into a zero-sum game, with one group's gain inevitably entailing another group's loss.
Such a zero-sum conception of politics relates directly to the second mechanism limiting political accountability—a shorter time horizon of leaders, who discount the future more heavily since there is no (credible) way of compensating ex post the political elites who lose their power.Footnote 64 Thus, leaders that face multiple principals should be expected to be more likely to resort to policies aimed at short-term self-enrichment. In line with this reasoning, unfulfilled reform promises—that is, moral hazard—are not necessarily best understood as a result of particularly malevolent leaders, but as a rational—and in some circumstances in fact even the only possible—response to the conditions under which they govern. In particular, moral hazard in the form of unfulfilled reform promises could from this perspective be understood as “adaptations” by leaders who govern in the context of constant threats stemming from the aspirations of other groups to “eat in office.”Footnote 65 Joel Migdal even refers to such practices as “politics of survival,” referring to the ways in which leaders' lack of political will to conduct reform at least sometimes reflects the constant threat they face rather than a simple “lie.”Footnote 66
Since the opportunities and incentives for leaders' self-enrichment are generally considerably larger in states where no unified principal exists to monitor and screen actual and potential agents, such systems should moreover be expected to attract more “lemons”—actors having an interest geared towards enriching themselves at the expense of others—to the political market. That is, the risk of adverse selection should be expected to be considerably greater in states lacking a shared social contract than in states having such a contract. In sum, the opportunity and incentive structure in states lacking a contract between the principal and the agent upon which the principals (quasi-)voluntarily and collectively agree should be expected to be highly vulnerable to both moral hazard and adverse selection. In the next section we will illustrate this logic empirically.
An Empirical Outlook
The dynamic of weak states and severe principal-agent problems is vividly illustrated in sub-Saharan Africa, the region most commonly considered to be severely plagued by bad leaders lacking the political will necessary to conduct reform. Compared to regions in which political will is more commonly observed, sub-Saharan Africa harbors the largest number of “failed states” lacking a social contract. As such, sub-Saharan Africa's forty-odd states are among the weakest in the world, and severe political instability—as indicated by coups, plots, and civil wars—has been more prevalent in this region than anywhere else.Footnote 67 Indeed, numerous African states—most notably Angola, Chad, Ethiopia, Nigeria, Sudan, Uganda, and Democratic Republic of Congo—have periodically ceded so much control of their country's territory and population that they arguably have ceased to be what Jackson and Rosberg refer to as “empirical states.”Footnote 68
Scholars such as Pierre Englebert, Richard Sandbrook, and Jeffrey Herbst trace this weakness of sub-Saharan African states to the earliest phases of modern state formation.Footnote 69 Compared to other regions, state creation in sub-Saharan Africa was by and large an exogenous process determined largely by European colonialism. Africa's post-colonial leaders simply inherited states rather than shaping them as instruments of an emerging national hegemony. As a result, most sub-Saharan African countries were born lacking legitimacy and, very soon after gaining independence, the majority of African leaders realized how truly little power they had inherited. The true challenge to their rule came not merely from highly diverse societies, but from the competing institutional claims to sovereignty advanced by multiple principals with conflicting interests inhabiting the same territory.Footnote 70 As Clifford Geertz wrote upon observing the lack of integration of the “new states” of the Third World, “we have not just competing loyalties, but competing loyalties of the same general order.”Footnote 71 As a result, a large number of citizens and political leaders alike came to conceive of the post-colonial African state not so much as the outcome of a social contract, an instrument of collective action based on “common ideological convictions”Footnote 72 or a shared national identity, but rather as an “alien” institution to be appropriated.Footnote 73 As described by Donald L. Horowitz, “everywhere the word domination was heard. Everywhere it was equated with political control. Everywhere it was a question of who were the ‘real owners of the country’ and of who would rule over whom.”Footnote 74
A brief look at cross-sectional data on the relationship between the character of social contracts and political will reveals that the explanatory power of the suggested analytical framework goes far beyond the case of sub-Saharan Africa. While the character of social contracts does not lend itself to easy measurement, Pierre Englebert's and Bruce Gilley's respective measurements of state legitimacy constitute fairly close proxies. Englebert's measurement takes as a point of departure his observation of the close relationship between the process of modern state formation and state capacity today.Footnote 75 As such, it builds on Kalevi Holsti's idea of “vertical legitimacy,” which refers to the strength of the relationship between society and political institutions.Footnote 76 Englebert measures state legitimacy on the basis of the degree to which a particular state's structures have evolved endogenously to its own society and the degree to which there is some level of historical continuity to its institutions. The measure is built on a series of five dichotomous outcomes, including whether or not a country was colonized; whether or not the country recovered its previous sovereignty upon independence if colonized; whether or not the country was created by colonialism and, if so, whether or not there was human settlement before colonialism; whether or not a civilization predating colonization was physically eliminated or marginalized in the process of colonization if colonized; and, finally, whether or not the post-colonial state in previously colonized countries exercises severe violence against the pre-existing political institutions. When these criteria together suggest a discontinuity in the evolution of state institutions—a “clash” between pre-colonial and post-colonial political institutions—then a country scores zero on the legitimacy variable. When there is no conflict between imported and domestic institutions, then a country scores one.Footnote 77
For Bruce Gilley, “a state is more legitimate the more that it is treated by its citizens as rightfully holding and exercising political power.Footnote 78 In his continuous measurement of state legitimacy, running from 0–10 where 10 represent a fully legitimate state, Gilley includes both attitudinal and behavioral variables, together covering what he calls the “legality,” “justification,” and “consent” aspects of legitimacy. The 72 countries included in his sample are ranked depending on the degree to which citizens believe that the state respects individual human rights; have confidence in the police; have confidence in the civil service; are satisfied with democratic development, with the operation of democracy, and with the current political system; use violence in civil protest; vote, and; quasi-voluntarily comply with taxes.
A basic cross-country, quantitative analysis of the relationship between Englebert's and Gilley's measures of state legitimacy and the number of opportunities and incentives for moral hazard and adverse selection available to leaders—as proxied by the level of corruption (i.e., “the misuse of public office for private gain”) and as measured by the World Bank's indicator “Control of Corruption”—reveals a strong correlation between the character of social contracts and political will.Footnote 79 More specifically, the correlation coefficient between Englebert's measure of legitimacy and corrupt rule is 0.51.Footnote 80 All of the 30 least corrupt countries in the world are legitimate according to this measure, while 23 out of the 30 most corrupt states are illegitimate.Footnote 81 On the basis of this correlation, we should in other words expect that of the 30 most thoroughly corrupt leaders lacking the political will to conduct reform, only about 7 could per se be assumed to be inherently kleptocratic “liars.” The actions of the other 23 could—at least in theory—be understood either as rational responses to the circumstances under which these leaders govern or, if we stick to the claim that leaders are at least to some extent guided by different moral standards, as the result of adverse selection. There is an equally strong relationship between Gilley's measure of the character of social contracts and political will, the correlation coefficient being 0.71.Footnote 82 The ten most legitimate states in Gilley's sample on average score 1.8 on the World Bank's index, while the ten least legitimate states score −0.56.
In sum, as the analytical framework outlined here suggests, the character of social contracts seems to be quite a strong determinant of whether or not the opportunity and incentive structure of leaders is geared towards moral hazard and adverse selection.
Conclusions: Rethinking the Limits of Political Will
Since the 1950s, countries in the developing world have followed a range of radically different development trajectories. For example, while in the 1950s Ghana, Cambodia, Bolivia, and South Korea were equally poor, today South Korea is 16 times as rich as Bolivia, 33 times as rich as Ghana, and 36 times as rich as Cambodia. Thus, what was before commonly referred to as the “Third World” is now more diverse than before. In fact, while for the last forty years the development challenge has typically been viewed as the disparity between a rich world of one billion people facing a poor world of five billion people, some scholars believe that this way of conceptualizing development will most likely need to be reconsidered in the near future. Instead, the real challenge of development will be, in the words of Paul Collier, “a group of countries at the bottom that are falling behind, and often falling apart.”Footnote 83
While many different factors have been put forward in an attempt to explain the increasing divergence in development patterns across the developing world, political will—or the lack thereof—is one of the factors most frequently stressed in the contemporary policy discussion. In line with this argument, successful development paths are attributed to the existence of political will to reform corrupt institutions, while less successful development outcomes are blamed on the lack of such will. In particular, the successful development paths embarked upon by countries such as Singapore, Hong Kong, Malaysia, Botswana, Mauritius, Brazil, and Chile are commonly attributed to the deft guidance of “a handful of heroes”—truly visionary leaders with the unconstrained power to defy history and define a brighter future for their people. In the same vein, the less successful development of especially many sub-Saharan African countries is often blamed on inherently kleptocratic “liars,” lacking any genuine political will whatsoever to conduct reforms that benefit the public good. That is, within the framework of the contemporary policy discourse, the credit given to the leaders of successful countries is only matched by the blame given to the leaders of less successful ones.
While we do not reject the claim that leaders are indeed crucial in the development process, we argue that the concept of political will as currently applied is of limited practical use and adds little to our understanding of development and reform. In particular, we argue that the contemporary policy discussion on political will lends itself to a circular, as well as a voluntaristic, analysis of leadership behavior. To escape these problems, we suggest that the discourse on political will be integrated into a more thorough analytical framework of analysis, combining principal-agent theory and state theory. In tandem, these two perspectives can add to our understanding of political will what the contemporary discourse lacks, namely a contextual and causal account of leadership behavior across time and space. In particular, the suggested analytical framework highlights the ways in which leadership behavior is to a significant extent conditioned by the character of underlying social contracts. Where a shared social contract is lacking, we should expect the incentive and opportunity structure of leaders to be geared towards moral hazard and adverse selection, making the existence of leaders demonstrating political will less likely. According to the same logic, where a shared social contract exists, the state is thought of as an instrument of collective action rather than a resource to be appropriated. Thus, the opportunities and incentives for kleptocratic rule and other forms of bad governance should be expected to be severely limited. In sum, the suggested analytical framework has the potential to provide a greater understanding of why some countries seem to be constantly blessed with “saints,” while others seem to be as constantly plagued by “tropical gangsters.”
The conceptualization of political will as something being conditioned by underlying social contracts, rather than something emanating from a particular leader's inherent qualities and wishes, has significant policy implications. In particular, our analytical framework suggests that, instead of the international community putting all its faith and hope in new generations of supposedly “responsive and responsible leaders”—and in fact actively supporting the toppling of certain leaders, often naïvely assuming that their successors will have the opportunities and incentives to act any differently—attention should instead be focused on better understanding the contexts in which these leaders function. In particular, policy-makers should pay much greater attention to how the relationship between the rulers and the ruled, as well as the relationship between the citizens, play out in different countries. In places where a shared social contract seems to be lacking, they should concentrate their efforts into helping to construct such a contract.
How, then, can a shared social contract be constructed—if it can at all be constructed? The synthesis of principal-agent and state theories that we propose suggests a general answer. For ultimately, the construction of a shared social contract demands exchanging a negative equilibrium, in which the state is generally perceived as a resource to be appropriated and politics as a zero-sum game which no one can afford to lose, for a positive equilibrium, in which the state is generally expected to be an instrument of collective action and public goods provision. In places suffering from politically-enforced corruption, kleptocracy, and economic decline, entire systems need to be transformed, from what Douglass C. North, John Joseph Wallis, and Barry R. Weingast call a “limited access order” of rent-seeking to what they call an “open access order” based on the rule of law and a structure of economic opportunity.Footnote 84 What is required, in other words, is nothing less than what Larry Diamond has described as a “revolutionary change in institutions.”Footnote 85
Yet, while such general prescriptions might make great sense at the level of theory, the coordinated changes that are envisioned require massive collective action. And, if we are to believe the tenets of institutional theory, we should expect social equilibria—functional as well as dysfunctional ones—to be extremely sticky and, hence, very difficult to change, especially for the better.Footnote 86 At the same time, as some modern examples attest, positive change does sometimes occur. The increasing heterogeneity of African states is a case in point. Indeed, recent “success stories” in places such as Botswana and Mauritius indicate that the character of social contracts can in fact be transformed over time.Footnote 87
At the same time, we must frankly acknowledge that while the social scientific literature is now quite strong in explaining the breakdown of political control and order, we know surprisingly little about what explains the development of effective social contracts (James D. Fearon and David D. Laitin have observed something similar about our understanding of the requisites of peace, noting that while peaceful ethnic relations are far more common than violent, interethnic conflicts, we still know considerably more about the sources of ethnic conflict than about the origins of peaceful relations).Footnote 88 Indeed, we know considerably more about which policies not to pursue than about which policies could actually help promote the development of shared social contracts. In particular, as briefly discussed already, what is increasingly coming through in the contemporary political science literature on development and reform is that the “one size fits all” models of development put forward primarily by economists and the discipline they represent—being intimately related to “New Public Management” and to Weberian ideas about the ideal bureaucracy, and hence promoting reforms focusing solely on the administrative and formal institutional, non-political dimensions of government—should be expected to have only a limited positive effect on the construction of social contracts, if any positive effect at all. In fact, departing from Dankwart A. Rustow's now-classic argument holding that the factors that sustain democracy where it is strong are not necessarily the same as those that brought it into being to begin with,Footnote 89 there is even a growing concern among some scholars that reforms failing to recognize the importance of some underlying social or political cohesion might even risk “kicking away the ladder” of development, undermining the very foundations upon which development success in the industrialized parts of the world was built.Footnote 90 Even more dangerous, according to these scholars, by encouraging (even forcing) developing countries to adopt the organizational and institutional features of industrialized and well-developed states, such reforms might even facilitate the establishment of pure skeletal and façade state structures, stripped of all basic state functions and, hence, providing a vast number of opportunities and incentives for opportunistic behavior.Footnote 91 In the words of Lant Pritchett and Michael Woolcock, “Simply mimicking … the organizational forms of a particular ‘Denmark’—has in fact been a root cause of deep problems encountered by developing countries.”Footnote 92
In the end, the critique directed against development efforts that focus exclusively on the administrative and managerial dimensions of reform very much echoes our critique of the ways the concept of political will is used in the contemporary policy-oriented development discourse. The problem is that the carriers of these discourses fail to take causality and political context seriously. In our view, this theoretical—as well as methodological—failure is to a significant extent the direct result of the fact that the contemporary development agenda has been more or less completely hijacked by economics (and economists), while other social science disciplines have been quiet bystanders to the show.Footnote 93 While we can of course not fully compensate for the relative silence of political science, history and sociology, we can at least pose this first step in overcoming the estrangement of political science in the development debate. As such, this acts as an attempt to keep the discipline alive and relevant for development policy and practice and, hopefully—if yet by extension—even for those suffering under rule of “predatory kleptocrats, military-installed autocrats, economic illiterates, and puffed-up posturers.”Footnote 94
