Policy Research Working Paper 10473 Economics of Political Clientelism and Corruption A Theoretical Pathway Rasit Pertev June 2023 Policy Research Working Paper 10473 Abstract This paper proposes a preliminary economic model of polit- challenge the sovereignty and territorial integrity of the ical clientelism and corruption in developing countries with country concerned. The impact on fragility is predicted weak rule of law. It explains why this corruption is often as greatest during monopolistic and dictatorial clientelism. chronic and persistent, and further examines its impact on Governance structures of inegalitarian or unjust local tradi- fragility, conflict, and violence. The basic model is built in tional authorities are shown to be an important independent three stages: (i) political party strategies vis-à-vis clientelist factor provoking separatism. Separatist movements are pre- options using a game-theoretical approach, (ii) strategies dicted to be left leaning or egalitarian in the beginning of using state repression and violence to complement elec- of their struggle. As economies grow and shift away from toral clientelism, and (iii) strategies of geographical/ethnic a patronage-based private sector toward a productive one, entities on remaining within a given republic or breaking individuals are likely to be protective of their enterprises and away. The model predicts that the first clientelist party in incomes against the aleatory decisions of a clientelist gov- power can monopolize government for long periods and ernment. Therefore, a substantially enhanced investment further consolidate power by blending in state violence. in a productive private sector may likely be a better longer Political clientelism and corruption are likely to provoke term anti-corruption strategy than exclusively focusing on geographically distinct communities and movements to governance, accountability, and accounting measures. This paper is a product of the World Bank Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at rpertev@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Economics of Political Clientelism and Corruption – A Theoretical Pathway Rasit PERTEV 1 JEL codes: P16, P48, C79, B51, O12 Keywords: Governance in Public Sector; Conflict and Fragility; Corruption Prevention; Political systems and Government; Democracy and Growth 1 Rasit Pertev is World Bank Country Manager for Chad. He has formerly served as the Secretary of International Fund for Agricultural Development of United Nations, as well as the Assistant Secretary General of World Federation of Farmers. He was Minister of Agriculture and Forestry, and the Chief Negotiator for the UN Peace talks in Cyprus. He also founded and served as the president of a political party championing good governance. Page 1 of 25 INTRODUCTION Clientelism and corruption in developing countries together with their impact on fragility and peace negatively affect millions of lives and seriously undermine the ongoing development efforts. Despite this vital importance, both aspects are poorly understood within economic theory. The current paper aims to address this gap, to enable development practitioners, economic analysts, and theoreticians to have a better grasp of what is happening. In economic theory, the spoils of government appear within the Samuelsonian paradigm, where the heterogeneity of individual preferences introduce an inherent welfare inefficiency into the system ((Samuelson, 1954; Stiglitz, 1958), especially in the context of a democracy. Differences in individual preferences can enable opportunistic candidates to target specific localities and lobby groups to achieve electoral success, through the promise and provision of pork-barrel projects (Lizzeri and Persico, 2000). Electoral strategies of such candidates, emanating from the non-identical nature of voter preferences, still operate within the law and rule of law, making the model more applicable to high-income economies. The Samuelsonian model becomes less relevant in the context of weak rule of law and is unable to explain the complexity of clientelism observed in developing countries. In a similar vein to the Samuelsonian paradigm, game theoretic approaches have been applied to reforms in political science, where it has been argued that legislative reforms are likely to pass for evenly distributed patronage (Geddes, 1991, 1994). These arguments have also limited themselves to the legislative domain within a functioning rule of law and have not ventured into the domain of a weak or non-existent rule of law. Furthermore, both have not attempted to sketch a possible mechanism for interaction with fragility and peace. In a different framework, corruption has also been analyzed through micro-economic calculations, where individuals decide to be or not to be corrupt based on available opportunities and risks. Proponents have therefore argued that corruption can be reduced by strengthening the legal framework, tightening public procurement, reinforcing public financial management, strengthening anti-criminal policing and judicial measures, institutional strengthening, and policy reforms (World Bank, 2020; Menezes, 2001; Kyarem et al, 2017). Consequently, multilateral and bilateral development agencies currently focus on such features to curb corruption in developing countries. While such measures may be appropriate to curb petty corruption, they are not particularly suitable for dealing with state-led clientelism and political corruption. It has also been argued that as countries move towards democracy, societies strengthen democratic checks and balances as well as their rule of law, resulting in a decrease of corruption (Acemoglu et al, 2003; Acemoglu et al, 2006), while the fixing of democratic imperfections can result in a similar effect in developing countries (Keefer and Khemani, 2005). Contrary to the above, the paper argues that free elections are likely to serve as a convenient vehicle for corruption in the context of a weak rule of law, by establishing clientelist voting Page 2 of 25 patterns and practices. The proposed theory is in line with literature which views clientelism as a viable conduit for corruption (Hicken, 2011) (Acemoglu, 2003). Clientelism is incorporated here as direct contractual arrangements between individuals and political parties. For clarity, the paper will rely on the Rothstein definition of good governance (Rothstein and Teorell, 2007), namely the impartiality of institutions that exercise government authority with respect to citizens. Intrinsic terminological issues when dealing with state-led or political systems-based corruption will be addressed through this definition, as the state has the power to legalize such actions through its legislative and administrative procedures and if needed, through sophistication in this domain. A theoretical model is built in three stages to study the impact of clientelism and corruption on fragility, conflict, and violence in developing countries. First, the model analyzes political party strategies vis-à-vis clientelist options using a game-theoretical approach. Second, it introduces the option of using state repression and violence to complement electoral clientelism. Third, it introduces geographical / ethnic entities and analyzes their option of either remaining within the country concerned or breaking away. The model makes several predictions. First, political parties opting for clientelism will have a definite electoral advantage and all political parties will shift to clientelism over time due to the presence of a Nash equilibrium. Current anti-corruption measures, as practiced by development agencies, will therefore remain superficial and unlikely to work. However, as national incomes increase, this prediction is likely to become weaker through the reduction of high discount rates prevailing in low-income economies as well as a reversal of individual preferences towards public goods. Equally, as economies grow and shift away from a patronage-based private sector towards a productive one, individuals are also likely to be protective of their growing enterprises and incomes against the aleatory decisions of a clientelist government. Therefore, investment in a productive private sector and human capital is a better longer term anti-corruption strategy. Second, the model also predicts that the first clientelist party in power can easily monopolize government for long periods even in the presence of free elections. Furthermore, any clientelist government can further consolidate and stay in power longer by blending in state violence, by supplementing the system of clientelist rewards with the clientelist application of repression and punishment. Impact on fragility and peace is predicted as greatest during such monopolistic and dictatorial clientelism. Third, governance structures of geographical and ethnic entities themselves are also shown to be a crucial factor contributing towards fragility and peace, counting as much as the governance modality of the national government. As an additional point, separatist movements arising are predicted to be left leaning or egalitarian in the beginning of their struggle. The paper further argues that as economies grow and shift away from a patronage-based private sector towards a productive one, individuals are likely to be protective of their enterprises and incomes against the aleatory decisions of a clientelist government. Therefore, a substantially Page 3 of 25 enhanced investment in the productive private sector is likely to be a better longer term anti- corruption strategy than only focusing on governance, accountability, and accounting measures. The model shows that multi-party electoral systems generate political clientelism and corruption when applied in the context of a weak rule of law and especially in absence of full- fledged capitalism. On a historical note, developing countries fulfilled both conditions in the wake of their independence. Most opted for multiparty democracy, recommended by their colonizers, who aimed to ensure the emergence of liberal economies during the Cold War. Furthermore, developing countries had none of the foundations to establish a strong rule of law. Colonialism had wiped their social landscapes clean, fragmented communities, diminished previous institutions, structures and values. As the emerging elite tried to accumulate wealth in the absence of full-fledged capitalism, they had little choice but to plunder their own state and public resources, and in so doing, inhibited economic, social and environmental development in the process. BASIC PREMISES OF THE MODEL In this model, the State will have one (1) unit of public resource available for use each year, with two options: (a) to allocate the totality of this public resource, with each citizen benefiting equally; (b) to allocate a proportion (S) of this resource as clientelist handouts, so that they benefit only the specific individuals concerned. This proportion S, referred to as Clientelist Surplus, is kept constant for the sake of simplicity. No multiplier effects arising from either allocation will be assumed. Private sector earnings of individuals will be considered exogenous to the model. At each electoral cycle, each political party will decide on a clientelist or non-clientelist strategy. A clientelist party will be defined as a political party which uses clientelism and corruption as its main modus operandi in a systematic, rather than an ad hoc manner, for at least one full electoral cycle. Individuals will support and vote for a clientelist political party in exchange for personal gains, guaranteed through direct informal contracts with the party. Contracts will be concluded at the national or local level, under the party leadership, utilizing the fractal nature of party structures. Contracts will include individuals’ obligations such as allegiance, explicit support, political mobilization, networking, carrying out the party’s specific instructions as well as providing monetary or in-kind support. In exchange, they will obtain specific individual gains, based on their loyalty, contributions and efforts, once the party is in power. A non-clientelist party will be defined as a party which aims to convert available state resources into public goods and services to benefit each citizen with impartiality (Rothstein, and Teorell, 2007). Individuals will adhere to and support the non-clientelist parties for promoting and implementing such policies, without expecting or receiving any unfair individual benefit. A clientelist party in power is obliged to deliver its part of the contract. Using deficiencies in rule of law, it will generate rewards for its clientelist voters by forgoing efficiency mainly in four categories: Page 4 of 25 (i) Public Sector Efficiency, including rationing and control of public services to favor party members (Olowu, 2000); inflating and distributing public-sector jobs to party members; basing civil servant rewards on party affiliation, leading to deficient performance and demise of talent; awarding contracts to party members, leading to sub- optimal services. (ii) Market Efficiency, including control and interference in markets leading to unjust privileges to party members and their enterprises; ensuring politicization and control of markets by avoiding complete liberalization; dictating privileged access through client-tailored changes to market rules and regulations, tariffs and import controls, ad- hoc ministerial decisions, allocation of subsidies and permits and controls of market entry; maintaining inefficient state-owned enterprises to provide gains for party members. (iii) Natural Resources, Public Property and the Environment, including acquiring surplus from extractive industries, natural resources, and other externalities; resources of public and parastatal institutions; from issuance of permits inflicting damage to environment, historical and cultural heritage. (iv) Social Welfare, including allocating less of the government budget to social welfare concerns in human capital, health, and education. In a clientelist party, individual rewards will be a function of the place of the individual in the party pecking order. The place in the pecking order will depend on: (a) length of uninterrupted time of allegiance to the party (b) quality and quantity of work contributed by the individual; (c) political “weight” based on the person’s network and sphere of political influence. Voters will be assumed to be fully aware of the policies of each party and their respective place in each party’s pecking order should they choose to support that party. In calculating their expected benefits from a political party, each person calculates the following: (a) voter expectations regarding the probability of electoral success of the party, which will be assumed to be uniform for the sake of simplicity; (b) voter expectations regarding the honoring of clientelist contracts by the political party, based on the party’s previous track record in this matter. Clientelist voters also know the clientelist reward they are expected to receive, based on (i) their contribution to the party; (ii) their place in the party pecking order, directly related to their loyalty, contributions and political weight. ELECTORAL COMPETITION: CLIENTELISM OR NON-CLIENTELISM? At the beginning of each electoral cycle, parties will decide on a clientelist/non-clientelist strategy. Parties opting for clientelism will then conclude individual clientelist contracts with Page 5 of 25 voters. Perceived chances of the electoral success of the parties A and B will be α and (1-α) respectively. The payoff matrix is given as Matrix 1 below. Non-clientelist party offer The non-clientelist party will propose to allocate the totality of resource available as public goods and services, to benefit each citizen equally. The total resource available is 1, the benefit 1 accruing to voter I is , where = , the total number of voters being n. Clientelist party offer Clientelist parties will extract a clientelist surplus S. Benefits to be received by citizens as public goods and services in an equal manner will therefore decrease from a value of 1 to a value of 1-S. Non-party members will only receive the reduced public goods and services, whereas clientelist party members shall additionally receive a clientelist reward. Benefits to non-party members shall be as follows: Benefits to voter i, non-member ( , ) = 1− of clientelist party, denoted by o ............................................... (1) Party members will additionally receive clientelist benefits: Clientelist reward to voter i, ( , ) = (i) ...................................... (2) from Clientelist Party P where is a function for clientelist rewards for Party (P) based on pecking order, such that: +1 2 ∑0 (i) = ............................................................................................ (3) Where a minimum of ( 2 + 1) number of voters will be necessary to win an election. Benefit accruing to voter i, member of the clientelist party in power (denoted as c) will be: 1− , = + (i) .................................................................................... (4) To secure the vote of ( 2 + 1) number of voters, the benefit received by the marginal voter 1 must be greater than the benefits offered by the non-clientelist party, being = 1− 1 , /2+1 = ( 2 + 1) > + .............................................................. (5) Page 6 of 25 Within these constraints, the party also needs to maximize the average of rewards received by its voters for a given S. The number of clientelist voters appropriated and rewarded by the party must therefore not exceed ( 2 + 1) number of voters. In this case, the undecided voter, for which the offers of clientelist and non-clientelist parties equalize is then the voter i, where i is ( 2 + 2). 1− 1 , +2 = + (2 + 2) = .................................................................. (6) 2 Such that; 1− 1 , = ( ) > + .............................................................................. (7) ( ) > ...................................................................................................... (8) For all values of i where i ≤ ( 2 + 1). Pay-off matrix Out of the possible four scenarios, two scenarios involve an electoral competition between same-strategy parties. By definition, a competition between two clientelist parties will result in a clientelist victory and the one for two non-clientelist parties, in a non-clientelist victory. and non- clientelist in the case of two competing non-clientelist parties. Regarding the competition, a clientelist and non-clientelist party, however, Voter i chooses between the clientelist and non-clientelist party depending on expected gains. If Party A is clientelist, where α is the probability of success of Party A, where (0 ≤ α ≤ 1): Expected gain of = (1 - α). Gain in case of a + (α). Gain in case of a Clientelist Voter i in case of Non-Clientelist Win Win voting for Party P If the voter i remains in the non-clientelist Party B, expected voter gains will be: Expected gain of = (1 - α). Gain in case of a + (α). Gain in case of a Voter i Non-Clientelist Win Clientelist Win 1 1− = (1 – α).( ) + α.( ) ...........................................(9) (1− αS) = ..........................................................(10) For the voter who opts for a clientelist contract with Party A, expected gains will be: Expected gain Voter = (1 - α). Gain in case of a + (α). Gain in case of a i where i ≤ ( 2 + 2) Non-Clientelist Win Clientelist Win Page 7 of 25 1 1− = (1 – α).( ) + α.( ) + α. (i) ...........................(11) (1− αS) = + α. (i) .....................................................(12) 1 Combining Equation (8) with (12), the expected gain of Voter i is greater than , where i ≤ ( 2 + 1), implying that ( 2 + 1) voters will vote for clientelist party, thus securing an electoral victory. Clientelist party will thus be the winner in a Clientelist versus Non-Clientelist Scenario. The same outcome holds true for the reverse scenario, Party B being the clientelist and Party A being the non-clientelist party. The payoff matrix resulting from an electoral competition between Party A and Party B is given in Matrix 1 below. For both parties, clientelist strategy is the dominant strategy, resulting in a Nash equilibrium, with both parties choosing the clientelist option. At that point, neither party would be able to do better by shifting to non-clientelism. On the other hand, both parties choosing non-clientelism is an unstable equilibrium, as either party can do better by shifting to clientelism. Matrix 1 PARTY B STRATEGY Non-Clientelist Clientelist Non- α , 1-α 0, 1 Clientelist (Non-clientelist Win) (Clientelist Win) PARTY A 1, 0 α , 1-α Clientelist STRATEGY (Clientelist Win) (Clientelist Win) The model predicts a high prevalence of clientelism in countries where rule of law is deficient, enabling the extraction of a clientelist surplus. Accordingly, (i) those parties that opt for clientelism will have a definite electoral advantage over non-clientelist parties; (ii) all political parties will become clientelist over time due to the presence of a Nash equilibrium. COMPETITION AMONG CLIENTELIST PARTIES In an electoral competition between two clientelist parties, the first variable, denoted by α in the model, is voter expectations of each clientelist party’s chances of winning. A party can try to influence this variable through communication campaigns including showing themselves to be powerful as a party as well as running effective negative campaigns against their rivals. Prior electoral success and time in office are also important contributors. Page 8 of 25 The second variable, denoted by Δ in the model, is the track record of the clientelist party in honoring its contracts. To have a high Δ, parties need to honor their clientelist contracts. However, any party which has not been in power would have non-existent track record and as a result, a low Δ. The third variable is the number of clientelist contracts targeted by the party. A clientelist party, may either aim to conclude clientelist contracts with the minimal requirement of ( 2 + 1), which would be just enough to win an election, or may aim to conclude a larger number of clientelist contracts, beyond the minimal requirement of ( 2 + 1). These two options can be applied to two parties, with Party A adopting the first option of targeting a minimum number of contracts, and Party B adopting the second option of targeting a greater number of contracts than Party A. If both parties are applying egalitarian distribution, ceteris paribus, all members within each party receiving an equal amount of the distributed surplus: 1− , = + ...................................................................................... (21) ( )+1 2 1− , = + ...................................................................................... (22) ( )+ 2 , will be greater than , for all values of t greater than 1. Clientelist Parties would therefore conclude only the minimal number of contracts, being ( 2 + 1) to win an election. The fourth variable is the shape of the curve (i), for distributing clientelist rewards, based on the place of the individual (i) in the pecking order within that party. The state of perfectly equal distribution of surplus shall be denoted as State 0 . In this state, the party concludes a minimal number of clientelist contracts, being ( 2 + 1) just enough to win the elections, and each member will receive an equal amount, being: 1− , = + ........................................................................................ (21) ( )+1 2 Starting from State 0 , the state where one person is allocated more than the rest of the members will be denoted as State −1 . At State −1 , at least one member must be allocated less than the rest of the members, compared to state 0 , so that one person may receive more than other members, such that the total of surplus distributed is maintained at the fixed total amount of . As such, State 0 will be electorally more competitive than State −1 . Similarly, if State − is the state where x persons are allocated more than the rest of the members, and where x members are allocated less than others to compensate, then, through the same reasoning, State − will be more competitive than State −(+1) . Thus, more egalitarian surplus distribution curves will be electorally more competitive compared to less egalitarian ones, provided that all other variables, such as α and Δ are the same for competing clientelist parties. Page 9 of 25 In cases where a party holds the advantage in terms of α and Δ, is also able to exercise a degree of unequal distribution among its members and still maintain competitivity, provided that no member in Party A receives less than min as below: 1− min = + .................................................................................. (24) ( )+1 2 where: = Electoral success of Party P = Track record of Party P in honoring its contracts and > and > In summary, clientelist parties are most competitive when they (i) aim at a minimal number of contracts to win the elections; and (ii) exercise an egalitarian surplus distribution among their members, if α and Δ are the same for all parties. However, parties having an advantage in α and Δ values can exercise a more unequal distribution without losing their electoral competitivity. SPECIAL CASE OF MONOPOLISTIC CLIENTELISM The model predicts the existence of a special case, termed Monopolistic Clientelism, where a clientelist party acquires a significant comparative advantage over other parties, enabling it to monopolize government over several electoral cycles. The first clientelist party in power at the start of a new sovereign state, or just after the re- constitution of a republic after a coup d’état, is likely to acquire advantages over its rivals. The first advantage concerns the track record in contract delivery, Δ . While Ruling Party A will have a reasonably strong Δ as the party in power, Party B will have a low Δ , as they have not yet been in power. The second advantage concerns voter expectations of each party’s chances of winning, being α and (1 – α) for political parties A and B respectively. The Ruling Party would use state resources to show party’s strength and force, thus raise expectations of its subsequent electoral win. The third advantage concerns the consolidation of members’ political investment in the ruling Party, who already represent the majority, during Party A’s time in power. Political investments made by members towards the clientelist party consists of the length of uninterrupted time the member has paid allegiance to the party, as well as the quality and quantity of work the member has contributed to the ruling party during this time. Under such circumstances, it would not make sense to switch to the rival party, as the member in question will have to forgo substantial previous investment. Furthermore, if the member does switch to Party B, in-party ranking in the Page 10 of 25 rival party is likely to be low because the member has not yet invested any time or energy in Party B. A combination of these factors gives the incumbent a significant comparative advantage, enabling it to monopolize government over several election periods, over and above any normal advantage an incumbent would acquire during an electoral cycle. Breaking up monopolistic clientelism requires one of the following possibilities. The first involves a political upheaval such as a coup d’état, military intervention, or popular uprising, analogous to rebooting of the republic, through a new constitution and the abolition of the monopolistic party. A new cycle of elections can then begin again from zero, with a new set of political parties. The second possibility involves the emergence of a strong alternative leader within the party, who decides to divide the party by creating a new party. This case is akin to a competition between two clientelist parties, but would differ in the following ways: - Coefficient Δ: Contrary to other rival parties who have not yet been in power, the alternative leader emerging from the ruling party as well as his team would have their own track records, which may be as good as that of the ruling party. - Function () and pecking orders: As a strategy, alternative party may target members with a lower pecking order in the ruling party and offer higher places to them. In this case, voter expectations of new party’s chances of winning, coefficient α*, will be the determining element. Voters will not follow the new party if coefficient α* is low. It is also important to note that the new party does not need to win more than ⁄2 votes to break up the monopolistic position of the ruling party. Rather, it must (i) push the ruling party into getting less than ⁄2 votes, and (ii) explore possibilities for a coalition with the opposition. TYPOLOGY OF NON-CLIENTELIST PARTIES AND COALITIONS The reasons why a party may remain non-clientelist over time are given below, each corresponding to a specific typology. (i) Pure non-clientelist parties: Party members may value non-clientelism rather than aim for electoral success at any cost, while on certain occasions, taking part in coalition governments. (ii) Tactical non-clientelist parties: When a non-clientelist party is confronted with monopolistic clientelism, party members may quickly realize that they will not win even if they adopt a clientelist strategy. By waiting out the period of monopolistic clientelism, the party can maintain the moral high ground and be more forceful and convincing in its opposition. This would also accord a more advantageous position in the aftermath of Page 11 of 25 monopolistic clientelism, by forging coalitions following the ruling party fracture, or during elections after a civil unrest or coup d’état. The party would then follow the normal game-theoretic path to clientelism once this period is over. (iii) “Non-Clientelist” Clientelism: When a non-clientelist party comes to power, they may opt to compensate the party members for the years of hardship and losses incurred, being ( ) for each period out of power per voter. The party may compensate for this loss when it is in government, by distributing rewards exclusively to its members out of the surplus over which the party exercises control. This option is de facto clientelism, differing from the clientelism of the other parties only in its voiced justification. As such, in this type of clientelism, parties still maintain a non-clientelist speech long after they have de-facto embraced clientelism. In a coalition between a clientelist and non-clientelist party, the resources at the disposition of the government are divided between the two parties. The two parties may divide the government resources such that β is the share of government resources going to Clientelist Party A and (1- β) is share of government resources going to Non-Clientelist Party B. In terms of extractable surplus S, the non-clientelist Party B will then receive S.(1-β). If Party B still chooses to remain non-clientelist in its strategy, it would distribute this resource equally to everyone, regardless of each person’s political position. (1−) Surplus redistributed to each voter by non-clientelist party = ............................(33) Party A will receive βS of surplus and distribute to party members according to function (). ( ) .....................................................(34) Surplus distributed by Clientelist party A = . such that: ∑ () = ....................................................................................................... (35) =1 . Where is the number of voters subscribed to Party A. As a result, the benefits received by Party A and Party B members are as follows: 1− (1−) ( ) + ( Total benefits received by Party A Voter (i) = . )+ ..............(36) 1− (1−) Total benefits received by Party B members = ( )+ ..................................(37) While Clientelist Party A members may receive less than during an outright Party A victory, they would still be receiving more than they would during a non-clientelist win. Non-clientelist 1 Party B members would be getting less than ( ), the amount they would have considered just Page 12 of 25 under a non-clientelist rule. The non-clientelist party may also come under pressure to compensate members for their commitment and past losses, triggering the transition to clientelism. DICTATORIAL RULE In the model described so far, multi-party elections are carried out in a free and fair manner, despite deficiencies in the rule of law, which enables the winner to extract a surplus and appease its clientelist supporters. By contrast, in dictatorships, the system is kept in check by instilling fear in the population. However, dictatorships, like clientelist democracies, require support and loyalty of a certain part of the population and the engagement of an elite for oversight and control in administrative, military, social and economic activities. This group must also be rewarded in a similar manner to ensure the functioning of the state (Acemoglu and Robinson, 2006). It will be argued, however, that reward is only part of the equation. Punishment, or rather protection from punishment, can also be used as the other side of the coin in an inherently clientelist system. In the model, dictatorships, analogous to clientelist parties, will extract a surplus S out of available resources for clientelist purposes. In clientelist parties, benefit accruing to voter i, member of the clientelist party in power (denoted as c) were: 1− , = + () ................................................................................... (5) In case of dictatorships, the party can also offer a degree of protection from state violence and repression for its loyal supporter i, as (), where () is a function of the position of the supporter (i) in the party. As such, the total benefit accruing to supporter (i) is: 1− , = + () + ().................................................................... (38) The important issue to note is the interchangeability of the clientelist reward (i) with the state protection from state violence and repression (). Furthermore, in pure dictatorships, there will be no competitive elections. Elections would be fake elections, with comprehensive election fraud. In the intermediary case, semi-competitive elections can be held, and their results would be binding. The state can exercise election fraud, as may be necessary, through recourse to state repression and violence. In such an intermediary case, + ≥ + 1 ....................................................................................... (39) 2 Where is the number of votes obtained through election fraud. It is important to note that the party can have access to a maximum number of votes , depending on the degree of dictatorship it chooses to exercise in a sustainable manner. ▪ In the case of a pure clientelist democracy, → 0 Page 13 of 25 ▪ In the case of a pure dictatorship, → In terms of implications of dictatorial rule in the model, varying degrees of state violence and repression can be used in tandem with clientelism to achieve the following: (i) Greater inequality in surplus distribution: Clientelist parties can exercise a greater inequality in surplus distribution among its members, only when the party holds an advantage in α and Δ values, as expounded in Section 4 above. The use of a degree of state violence and repression can act in the same way, enabling a more unequal surplus distribution among members. As above, Party A with an advantage in α and Δ values can exercise unequal distribution among its members provided that no member in Party A receives benefits less than min as below: 1− min = + ................................................................................. (24) ( )+1 2 Where: = Electoral success of Party X; = Track record of Party X in honoring its contracts; Where > and > If a ruling clientelist party can exercise a degree of state violence and repression, the value min ∗ is further reduced to min as follows: ∗ 1− min = + − () ................................................................... (40) ( )+1 2 Where () is the value of protection from state violence for the marginal voter. As such, a ruling clientelist party who is able to use a degree of state violence and repression can exercise a greater inequality in its surplus distribution among members, without losing its competitivity over rival parties. (ii) Sustaining Monopolistic Clientelism: One of the ways of breaking Monopolistic Clientelism is by fracturing the ruling party itself, into two parties, as expounded in Section 4. As both the ruling party and the alternative party emerging from the ruling party would have similar values in terms of Δ, then the competition would mainly center on the differences in variable α, being the voter expectations on whether the emerging new party is likely to win or lose. However, in the case of ruling clientelist parties who use a degree of state violence, the ruling party would have an additional advantage, being its privileged access to (), which would further consolidate the power of the ruling party and its incumbent leader. (iii) Sustaining clientelism in a poorer context: It may be more difficult to run clientelism in its pure form in a low-income country with scarce resources, especially as pure clientelism requires a surplus transfer to a large group of people. Furthermore, as the resources of the country are limited, the political leadership may also not end up appropriating consequential resources for themselves. However, by using (), protection from state violence can be used as an economic substitute Page 14 of 25 for clientelist rewards. As such, political leadership would be able to run a clientelist system using (), while taking a larger surplus for themselves through the greater inequality in surplus distribution among members. In summary, ruling clientelist parties can further strengthen their power by combining clientelism with degrees of state oppression and violence, allowing the party (i) to consolidate monopolistic clientelism, (ii) to exercise more inequality and exclusivity in surplus distribution; (iii) to use this as a cheaper way to run clientelist systems in low-income, resource-poor countries. FRAGILITY AND GEOGRAPHICALLY DISTINCT COMMUNITIES This section analyzes geographically distinct communities, defined along ethnic, religious, socio-political, cultural, or linguistic lines, and their significance for the model. Contrary to individual citizens, such communities or movements have the option of challenging the sovereignty and territorial integrity of the republic, if they feel that the existing clientelist system is unfairly skewed against their interests. The model expounded so far is based on clientelist parties taking over the government apparatus and favoring their party members in the allocation of public resources and services. While other voters are discriminated against during a given electoral cycle, all voters and their parties have a chance to win in subsequent electoral cycles and to benefit from public resources in a similar, discriminate, and exclusive manner. As such, the clientelist system, in its pure form, may be considered “free and fair”, analogous to the “free and fair” workings of a casino. The clientelist system, in this logic, becomes “very unfair” in the context of monopolistic clientelism, where the rival clientelist parties no longer have an equitable chance to win. In the analogy of a casino, this may be likened to the first winner acquiring an unfair advantage over others during successive rounds. Introduction of dictatorial elements into monopolistic clientelism, including state violence, vote rigging and state repression, further skews this imbalance in favor of the incumbent and increases the feeling of injustice. In monopolistic or dictatorial clientelism, rival political parties have a restricted access to the panoply of instruments to counteract such “unfairness”. Incumbent’s effective control over the state, the parliament and the media serve as major impediments. Furthermore, the government has the means to thwart protests organized by rival parties by the actions of the state police, military, and justice system which it controls. In the clientelist model, individual voters normally have no option but to accept existing parameters of sovereignty as a given, such as the framework of the constitution, delineation of territory and national borders as well as the territorial integrity of the republic in question. Page 15 of 25 However, geographically distinct communities or their members through an independence movement, have the option of questioning and challenging, albeit at a cost, the existing parameters of sovereignty, including the territorial integrity of the republic. They would do so if their expected gains from independence would outweigh the sum of costs incurred during an independence struggle and the gains currently received under the present sovereignty arrangement. Communities as mega-voters Each geographically distinct ethnic, religious, socio-political, cultural, or linguistic community shall be denoted by x . While community members will have similar preferences because of their shared background, each person will also have the choice to act either in line with her community x or to act independently. If the person acts independently, then the person will incur a cost imposed by the community. These actions where conformity is expected may pertain to social, economic, traditional, cultural as well as political domains. When this cost is zero, the community will only be of a symbolic nature. In such a case, all members would be acting independently of the community as individual citizens and be bound only through their shared preferences. When this cost is non-zero, members will take the cost into consideration as part of their individual calculations when they determine their individual choices and course of action. Maximal levels may mean social exclusion or severe punishment, forcing the quasi-totality of members to act in line with the community decision. When ≫ 0, most members will be acting in line with political decisions of community x , enabling the community to act as a single mega- voter. As such, the community can negotiate an advantageous place within a political party or to set up a political party. In an electoral victory, the community would benefit much like an individual voter, depending on its pecking order within the party. Sovereignty and territorial integrity as policy options for a community Large geographically distinct communities, or their members have the option of challenging the existing parameters of sovereignty, including the territorial integrity of the republic, through an independence movement, albeit at a cost. If a community opts for independence, the community’s annual benefit would be an 1 average of ( ) per person from the moment they gain independence, as the community would have total control over the new state and associated public resources. This is done with the assumption of economic homogeneity among geographical regions, and that any new country arising from a partition would have a gross domestic product proportional to its population. The opportunity cost per person of remaining within the current republic would be denoted 1 by L, being the difference between a benefit of ( ) per person in the new Republic and the average annual benefits per person received in the current Republic. If the mode of governance of Page 16 of 25 1 the current republic is non-clientelist, annual benefits per person would also be ( ), and the community would have no incentive to opt for independence. If the mode of governance of the current republic is clientelist, the community may have gained or lost depending on the political context and party affiliations. The calculation would then be based on the average of past, present, and expected benefits 1 in the current Republic and whether this is greater or less than a benefit stream of ( ) per person per year in the new Republic. If the mode of governance of the current republic is monopolistic or dictatorial clientelism, and if the community is not affiliated to the ruling party in this context, then the opportunity cost of remaining in the current Republic would be high. When the opportunity cost (L) is positive, then the community must contemplate paying the cost of an independence struggle. This would be an average cost of W per person per year over w years, at the end of which the break-away republic would be attained, at a discount rate of r: 1− Total cost of an independence struggle per person = W.( ) ................................... (43) 1− Net benefit per person of opting for a break-away Republic (L*) would then be as below, considering that the community would gain independence after w years: Net benefit/person opting for a new 1+ 1− = . ( ) − ( + ). ( ) ............... (45) 1− republic at discount rate r (L*) When L* is positive, then it is more logical for the community to opt for independence. In the specific case of a community not affiliated to the ruling party under monopolistic or dictatorial clientelism, L* will be as below, considering that the community would be receiving 1 − only an average ( ) per person per year. L* for community not affiliated to the 1+ 1− = ( ) . ( ) − ( + ) . ( ) ...................... (46) 1− ruling party in monopolistic clientelism Community internal distribution systems In this model, while communities negotiate clientelist contracts as a single entity, communities are also independent to determine the modality of their internal distribution system of benefits to individual community members, regardless of the political party in question. Benefits accruing to individual community members, therefore, may not be proportional to community’s actual gain. Communities where ≫ 0 can easily exercise greater inequality in surplus distribution, analogous to dictatorial clientelism. This may be the case even if the political party the community is affiliated to may have an egalitarian distribution system. Page 17 of 25 In effect, clientelism can have a two-tier distribution system, one belonging to the clientelist political party and the other belonging to the community itself. Furthermore, in communities where ≫ 0, community decisions would be made by the community leadership, who may receive substantial clientelist rewards, while keeping the rest of the community at a bare 1 − minimum of ( ) per person per year, regardless of which party carries the elections. Independence movements Community members who have positive L* values would opt for independence, either through their community leadership or through independence movements organized outside of the community leadership. Where the internal distribution of benefits is done in a relatively egalitarian way, the community leadership can also take the lead in an independence struggle. In cases where the internal distribution of benefits is done in a highly inegalitarian manner within the community, the leadership would be getting the most of the clientelist rewards and would not be opting for independence, unlike other community members who would be getting little of clientelist rewards. Independence movements must mobilize a maximum number of community members to 1 achieve success. They can maximize the number of supporters if they promise an average of ( ) for each person per year after w years of struggle as above. This can only be promised if the independence movement can advocate an egalitarian regime, either as a non-clientelist or socialist as a governance modality. Any proposal to set up a clientelist, monopolistic or dictatorial clientelist regime would substantially lower L and L* values. It would also reduce the number of community members who would be opting for and supporting independence. As such, separatist independence movements are likely to be left leaning in their politics in their initial stages. They are however likely to follow the clientelist logic once they are in power, because of the presence of the Nash equilibrium expounded above. Risks related to fragility In line with the above arguments, risks related to fragility in cases of geographically distinct communities are summarized in Table I below. Under non-clientelist governance, there is no incentive to opt for independence. Clientelism, in its pure form, poses low to moderate risks, while substantial to high risks is associated with monopolistic and dictatorial clientelism. Internal governance modalities of communities are an important factor. Members of undemocratic communities would be more inclined to opt for a liberation struggle if their traditional community leadership is not passing clientelist rewards to community members. As such, fragility risks would be exacerbated, as the combined governance modality would be close to dictatorial clientelism. Clientelism and monopolistic/dictatorial clientelism is likely to increase the fragility of a country, encouraging communities or their individual members to challenge territorial integrity Page 18 of 25 through separatist movements. Undemocratic and unjust internal resource distribution systems within geographically distinct communities are likely to exacerbate such risks. Separatist movements arising from such divisions are likely to be left leaning and egalitarian in their politics in their initial stages. Once in power, such separatist movements will be prone to clientelism due to the presence of the Nash equilibrium expounded within the scope of this paper. Table I: Fragility Risks -Governance Modalities of Government and Communities Ruling Party Modality Community Internal Modality Fragility Risk A. Cases where the community is affiliated to a ruling party ▪ Not relevant as the state does not give Non-clientelist None out clientelist rewards Low or moderate, depending on ▪ Low inequality of S, , and (x ) values Clientelist or ▪ Moderate inequality Moderate or substantial Monopolistic Clientelism High, as most community ▪ Highly inequality, community elites members are close to the bare get the most of clientelist rewards 1 − minimum of ( ) per person Low or moderate, depending on where the community is in terms ▪ Low or moderate inequality of S, and (x ) values and provision of state protection Dictatorial Clientelism High, as most community ▪ Highly inequality where community 1 − members are close to ( ) per elites get most of clientelist rewards person per year B. Cases where the community is not affiliated to the ruling party ▪ Not relevant as the state does not give Non-clientelist None out clientelist rewards ▪ Not relevant as the community would High, as community members Clientelism not receive any clientelist rewards would be receiving the bare 1 − when =0 minimum of ( ) per person Page 19 of 25 Monopolistic High, as community members ▪ Not relevant as the community would would be receiving the bare Clientelism or not receive any clientelist rewards 1 − Dictatorial Clientelism minimum of ( ) per person BREAKING THE VICIOUS CYCLE? The obvious question remaining is: how can this vicious cycle be broken? The model shows that there will be a strong trend towards political clientelism and corruption in the context of a deficient rule of law, with a negative impact on economic, social, and environmental development. Moreover, ruling clientelist parties can further consolidate their power through monopolistic and dictatorial clientelism. While this is outside the scope of this paper, the obvious question is, how can such a vicious cycle be broken? Several suggestions can be advanced within the confines of the model. The first, and most probable option, is applying improved private sector development strategies and scaled-up finance to increase productive private sector investments, accompanied by measures to protect such investment through sovereign and other guarantee schemes. As can be seen in the model, political clientelism and corruption provides privileges to inefficient clientelist enterprises while dissuading investments by productive private sector capital. Once productive enterprises take hold, voters are more likely to steer away from clientelism with the aim of protecting their own private sector incomes from the aleatory decisions of clientelist governments. The second, and a lesser possibility is the correct use of the window of opportunity, when a republic is reconstructed anew, headed by a non-clientelist government. In such a case, before the ruling party eventually slides into clientelism, the opportunity may be used to ensure a tighter rule of law backed by regional checks and balances, proper legislation as well as substantive investments in productive capital. However, any temporary gains obtained may be reversed once the authorities turn to clientelism, due to the sovereign nature of the states. Third is effective regional peer pressure exercised for the mitigation of political clientelism and corruption. This requires effective regional mechanisms for recourse to law including for the protection of productive investments and human and social rights, as well as through regional integration measures in legislation and policy formulation. Fourth is a contentious and often deceptive short-cut, through development-oriented benevolent dictatorships. As such, benevolent dictatorships can reach their purported development objectives as well as attain a higher economic level without the interference of clientelism, albeit at a high democratic, human and social cost. Benevolence of dictatorships is also questionable, as Page 20 of 25 it is possible for the authority to function in a dual modality, partitioning clientelist and personal gains from development-oriented benevolence, while pursuing both objectives in a compartmentalized manner. In certain instances, benevolent dictatorships may also extend to include socialist systems and measures which is outside the scope of this work. A HISTORICAL PERSPECTIVE AND CONCLUDING REMARKS In concluding, it may be important to situate the model in the wider historical setting, as it shows that free and fair elections result in clientelist outcomes in the context of a weak rule of law. The rise of multi-party electoral systems coincides with the beginnings of capitalism, at a time of major social change and turmoil for those nations directly involved in colonialism and transatlantic trade. One of the important drivers of change resided in the vested interests of emerging capitalism, which required its own enabling and regulating environment. These interests could not be satisfied if the levers of political and economic power were left with the erratic will of a single royal family. Colonialism and slave related activities were key factors inducing institutional change in the Atlantic trading nations of Britain, France, Netherlands, Portugal and Spain, imposing checks and balances on the authority of the monarchies (Acemoglu et al, 2005). As economic activities evolved and wealth increased, there were demands for corresponding changes in the systems of political, economic, and financial governance. With the introduction of multiparty electoral systems, the will of the monarchs was replaced by the collective will of representatives of powerful interest groups and their evolving individual rights and rule-based systems (Rousseau, 1762). The ballot box was mainly an effective, and even an innovative, tool for replacing the authority of the monarchy. Through the ballot box, powerful interest groups could resolve the conflictual question of “who”, under a time-bound framework of a “term of office”. Capitalism needed control over state authority for many purposes. These included provision of public services, facilitation of trade, generating and enforcing appropriate administrative rules and regulations encompassing financial systems, enforcement of contracts. Modern banking emerged in the 17th century, with the Bank of England issuing banknotes in 1695 (Bagehot, 1873). Modern contract law made its appearance during the Industrial Revolution (Horwitz, 1974; Simpson, 1979). Capitalism also needed effective state authority for policing and protecting property, wealth, and equipment as well as a strong army and colonial administration able to venture and capture resources in the new worlds. The introduction of the ballot box, however, did not have the trimmings of democracy in the sense the term is understood today. Only a privileged minority of the population, approximately one in four men, could vote in the 17th and early 18th century England, with women and Catholics effectively banned from voting (Holmes, 1986; Plumb, 1967). Modern democracy started emerging with social and historical processes, pushbacks, and reactions to emerging inequalities. Page 21 of 25 Capitalism and its emerging elite did not need to plunder the state. The powerful colonial, commercial, and industrial machinery was already producing substantial wealth both at home and abroad. Furthermore, capitalism also gathered wealth by creating and deepening inequalities. On the internal front, enclosure laws in Britain took away the common rights (Thomson, 1991), opening the way for the creation of a working class. Harsh working conditions and low pay led to an unequitable distribution of benefits away from labor, especially in the initial stages of the Industrial Revolution, until there was partial redress due to trade unions. On the external front, colonialism had already opened the doors of newly discovered worlds, for exploitation of natural and mineral resources as well as for slavery and cheap labor. The state was an essential instrument and an unavoidable ally of capitalism in the generation of wealth, through its facilitating and enabling role, its regulatory functions, army and police. Despite the crucial importance of the state, the introduction of multiparty electoral systems did generate significant clientelism, like those experienced in developing countries today. Corruption was rampant even though strongly represented interests of aristocracy, landlords and emerging capitalist classes could have potentially kept each other in check as countervailing forces. In the 18th and early 19th century Britain, “the old corruption” was a well-entrenched clientelist system of buying votes through key representatives, with practices that are reminiscent of those in developing countries today. Electoral contests included “bribery, intimidation and finagling”, with local voters being wooed with “cash, trinkets, public entertainments, speeches, flattery and open houses” (Olsen, 1999). As the system evolved, further checks and balances were applied, accompanied by the gradual move towards increased suffrage and the entry of other stakeholders. In Britain, anti- clientelist measures included reforms regarding the size of constituencies, the Corruption Practices Act against bribery in 1854 and the introduction of the secret ballot in 1872 (Prenzler, 2009). Greater recognition of the trade union and labor movements (Laybourn, 1992), representation of broader interests and enhanced civil rights all played their part. Clientelism extended to the functioning of the economy. Leading firms often secured and exploited monopolistic or oligopolistic opportunities and explicitly lobbied their governments to maintain such positions. In the United States, certain companies maintained monopoly positions until the passage of the Sherman Anti-Trust Act in 1890 (Cioli, 2017; Kovacic and Shapiro, 2000), leading to the loss of their privilege only in early twentieth century. By the time the post-colonial era arrived, most people had forgotten that elections, per se, do not mean democracy, but that many upheavals were required to put in checks and balances. In the wake of their independence, most developing countries adopted multi-party electoral systems as their future system of governance. They often did so at the recommendation of their colonizing powers, who equated multi-party electoral systems with their own contemporary understanding of democracy. Former rulers aimed to ensure the emergence of liberal economies, Page 22 of 25 to work in tandem with their own economies and political systems. The notion of free and fair elections also became doubly important, especially due to the Cold War. In adopting multi-party electoral systems, most developing countries had none of the required deep-seated social checks and balances. Colonialism had often wiped entire social landscapes clean, redrawn borders and boundaries, fragmented communities, diminished previous institutions, community structures and values. In applying multi-party electoral systems on quasi- blank social pages, developing countries were to become, for many decades, centers of systematic clientelism. Furthermore, as the emerging elite tried to accumulate their own capital in the absence of full-fledged capitalism, they had little choice but to plunder their own state and public resources in a clientelist setting. Clientelism thus became an unnamed historical epoch, especially for low- income countries, defining the post-colonial era up to today. REFERENCES Acemoglu, D. and Robinson, J. “Economic Origins of Dictatorship and Democracy”, 2006. Acemoglu, D. Robinson, J. and Verdier, T. “Kleptocracy and Divide-and-Rule: A Model of Personal Rule”, 2003, Massachusetts Institute of Technology, Department of Economics Acemoglu, D.; Johnson S. and James R., “Rise of Europe: Atlantic trade, Institutional Change and Economic Growth”, American Economic Review, 2005, Vol. 95, No. 3. Acemoglu, D. Ticchi, D. and Vindigni, A. “A Theory of Military Dictatorships”, 2008, Massachusetts Institute of Technology, Department of Economics, Working Paper Series Bai, Jie et al, “Does Economic Growth Reduce Corruption? Theory and Evidence from Vietnam”, (2013), NBER Working Paper No: 19483. Bagehot, Walter (1873) “Lombard Street: a description of the money market”, Henry S. King and Co., London. Bloch, Marc (1961) Feudal Society Volume 2 Campos, Nauro F. et al, "Whither corruption? A Quantitative Survey of Literature on Corruption and Growth”, (2010) Discussion Paper no: 5334, Institute for the Study of Labor. Cioli, Alessio (2017) “Rethinking Early Antitrust History: From the Sherman Act to Theodore Roosevelt”, Dissertation for the Advanced Master of American Studies, University of Gent. Dixit, A. “Governance Institutions and Economic Activity”, American Economic Review (2009) Gander, James P., “Microeconomics of Corruption among Developing Countries” Economics, Management & Financial Markets, Sep 2011, Vol. 6 Issue 3, p11-22. Geddes, Barbara, “A game theoretic model of reform in Latin American democracies”, American Political Science Review, June 1991, Vol. 85, No 2. Geddes, Barbara, “Politician’s Dilemma: Building State Capacity in Latin America”, Berkeley, University of California Press, 1994. Page 23 of 25 Gerring, J. Thacker, S.C. Alfaro, R. “Democracy and Human Development”, (2012) Journal of Politics, Vol. 74, No. 1 Hicken, Allen, “Clientelism”, 2011, Annual Review of Political Science, Vol. 14:289-310 Horwitz, M. J. (1974), “The historical foundations of modern contract law” 87(5) Harvard Law Review. Holmes, Geoffrey (1986) Politics, Religion and Society in England: 1679-1742, Hambledon Press. Jain, Arvind K., “Models of Corruption” in Jain, Arvind K. (Ed), “Economics of Corruption”, 1998, Recent Economic Thought Series, Springer Science + Business Media. Jain, Arvind K. (Eds), “Economics of Corruption”, 1998, Recent Economic Thought Series, Springer Science + Business Media, LLC. Kaufmann, D. “Corruption, Governance and Security: Challenges for the Rich Countries and the World." Available at SSRN: http://ssrn.com/abstract=605801, (2004) Kaufmann, D. and Vicente, P. C., “Legal Corruption" Available at SSRN : http://ssrn.com/abstract=829844 (2005) Kaufmann, D. “Myths and Realities of Governance and Corruption", Available at SSRN: http://ssrn.com/abstract=829244, (2005) Keefer, P. and Khemani, S. “Political Economy of Public Expenditures”, Background Paper for WDR 2004: Making Services Work for Poor People, June 12, 2003. Keefer, P. and Khemani, S. “Why Do the Poor Receive Poor Services?”, Economic Political Weekly (2004) Keefer, P. and Khemani, S. “Democracy, Public Expenditures and the Poor: Understanding Political Incentives for Providing Public Services”, The World Bank Research Observer, (2005). Keynes, J. M., “The General Theory of Employment, Interest and Money”, 1936. Khan, Mushtaq, “Determinants of corruption in developing countries: the limits of conventional economic analysis”, in Rose-Ackerman, S. (Ed), “International Handbook on the Economics of Corruption”, 2006, Edward Elgar Publishing Limited. Li, Shaomin and Judy Jun Wu, “Why countries thrive despite corruption? The role of trust in corruption-efficiency relationship”, Review of International Political Economy, (2010). Lizzeri, Alessandro and Persico, Nicola, “The Provision of Public Goods under Alternative Electoral Incentives”, American Economic Review, Vol 91, No 1, March 2001. McGregor, R.R., “Torsten Persson and Guido Tabellini, Political Economics: Explaining Economic Policy”. Public Choice 108, 204–206 (2001). Menezes, Flavio M., “The Microeconomics of Corruption: The Classical approach” in “Corruption and Anti-Corruption”, by Peter Larmour and Nick Wolanin (Eds), 2001, Canberra Institute of Criminology, Asia Pacific Press Meon, P.-G. and Khalid Sekkat, “Does Corruption Grease or Sand the Wheels of growth?”, (2005) Public Choice. Page 24 of 25 North, D. Acemoglu, D. Fukuyama, F. Dani Rodrik “Governance, Growth, Development and Decision Making”, (2008) World Bank Olowu, Dele. "Bureaucracy and Democratic Reform" In African Perspectives on Governance, Trenton, NJ: Africa World Press. (2000). Olsen, Kirsten (1999) Daily life in 18th century, Greenwood press. Persson, Torsten and Tabellini, Guido, “Political Economics: Explaining Economic Policy”, Massachusetts Institute of Technology, 2000. Plumb, J. H. (1967) “The Growth of Political Stability in England: 1675-1725”, Macmillan. Poundstone, William “Prisoner’s Dilemma”, 1992, Doubleday Prenzler, T. “Police Corruption: Preventing Misconduct & Maintaining Integrity”, 2009, London: Taylor & Francis. Rodrik, D. “Getting Institutions Right”, 2004, Harvard University Rose-Ackerman, S. (Ed), “International Handbook on the Economics of Corruption”, 2006, Edward Elgar Publishing Limited. Rothstein, B. “Anti-Corruption – A Big Bang Theory”, QoG Working Paper Series, 2007, The Quality of Government Institute, University of Gothenburg Rothstein, B. Teorell, J. “What is Quality of Government: The Theory of Impartial Government Institutions” 2007, The Quality of Government Institute, Gothenburg University Rothstein, B. Uslaner, E. “All for All: Equality, Corruption and Social Trust”, 2005, World Politics Rothstein, B. “Preventing Markets from Self-Destruction: The Quality of Government Factor”, 2009, QoG Working Paper Series, University of Gothenburg. Rousseau, Jean Jacques (1762) The Social Contract or Principles of Political Rights. Samuelson, Paul A., “The Theory of Public Expenditure”, 1954, Review of Economics and Statistics 36, pp. 386–389 Simpson, A. W. B. (1979) “The Horwitz Thesis and the History of Contracts”, The University of Chicago Law Review. Stiglitz, Joseph, “Economics of the Public Sector”, New York, Norton, 1988. Thompson, E. P. (reprint 1991) The Making of the English Working Class, Penguin World Bank Global Report “Enhancing Government Effectiveness and Transparency: The Fight Against Corruption”, 2020. Page 25 of 25