KLEPTOCRACY AND DIVIDE- AND- RULE: A MODEL OF …KLEPTOCRACY AND DIVIDE-AND-RULE: A MODEL OF...

ALFRED MARSHALL LECTUREKLEPTOCRACY AND DIVIDE-AND-RULEA MODEL OF PERSONAL RULE

Daron AcemogluMassachusetts Institute of Technology

James A RobinsonUniversity of California Berkeley

Thierry VerdierDELTA-ENS

AbstractMany developing countries have suffered under the personal rule of kleptocrats whoimplement highly inef cient economic policies expropriate the wealth of their citizens anduse the proceeds for their own glori cation or consumption We argue that the success ofkleptocrats rests in part on their ability to use a divide-and-rule strategy made possible bythe weakness of institutions in these societies Members of society need to cooperate in orderto depose a kleptocrat yet such cooperation may be defused by imposing punitive rates oftaxation on any citizen who proposes such a move and redistributing the bene ts to thosewho need to agree to it Thus the collective action problem can be intensi ed by threats whichremain off the equilibrium path In equilibrium all are exploited and no one challenges thekleptocrat Kleptocratic policies are more likely when foreign aid and rents from naturalresources provide rulers with substantial resources to buy off opponents when oppositiongroups are shortsightedwhen the average productivity in the economy is low and when thereis greater inequality between producer groups (because more productive groups are moredif cult to buy off) (JEL O12 H00)

1 Introduction

Many countries in Africa and the Caribbean suffer under ldquokleptocraticrdquo regimeswhere the state is controlled and run for the bene t of an individual or a smallgroup who use their power to transfer a large fraction of societyrsquos resources tothemselves Examples of kleptocratic regimes include the Democratic Republicof the Congo (Zaire) under Mobutu Sese Seko the Dominican Republic underRafael Trujillo Haiti under the Duvaliers Nicaragua under the SomozasUganda under Idi Amin Liberia under Charles Taylor and the Philippines

Acknowledgments Paper presented as the Marshall Lecture at the European Economic Associ-ationrsquos Annual meetings in Stockholm August 24 2003 We thank Silje Aslaksen for pointing outan algebraic mistake in the rst version and Alexander Debs for excellent research assistanceE-mail addresses Acemoglu daronmitedu Robinson jamesarsocratesberkeleyedu Verdiertvjavaensfr

Journal of the European Economic Association AprilndashMay 2004 2(2ndash3)162ndash192copy 2004 by the European Economic Association

under Ferdinand Marcos In all these cases kleptocratic regimes appear to havebeen disastrous for economic performance and caused the impoverishment ofthe citizens

A study of the political economy of such regimes must depart from thestandard presumptions of most research in economics and political sciencewhich assume that rulers make choices within strongly institutionalized poli-ties1 In these polities formal political institutions such as the constitution thestructure of the legislature or electoral rules place constraints on the behaviorof politicians and political elites and directly in uence political outcomes Incontrast kleptocracy emerges in weakly institutionalized polities where formalinstitutions neither place signi cant restrictions on politiciansrsquo actions nor makethem accountable to citizens

While the academic study of strongly institutionalized polities is welladvanced (eg Shepsle and Weingast 1995 Cox 1997 Persson and Tabellini2000 2003) there are few studies and less of a consensus on the nature ofweakly institutionalized polities What determines corruption rent extractionand bad policies when institutions are weak Indeed the qualitative nature ofpolitics appears to differ markedly between strongly and weakly institutional-ized polities when institutions are strong citizens punish politicians by votingthem out of power when institutions are weak politicians punish citizens whofail to support them When institutions are strong politicians vie for the supportand endorsement of interest groups when institutions are weak politicianscreate and control interest groups When institutions are strong citizens demandrights when institutions are weak citizens beg for favors

The research program proposed in this paper is a systematic study ofpolicymaking in weakly institutionalized societies and ultimately a study of theprocess by which strongly institutionalized societies emerge (see North andWeingast (1989) for a classic account of such a process) To take a rst step inthis program we construct a model to study kleptocratic politics Perhaps themost puzzling feature of kleptocracies illustrated by the examples from theCongo the Dominican Republic Nicaragua or Haiti is their longevity despitethe disastrous policies pursued by the rulers2 This longevity is made even moreparadoxical by the fact that such regimes apparently lacked a political base (acore constituency) that supported them Despite the absence of formal institu-tional mechanisms for deposing unpopular rulers constraints on the behavior ofrulers exist even in weakly-institutionalized societies (eg the threat of revo-

1 We owe this terminology and the distinction between strongly and weakly institutionalizedpolities to Robert Powell2 Despite the appealing intuition perhaps derived from many political analyses of stronglyinstitutionalized societies that rulers who reduce the utility of citizens ought to be replaced thisappears not to be the case in weakly-institutionalized societies Many kleptocrats rule for longperiods Mobutu for 32 years Trujillo for 31 years and the Somozas for 42 years

163Acemoglu Robinson amp Verdier Kleptocracy and Divide-and-Rule

lution or competition from other strongmen) Why do then the heavily taxedproducers or the poverty-stricken citizens not replace the kleptocrat Why dothey rarely form an effective opposition constraining the kleptocrat How can aregime that apparently bene ts nobody outside of the narrowest of cliquessurvive Our basic answer is that this is because the absence of strong institu-tions allows rulers to adopt political strategies which are highly effective atdefusing any opposition to their regime

The seminal book by Robert Bates Markets and States in Tropical Africaprovides many clues towards an answer Bates described the web of inef cienttransfers and policies in effect in many parts of Africa but most notably inGhana and Zambia and suggested the following logic many of these inef cientpolicies are in place to transfer resources from the population to the rulinggroups while at the same time ensuring their political survival In particular thenexus of inef cient policies appeared to be useful for creating an environmentwhere any group that became politically mobilized against the rulers could bepunished while those that remained loyal were rewarded With this logic theGhanaian government heavily taxed cocoa producers while at the same timesubsidizing their inputs of seeds and fertilizers The subsidies could be allocatedselectively as a potential reward for not attempting to change the status quoSimilarly the exchange rate was kept overvalued because then the governmentcould allocate or withhold valuable rations of foreign exchange in order toguarantee support

In this paper we suggest a generalization of this reasoning which we dubthe divide-and-rule strategy Divide-and-rule is a method used by kleptocrats tomaintain power in weakly-institutionalized polities while simultaneously pur-suing policies costly to society The logic of the divide-and-rule strategy is toenable a ruler to bribe politically pivotal groups off the equilibrium pathensuring that he can remain in power against challenges To remove a ruler frompower requires the cooperation of distinct social groups which is made dif cultby the collective action problem (Olson 1965) By providing selective incentivesand punishments the divide-and-rule strategy exploits the fragility of socialcooperation when faced with the threat of being ousted the kleptocratic rulerintensi es the collective action problem and destroys the coalition against himby bribing the pivotal groups

More explicitly we consider a dynamic game between the ruler and twoproducer groups The kleptocratic ruler taxes production and uses the resultingtax revenue the rents from natural resources and potential foreign aid fromoutside donors for his own consumption The two producer groups if they cancooperate can remove the ruler from power and establish democracy (a regimemore favorable to their interests) We model this cooperation as follows one ofthe two groups (the proposer) makes a proposal to remove the ruler from powerand if the other group (the proposed) agrees the ruler is removed and democracy

164 Journal of the European Economic Association

is established The ruler-friendly political institutions however imply thatbefore the proposed group responds to the proposal the ruler can make acounteroffer This counteroffer enables him to use a divide-and-rule strategyfollowing a challenge the ruler uses all his resources and tax revenues to bribethe proposed group (and compensate them for future higher taxes if they turndown the proposal and keep the ruler in power) If he can do so successfully hecan ght off the challenge and anticipating this no group will challenge theruler Therefore the divide-and-rule strategy remains off the equilibrium pathand its anticipation implies that the ruler can follow highly distortionary (klep-tocratic) policies without being challenged Not only is the kleptocrat able tostay in power but the threat of divide-and-rule implies that there will be nochallenges to remove him from power along the equilibrium path3

In addition to providing an answer to the question of how kleptocraticregimes survive and formalizing the idea of divide-and-rule the model alsoimplies a number of interesting comparative static results These results stem notfrom varying the formal political institutions but from how the feasibility ofdivide-and-rule depends on the socioeconomic environment In particular theyfollow directly from the logic of divide-and-rule Kleptocracy survives if off theequilibrium path it can successfully bribe the pivotal group Consequentlykleptocracy is more likely (a) when producer groups or their political repre-sentatives are more shortsighted because they put less weight on future returnsunder democracy (b) when there is more foreign aid to the ruler which he canuse to bribe pivotal groups (c) when there are greater natural resource rents thatcan be used to bribe pivotal groups4 (d) when average productivity in theeconomy is low (loosely speaking because this makes natural resource rents andfunds from foreign aid more effective instruments to bribe the pivotal group)5

(e) when there is less inequality between producer groups This is because themain challenge to the ruler comes from the more productive group which hasmore to gain from democracy If inequality increases it becomes more dif cultfor the ruler to sustain his kleptocratic policies

These comparative static results are useful in thinking about the experiencesof a number of African and Caribbean countries First kleptocratic regimesseem to be more common in natural-resource-rich countries such as the Congo

3 We conjecture that these results extend to an environment with uncertainty where along thelines of the analysis in Green and Porter (1984) and Abreu Pearce and Stacchetti (1990)punishment will sometimes occur along the equilibrium path Obviously most of the examples inthe case study literature feature situations where divide-and-rule is actually observed in practice4 However this result requires producer groups not to be too patient Otherwise the prospect ofobtaining some of those natural resource rents in the future once the ruler is deposed may makeit suf ciently attractive to oust the ruler making kleptocratic policies harder to sustain5 However as with the comparative static with respect to natural resource rents the effect ofaverage productivity on policies depends on the discount rate of the citizens if citizens are veryimpatient greater average productivity may make kleptocracy more likely


Liberia Sierra Leone and Nigeria which is consistent with our result thatas long as producer groups are not too patient greater natural resource rentsfacilitate kleptocracy For example Ross (2001) shows that other thingsequal the presence of oil increases the probability that a country will be adictatorship6

Second many accounts of the worst kleptocracies also illustrate how theruler was able to use foreign aid (as well as rents from natural resources) to stayin power and ght off challenges The case of Mobutu which we discuss in thenext section is especially pertinent here This negative role of foreign aid onequilibrium policies might also explain the common nding in the empiricalgrowth literature that foreign aid seems to have little positive effect on the rateof economic growth in less-developed countries and in fact it might even havea negative effect on non-democracies (eg Burnside and Dollar 2000 EasterlyLevine and Roodman 2003) In addition there is a large amount of anecdotalevidence (eg Dollar and Pritchett 1998 Brautigam 2000 Easterly 2001 vander Walle 2001) and some tentative econometric evidence (Knack 2000) thatforeign aid leads to worse governance

The comparative static result with respect to inequality is also interestingOne interpretation of this result is that kleptocracy is less likely to arise whenthere is a balance of power between the ruler and a powerful producer group insociety This is consistent with the experiences of a number of more successfuleconomies in sub-Saharan Africa such as Botswana and Mauritius which haveboth managed to become stable democracies and refrain from the most distor-tionary policies A possible explanation is the political power of major producergroups such as the cattle owners in Botswana and the sugar planters inMauritius which has placed real constraints on the behavior of political elites(see for example Leith (2000) and Acemoglu Johnson and Robinson (2003)on Botswana and Bowman (1991) on Mauritius) Our result is also consistentwith Batesrsquo emphasis that economic policies were less distortionary in Kenyathan in Ghana because large landowners in Kenya were better able to solve thecollective action problem and act as an important force in politics The inter-pretation suggested by our model is that in Kenya the presence of a social classwith suf cient wealth and power (thus greater inequality among producers)

6 Moreover the case study evidence also suggests that whether or not resources are a cursedepends crucially on institutions (see Ross 1999 for a survey) Similarly Mehlum Moene andTorvik (2001) show that the negative effects of resources pointed out by Sachs and Warner (1995)are conditional on institutionsmdashonly countries with poor institutions speci cally weak constraintson the executive suffer a resource curse In our model this follows from the fact that once politicalinstitutions do not constrain rulers natural resources help to facilitate divide-and-rule (in thiscontext see also Robinson Torvik and Verdier (2002) for a related argument that when institu-tions are weak resource booms are a curse because they intensify incentives for regimes to useinef cient forms of redistribution to stay in power)


counterbalanced the power of the ruling party KANU (Kenya African NationalUnion) under Jomo Kenyatta7

In addition to these comparative static results we should emphasize thatkleptocratic rule can be sustained because of two features of our political game(1) the ability of the ruler to make a counteroffer after the proposal of one of thegroups (2) the ability of the ruler to charge different tax rates and makedifferential transfers to the two groups The rst feature is plausible in aweakly-institutionalized society where politicians can exploit their power toconsolidate their rule The second feature is also plausible in the context ofmany less developed countries where ethnic geographic and economic divisionsenable policies with very different distributional consequences for variousgroups to be implemented In the next section we provide a brief discussion ofhow these divisions were important in the survival of some kleptocratic regimes

This paper is related to a number of different literatures As we notedweakly-institutionalized polities have received little attention by political econ-omists For instance the political economy literature on dictatorships is veryunderdeveloped Most models are similar to ours in that they formalize dicta-torship as the maximization of an individualrsquos or grouprsquos utility function subjectto the constraint of staying in power (eg North 1981 Levi 1988 Grossman1991 Grossman and Noh 1994 McGuire and Olson 1996 Robinson 1998Wintrobe 1998 Acemoglu and Robinson 2000b La Ferrara and Bates 2001Bueno de Mesquita et al 2003 Acemoglu 2003)8 However these models donot capture the notion that the nature of politics under weak institutions isdifferent than in strongly institutionalized polities Our model develops this ideaand shows how this can lead to the divide-and-rule strategy in weakly institu-tionalized societies and how divide-and-rule leads to very different comparativestatic results

Political scientists often distinguish between different types of autocraciesor dictatorships For example the seminal study of Linz and Stepan (1996)posits the existence of four types totalitarian posttotalitarian authoritarian andsultanistic Sultanism which is close to what we mean by kleptocracy is alsoequivalent to other concepts used by political scientists Scholars working inAfrican politics for instance refer to this as neopatrimonialism personal ruleor prebendalism In the words of Bratton and van der Walle (1997 p 62)

7 In this context it is also interesting that in the most extreme examples of kleptocracy in theAmericas the Dominican Republic Haiti and Nicaragua revolution and US intervention de-stroyed traditional elites that might have blocked the power of kleptocrats8 There are interesting comparative static results from the literature on dictatorships Existingideas suggest that better economic policies are chosen by dictators with long-time horizons (Levi1988 Grossman and Noh 1994) who are encompassing in the sense that a large fraction of incomeaccrues to them (McGuire and Olson 1996) who do not fear losing their political power (Robinson1998 Acemoglu and Robinson 2000b) or who have to gain the support of a large ldquowinningcoalitionrdquo (Bueno de Mesquita et al 2003) See also Acemoglu and Robinson (2000b 2004) fora framework to analyze transitions from dictatorship to democracy


the right to rule in neopatrimonial regimes is ascribed to a person rather thanto an of ce despite the of cial existence of a written constitution Oneindividual often a president for life dominates the state apparatus andstands above its laws Relationships of loyalty and dependence pervade aformal political and administrative system and of cials occupy bureaucraticpositions less to perform public service than to acquire personal wealthand status Although state functionaries receive an of cial salary they alsoenjoy access to various forms of illicit rents prebends and petty corruptionwhich constitute an entitlement of of ce The chief executive and hisinner circle undermine the effectiveness of the nominally modern stateadministration by using it for systematic patronage and clientelist practices inorder to maintain political order

That such neopatrimonialism is at the heart of Africarsquos economic crisis is theconventional wisdom in political science (see Sandbrook 1985 Herbst 2000Bates 2001) Jackson and Rosberg (1982 pp 17ndash19) note that personal rule isldquoa system of relations linking rulers with patrons clients supporters andrivals who constitute the system If personal rulers are restrained it is by thelimits of their personal authority and power and by the authority and power ofpatrons associates clients supporters andmdashof coursemdashrivals The system islsquostructuredrsquo not by institutions but by the politicians themselvesrdquo

Our contribution to this literature is to provide a formal model of kleptoc-racy (personal rule) to systematically investigate how a divide-and-rule strategysupports kleptocracies and to develop a number of comparative static results onthe likelihood of kleptocratic regimes and policies These are all prerequisites toa full understanding of politics in weakly institutionalized societies

The rest of the paper is organized as follows In the next section we brie ydiscuss the experiences of two countries under kleptocratic regimes with specialemphasis on how these regimes were sustained Section 3 presents a model thatformalizes the main ideas discussed above and derives the main comparativestatic results Section 4 concludes

2 The Kleptocratic Experience Two Case Studies

In this section we brie y discuss two classic kleptocracies The rule of MobutuSese Seko in the Democratic Republic of the Congo from 1965 to 1997 and therule of Rafael Trujillo in the Dominican Republic from 1930 to 1961 Ourdiscussion emphasizes several characteristic features of these regimes Firstboth rulers were true kleptocrats in that their prime aim was to accumulatepersonal wealth for themselves and their families Second neither regimedespite their longevity had an extensive base of support in society Third each


ruler maintained control of power by manipulating people and groups along thelines of divide-and-rule Finally the grip on power of both rulers was greatlyfacilitated by the nature of society natural resource rents and foreign aid

21 Mobutu Sese Seko and Le Mal Zairois

Joseph Mobutu seized power in the Congo which became independent fromBelgium in 1960 in a military coup in 1965 and quickly established himself asa dictator at the head of a one-party state the MPR (Mouvement Populaire dela Revolution) There is no doubt that the aim of Mobutu was to use the state forthe enrichment of himself and his family He was a true kleptocrat In the 1970s15ndash20 of the operating budget of the state went directly to Mobutu In 1977Mobutursquos family took $71 million from the National Bank for personal use andby the early 1980s his personal fortune was estimated at $5 billion (Leslie 1987p 72) Turner and Young (1985 pp 178 ndash183) devote six pages to listingMobutursquos assets and wealth

The social base of the Mobutu regime was very narrow Mobutu himselfcame from a small ethnic group the Ngbandi in the Equateur province andthere is some evidence of systematic favoritism towards fellow Equateuriansespecially towards the end of his regime For example in 1990 46 percent ofthe of cer corp 34 of diplomats and 19 of the MPR Central Committeewere from Equateur Nevertheless it was relatively few who constituted thebase of support These people called by Zairians les gros legumes or les baronsdu regime constituted ldquoan inner core of persons with especially close kinshipethnoregional or personal ties to the president variously known as the lsquopresi-dential brotherhoodrsquo or the lsquountouchablesrsquo they were able to conduct theirmercantile affairs in blatant disregard for normal legal regulationsrdquo (Turner andYoung 1985 p 398)

Mobutu used various strategies to maintain power including the militaryand nationalist ideology and most notably the philosophy of ldquoMobutuismrdquo after1974 However his most important strategy was similar to what we refer to asdivide-and-rule creating an environment in which any person or group could berewarded or punished selectively For example Leslie (1987 p 70) describesthis as ldquoIndividuals in public of ce are totally dependent on him for selectionand maintenance of power By frequently rotating government posts Mobutumanages to maintain uncertainty and vulnerability He plays the role of bigchief bestowing favors on his subjects based on personal discretionrdquo

Turner and Young in their seminal 1985 book on the Mobutu statedescribe the same situation as follows (pp 165ndash166) ldquoClient of ce holders havebeen constantly reminded of the precariousness of tenure by the frequency ofof ce rotation which simultaneously fuels the hopes of those Zairians anxiouslywaiting just outside the portals of power The MPR Political Bureau for


example was revamped a dozen times in the rst decade of party life No one(except of course Mobutu) has been continuously a member and only sixpersons have gured on as many as half of the membership lists Insecurityhas been sustained among top state and party personnel by the frequent appli-cation of sanctions During the rst decade of the Mobutu regime 29 of the 212leaders went directly from their posts to prison on either political or corruptioncharges An additional 26 were removed on grounds of disloyalty or dishonestywith penal sanctions Frequently in presidential speeches reference is madeto unnamed disloyal or corrupt persons in the leadership ranks Cumula-tively these devices constitute a powerful mechanism of informal intimidationand suggest why systematic opposition has never arisen within the top organs ofthe staterdquo Later they continue (pp 397ndash398) ldquoAs a co-optative resource a poolof vacant high of ces was sustained the sanction for not cooperating was imprisonment on corruption nepotism or subversion charges or exileAccess to high rank in all state agencies depended upon presidential favorrdquoThey sum up the essence of Mobutursquos state as (p 193) ldquothe shifting patterns ofmembership [of the Political Bureau] constituted the very essence of patri-monial politicsrdquo

Gouldrsquos (1980 p 83) analysis concurs when he notes that ldquothe frequentcabinet shuf es and transfers of of cials from region to region may beexplained as largely re ections of the presidentrsquos skill at using people while theycan provide assistance to him and at the same time keeping factions separatedfrom each other thus preventing autonomous power centers from developingrdquo

The career of Nguza Karl-i-Bond who started as an opponent of Mobutu inthe 1960s illustrates the workings of divide-and-rule and the nature of shiftingpersonal alliances In the 1970s he was brought into the Mobutu camp and at onepoint became the foreign minister In 1977 he was accused of treason andsentenced to death In 1979 he was released and made prime minister By 1981he was in exile and gained fame by his book on the Mobutu regime (Karl-i-Bond1982) By 1985 he was back in favor and became the ambassador to the UnitedStates (Leslie 1987 pp 70ndash71) There are many other cases like this forexample Kamitatu-Massamba and Mbujimayi Belshika (see Gould 1980 p 83)Turner and Young (1985 p 166) note in this context ldquothough sanctions couldbe severe a fall from grace was not necessarily permanent Those jailed seldomremained in prison for very long Repentance and renewed cultivation of thefavor of the sovereign could make possible a return to full gracerdquo

The impact of divide-and-rule was catastrophic for the ef ciency of thebureaucracy and the state Leslie (1987 p 6) notes that in Mobutursquos regimeldquowhat is considered to be simply bureaucratic disorganization and economicmismanagement by external actors such as the [World] Bank and the IMF is toZairersquos ruling elite a rational policy of lsquoorganized disorganizationrsquo designed tomaintain the status quordquo


Added to this the personal wealth accumulation of Mobutu and his cliquedestroyed much of the economy The nationalization of foreign owned rmsmost of which ended up in the hands of Mobutu and the regime under theZairianization program initiated in November 1973 was particularly disastrousdestroying what was left of the economy According to Maddison (1995) GDPper-capita in 1992 was less than half of what it had been at independencein 1960

What factors facilitated Mobutursquos use of divide-and-rule During the rule ofMobutu the Congo was clearly very poor and characterized by low productiv-ity making it easy for him to buy off potential challengers Moreover theCongo is endowed with enormous natural resource wealth including 15 percentof the worldrsquos copper deposits vast amounts of diamonds zinc gold silver oiland many other resources (Leslie 1993 p 3) This vast natural wealth gaveMobutu a constant ow of income to help sustain his power

The Congo is also a huge and geographically diverse country (see Herbst(2000) for an analysis of the ldquodif cult geographyrdquo of the Congo) and manyscholars have seen it as having a uniquely complex cultural heritage Thecommon estimate is that there are about 200 distinct ethnic groups in the Congo(Leslie 1993 p 68) and Young (1965 p 271) argued that ldquoThe particularcolonial experience has worked to make ethnic loyalties in the Congo strongerand supra-tribal identi cation probably weaker than in most other Africanstatesrdquo According to numbers reported in Easterly and Levine (1997) theCongo is the second most ethnically diverse country in the world after TanzaniaThe fragmented nature of the country combined with the impact of the Belgiancolonial state meant that there was no large group or socioeconomic class thatcould offset the power of Mobutu and help obstruct his personal rule As wenoted in the Introduction this is in stark contrast to Botswana and Mauritiuswhere the existence of such classes placed checks on state power and limited thepotential for kleptocracy Moreover the regional and ethnic identi cation ofdifferent groups made it straightforward for Mobutu to engage in the types oftargeted redistribution and punishment emphasized in our model

Finally there is general consensus among scholars that foreign aid was akey factor enabling Mobutu to stay in power for so long Lesliersquos (1987)analysis of the relationship between the Mobutu regime and foreign aid agenciessuggests that aid and foreign nancial support was crucial in propping up theregime She notes ldquothe regime views aid agencies such as the [World] Bankand bilateral donors as sources of foreign exchange not partners in develop-mentrdquo (p 135) and also emphasizes that Mobutu and his clique ldquosees develop-ment assistance as yet another means to accumulate personal wealthrdquo Mobutumade efforts to reform the regime and its economic policies only when ldquo[theabysmal situation] undermined the con dence of Western governments and


investors and thus ows of foreign exchange resources vital for Mobutu andhis cliquerdquo (p 142)9

Thus the Mobutu regime is a classic example of how divide-and-rule can beused to sustain a regime with little if any popular support One could ask whyMobutu faced with ruling such a polity could not simply sustain himselfthrough repression rather than resort to the relatively complex strategy we haveoutlined One reason is that Mobutu always kept the military relatively weakbecause he could never be sure of its support Divide-and-rule was used withrespect to military of cers as well as civilian political elites When Mobutuneeded effective military forces he relied instead on the Belgians and theFrench who could not be expected to repress domestic opponents The expe-rience of the Congo during this awful era also illustrates the importance of thefactors highlighted by the comparative statics of our model

22 The Dominican Republic under Rafael Trujillo

The Dominican Republic the eastern half of the island of Hispaniola becameindependent from Spain in 1821 However in 1822 it was invaded by Haiti andregained its freedom only in 1844 after a war of independence The Haitianoccupation destroyed the traditional ruling elite of large landowners who edthe country never to return In the subsequent decades the country was plaguedby serious political instability and a series of dictatorships In 1916 the UnitedStates took over the country (previously they had controlled the customs since1906 to recover delinquent loans) an occupation that lasted until 1924 Duringthis period the country was run by a military governor To establish control ofthe country the United States trained a domestic constabulary which becamethe Nation Army after 1924 Trujillo became the head of this constabulary in1928 and in 1930 used it as a power base to help a coup against the governmentafter which he had himself elected President in a fraudulent election Through-out his rule Trujillo carried out the facade of elections regularly being reelectedby 100 of the vote and even on two instances from 1938 to 1942 and from1952 to 1961 stepping down from the Presidency and controlling power frombehind the scenes

As with Mobutu there is no doubt about the kleptocratic tendencies ofTrujillo Indeed relative to the size of the economy Trujillorsquos regime wasprobably the most successful case of kleptocratic accumulation in history Bythe end of his regime the fortune of the Trujillo family was equal to about 100of GDP at current prices and the family ldquocontrolled almost 80 of the countryrsquos

9 The reforms were never properly implemented however because ldquo[such] reforms strike at theheart of Mobutursquos patrimonial system and will only serve to undermine patronndashclient networksefforts to buy loyalty and wealth seeking strategiesrdquo (Leslie 1993 p 123)


industrial production About 60 of the countryrsquos labor force depended on himdirectly or indirectly with 45 employed in his rms and another 15 workingfor the staterdquo (Moya Pons 1995 p 398)

Trujillo used many tools to maintain power He used coercion and forceagainst potential rivals and murdered and tortured many opponents But as inthe Congo divide-and-rule was a key strategy As a result Hartlyn (1998)explains that ldquoTrujillo eventually became the single dominant force in thecountry by combining abuse of state power threats and co-optation Althoughcertain of the countryrsquos economic elite maintained some individual autonomythere was no possibility for independent organizationrdquo (pp 99ndash100)10

How the Dominican version of divide-and-rule worked is explained inTurits (2003 p 7) ldquoTrujillo continually shuf ed cabinet members and otherimportant of cials in and out of government posts to prevent their developing anindependent power base He exerted still tighter control over the nationallegislature Upon assumption of of ce members of the National Congress wereobliged to sign their own resignations a document that on Trujillorsquos order couldbe handed to them at any moment In certain cases these resignations werereportedly delivered to legislators in the middle of their speeches before theassembly In the 1942 to 1947 period for example there were 32 resignationsin a Senate of 19 members and 139 resignations in a House composed of 42deputiesrdquo

Wiardarsquos (1968) analysis is similar ldquoTrujillorsquos was a highly personalisticdictatorship in which power was not shared even among a small clique butconcentrated in the hands of one manrdquo (p 26) He further demonstrates thatldquoTrujillorsquos principal method of controlling the governmental machinery was theconstant shuf ing and reshuf ing of political of ceholders He had the consti-tutional authority to appoint and remove almost all government personneland he used his power extensively In addition he kept a le of signed butupdated resignations for all government employees and of cials frequentlyarrived at work only to learn that Trujillo had lled in the date and that they hadresigned The technique of shuf ing government personnel was employed forthe same reason as Trujillorsquos frequent changing of the armed forces and policecommands Anyone who gained a powerful position could expect to bereplaced no potential opposition power centers were allowed to developrdquo(p 62) Wiarda continues ldquoTrujillorsquos technique was to fragment the power of

10 Those who bene tted from the regime were few outside Trujillorsquos extended family A 1953New York Times article (quoted in Wiarda 1968 p 74) found that 153 relatives were employed bythe government Trujillo himself was minister of foreign relations minister of social securityambassador at large and special ambassador to the United States Family members held thepresidency (his brother) two senatorial posts six major diplomatic assignments the positions ofcommander in chief of the armed forces undersecretary of defense chief of staff of the air forceinspector general of the army inspector of embassies plus a great number of other posts


the many and correspondingly to concentrate all authority in his own handsrdquo(p 62)

As in Zaire people who were demoted jailed or exiled were often pardonedand returned to positions of power (see Wiarda (1968 p 63) and Crassweller(1966) for many examples) Hartlyn (1998 p 100) adds ldquoPolitically Trujillocombined guile cynicism ruthlessness and cooptation There was incred-ible manipulation of individuals who found themselves moved and removedfrom public of ce in complex and disconcerting fashion even as personalrivalries were promoted and testedrdquo

Trujillo also employed the strategy of divide-and-rule to control the ruralsector Turits (2003 pp 82ndash83 95) describes this as follows ldquoin exchange forland access and state assistance peasants were compelled to farm in a moresedentary intensive and land ef cient manner The dual nature of theTrujillo statersquos proteccion or support of the peasantry was evident in the severemethods it used to augment peasant production The state distributed rewards tothose it deemed lsquoman of workrsquo by variously providing them with land irriga-tion tools seeds credit and technical assistance while it harshly punished withvagrant penalties (jail and forced labor) and withdrawal of usufruct rights thoseit considered to be idle The distribution of lands under Trujillo was repre-sented as a gift more than a right and thus served to dramatize the dictatorrsquospersonal powerrdquo Most notably Turits (2003 p 113) explains11 ldquothe statersquosmediation of peasantsrsquo access to land fostered a high level of political controlAll individuals who wished to obtain land had to supply a certi cate of lsquogoodconductrsquo from the neighborhood authority Certainly anyone suspected ofpolitical disloyalty would be excludedrdquo

The divide-and-rule strategy was facilitated by the extraordinary controlthat Trujillo accumulated over the economy According to Wiarda (1968 pp 8790 ndash91) ldquoTrujillo could hire and re whom he pleased when he pleased Sincethe great majority of the population was absolutely dependent on him forday-to-day existence his control over it was assured there was so littleeconomic independence that even a bare minimum of political independencewas impossible Trujillo frequently employed his economic power to de-stroy his political opponents Banks could and did refuse loans and foreclosemortgages government agencies refused export or import permits electricityor phone services were interrupted and streets and sidewalks infront of selectedbusiness establishments were torn uprdquo

As with divide-and-rule in the Congo one effect was to destroy theef ciency of much of the state machinery Turits (2003 p 140) notes ldquoconfu-sion within the state was an almost inevitable by-product of Trujillorsquos system of

11 Turits (2003) interestingly claims that despite the manipulative nature of the relationshipbetween Trujillo and the peasantry the fact that Trujillo did actually give peasants access to landcreated quite a large degree of support for his regime


continually circulating high-level functionaries into different positions almostevery year for the purposes of maintaining his controlrdquo Nevertheless economicperformance under Trujillo was not as bad as under Mobutu Though thereappears to be no reliable national account data for this period the scholarlyconsensus appears to be that the Dominican Republic experienced some growthduring Trujillorsquos rule

What facilitated divide-and-rule in the Dominican Republic Unlike theCongo not all the factors suggested by our model t the Dominican experienceFor example though the country has very rich agricultural lands it does nothave large quantities of minerals oil or other natural resources Also theTrujillo regime took place in a period before the existence of the IMF WorldBank and much international development assistance so this cannot haveplayed an important role in sustaining the regime Although the United Stateswas generally friendly to the regime it did not provide great quantities of nancial assistance However two factors clearly are important When Trujillotook power in 1930 the Dominican Republic was a very poor country and thismade it relatively feasible for Trujillo to buy off support Moreover althoughthe country is not ethnically heterogeneous in the way the Congo is the impactof the nineteenth-century Haitian occupation and the United States militaryintervention meant that there was no organized traditional elite to counterbal-ance Trujillorsquos rise to power and kleptocratic tendencies

3 A Model

We now outline a simple model to formalize the ideas discussed above and toillustrate how in weakly institutionalized societies a kleptocratic regime sus-tained by a divide-and-ruled strategy can emerge Our purpose is not generalitybut to construct a simple framework which can be used to derive new com-parative static results on the incidence and sustainability of kleptocratic regimes

31 The Environment

We consider a small open economy (alternatively an economy with lineartechnology) producing three goods a natural resource Z and two goods q1 andq2 We normalize the prices of all goods to 1 which is without loss of anygenerality since we will allow differences in the technology of production of thetwo goods To start with we focus on the case where the production of thenatural resource good Zt is constant in all periods

Z t 5 Z

Natural resources create rents in this economy which in turn affect political


equilibria We assume that the natural resource rents accrue to the governmentand can then be distributed to the producers or consumed by the ruler

There are two (large) groups of agents n1 that produce q1 and n2 thatproduce q2 We normalize n1 5 n2 5 1 Both groups have utility at time t givenby

Os5t

`

bsu is~ y is lis 5 Os5t

`

bsS y is 2h

1 1 hl is~11hhD (1)

where b 1 is the discount factor yit denotes their after-tax income and lit islabor supply at time t This speci cation implies that labor is supplied withelasticity h 0

For each producer of group i the production technology is

q it 5 vilit (2)

where vi is the productivity of group i 5 1 2 Without loss of generality weassume that group 1 is more productive that is v1 $ v2 To parametrize thedegree of inequality between the two groups we denote

v1 5 v~1 1 x and v2 5 v~1 2 x (3)

where by construction v is the average productivity of the economy and x [[0 1] A greater x corresponds to greater inequality between the two groups

The only redistributive tools in the economy are a linear income tax that ispotentially speci c to each group and group-speci c lump-sum transfers Theoption to use group-speci c taxes and transfers are important for the results andplausible in the context of African societies where there are clear geographicand ethnic distinctions between producer groups12 The after-tax income of thetwo groups are

yit 5 ~1 2 titvil it 1 T it (4)

where tit [ [0 1] is the income tax imposed on group i at time t and Tit [[0 `) is a (nonnegative) lump-sum transfer to group i

In each period each producer maximizes his utility function (1) taking thetax rate tit as given This implies a labor supply function

l it~t it 5 ~1 2 titvih (5)

This equation relates labor supply and therefore output to taxes and illustrates

12 In practice differential taxes also include indirect taxes levied on the producer and consumergroups as well as the use of marketing boards to manipulate agricultural prices a frequently usedpolicy in many African countries (see Bates 1981)


the distortionary effects of taxation greater taxes reduce labor supply andoutput

Using (5) the instantaneous indirect utility of a representative agent ingroup i is found to be

U i~t it T it 51

1 1 hvi~1 2 t it

11h 1 Tit (6)

and tax revenues are

R~t1t t2t 5 t1tq1t 1 t2tq2t 5 t1t~1 2 t1thv1

11h 1 t2t~1 2 t2thv2

11h (7)

The government budget constraint is

T1t 1 T2t 1 CKt R~t1t t2t 1 Z 1 F t (8)

where CKt [ [0 `) is the consumption of the (kleptocratic) ruler R(t1t t2t) istax revenue given by (7) and Ft [ [0 `) is foreign aid if any The ruler isassumed to have the utility function at time t

Os5t

`

bKs CKs

where bK 1 is the discount factor of the ruler which could differ from thoseof the citizens

The political system is either ldquodictatorshiprdquo (controlled by the ruler) K ordemocracy D Our focus is whether dictatorship can survive and to what extentit will be kleptocratic (ie to what extent the ruler will be able to tax producersfor his own consumption while ensuring the survival of the dictatorship) Tofocus on this question we model democracy in the simplest possible way andassume that in democracy the two producer groups are in power jointly thusthey set zero taxes and share the natural resource rents and foreign aid equally(and therefore set CKt 5 0)13

In what follows we assume that only the ruler receives foreign aid (iethere is no foreign aid in democracy) More formally denoting the political stateby St21 we have Ft 5 0 if St21 5 D and Ft 5 F if St21 5 K This assumptionis not as extreme as it appears since any part of foreign aid that is perpetual canbe included in the natural resource rents Z and therefore Ft can be interpretedas the additional portion of foreign aid that the ruler receives because under hisrule there is more poverty or famine or because the ruler pursues a foreign

13 This formulation of policies in democracy is adopted to simplify the discussion and theresults are robust to more realistic models of decision making in democracy as long as democracyis more attractive for both producer groups than the kleptocratic regime For example we couldallow the two groups to be of different sizes and the larger group to impose limited redistributivetaxes on the smaller group


policy in line with the donorsrsquo interests Moreover if a democracy will receivemore foreign aid than the ruler we can allow this by letting F 0 Finally wecan interpret F as the fungible part of foreign aid In this case even if foreignaid will continue after democracy it will be allocated to some speci c purposessuch as poverty reduction or education for disadvantaged groups In this case indemocracy the two producer groups will not have access to aid and thissituation is equivalent to the one here with Ft 5 0 when St21 5 D

Given this assumption the instantaneous utilities of the two groups indemocracy are

UiD 5

vi11h

1 1 h1

Z

2 (9)

In contrast in kleptocracy the ruler will maximize his consumption subjectto the constraint that he keeps power (alternatively he can be removed frompower and in this case democracy will result and CKt 5 0 for all futureperiods) Before describing the constraints facing the ruler in detail let us writethe unconstrained solution This is given by maximizing R(t1t t2t) which isachieved at the tax rates

t1t 5 t2t

5 t 1

1 1 h (10)

and paying 0 transfers ie T it 5 0 thus setting CKt 5 R(t t) 1 Z 1 F The

instantaneous utilities of the two groups under these tax rates are given by

U i 5 U i~tit 5 t Tit 5 0 5

vi11h

1 1 h S h

1 1 hD 11h

(11)

32 The Political Game and Denition of Equilibrium

The timing of events in the political game are as follows In each period tsociety inherits a political state either St21 5 D or St21 5 K Note that St21 5D is an absorbing state so if the economy ever becomes a democracy it remainsso forever If St21 5 D the two producer groups play the simple game describedabove denoted by Gt(D) for convenience where they set the taxes and share thenatural resources rents equally If society is a dictatorship that is St21 5 K thenthe following game Gt(K) is played

1 The ruler announces tax rates (t1t t2t) and transfers (T1t T2t)2 Each group i decides whether to make a proposal to remove the ruler from

power pit 5 1 denotes that group i has made a proposal and pit 5 0 denotesotherwise (if both groups simultaneously choose to make a proposal one ofthem is chosen randomly to have pit 5 1 and the other one has pit 5 0) If


one of the two groups makes a proposal to replace the ruler we denote thisby Pt 5 1 with Pt 5 0 otherwise

3 If Pt 5 0 then (t1t t2t T1t T2t) is implemented and the political systemremains at St 5 K

4 If Pt 5 1 ie if pjt 5 1 for one of the groups then the ruler makes a newoffer of taxes and transfers (t1t

r t 2tr T 1t

r T 2tr ) such that this policy vector

satis es the government budget constraint (8) Group i THORN j then respondsto the proposal of ldquoproposerrdquo group j and the rulerrsquos new policy vector Ifthe proposed group i chooses dit 5 1 the ruler is replaced and there is aswitch to democracy ie St 5 D If dit 5 0 the political system remainsat St 5 K and (t 1t

r t 2tr T 1t

r T 2tr ) is implemented

5 Given the policy vector either (t1t t2t T1t T2t) or (t 1tr t 2t

r T 1tr T 2t

r )individuals in both groups choose labor supply

6 Output is produced tax revenues are collected and consumption takes place7 If Pt 5 1 and the proposed group plays dit 5 1 then in the next period the

stage game switches to Gt11(D) and otherwise it is Gt11(K)

There are a number of noteworthy features rst we assume that allindividuals within a producer group act in cohesion in the political game Thisis a natural assumption here since there are no costs of political action and allagents within a group have the same preferences so there is no free-riderproblem Second there is a speci c (political) structure built in the timing of thepolitical game the ruler can only be replaced if the two groups agree to replacehim This assumption captures the fact that in weakly-institutionalized societiesthose controlling the state may have considerable power and cannot be easilyremoved from of ce by one of the social groups alone14 An alternative politicalgame where the party in power needs to receive support from all social groupsor compete against potential rivals would correspond to political institutionsplacing checks on politicians However such strong political institutions areabsent in a number of less developed countries In these weakly institutionalizedpolities the implied power of the ruler combined with the fact that after theproposal to remove him from power he can offer a different policy vector giveshim the opportunity to use a divide-and-rule strategy which will be the focus ofour analysis

To simplify the analysis we will focus on the (pure strategy) MarkovPerfect Equilibria (MPE) of the above game (though this restriction is notimportant for the results) An MPE is a mapping from the current state of thegame (and from the actions taken previously in the same stage game) to

14 In addition in some countries the ethnic and social divisions enabling different taxes andtransfers on distinct groups may be absent thus also undermining the logic of the divide-and-rulestrategy


strategies Here the only state variable is St21 which denotes whether thepolitical state is either democracy or dictatorship

33 Analysis

The MPE will be characterized by backward induction When St21 5 D thereare no interesting actions and the ruler receives zero utility while the twogroups receive lifetime utilities of

V iD 5

U iD

1 2 b (12)

with UiD given by (9) Note also that Vi

D given by (12) is what the proposedgroup will receive if it chooses dit 5 1 and removes the ruler from power

On the other hand if in response to the reaction of the ruler (t itr T it

r)15 theproposed group chooses dit 5 0 its members will receive

V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)

where Ui is given by (6) and (t ie T i

e) is the MPE tax transfer combination thatapplies to this group The reasoning for this expression is that in this period theproposed group receives (t it

r T itr) and the kleptocrat remains in power so in the

future the play goes back to the equilibrium policy of (t ie T i

e)16

In addition the response of the ruler must satisfy the government budgetconstraint

T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)

The divide-and-rule strategy will be successful and the ruler will keeppower only if

V iC~t it

r T itr ut i

e T ie $ V i

D (15)

It is useful to distinguish two cases

1 The ruler will be able to maintain power with the equilibrium strategy oftit

e 5 t [ 1(1 1 h) as given by (10) and T ite 5 0 for i 5 1 2 and for all

t We denote the set of parameters such that this happens by S ie if s9 5(h b Z F v x) [ S then (t 1t

e t2te T 1t

e T 2te ) 5 (t t 0 0)

2 The ruler will not be able to maintain power if he set (t t 0 0) thus

15 The reaction of the ruler is the vector (t 1tr t 2t

r T 1tr T 2t

r ) but since only the component (t itr

T itr) is relevant for group i we will use this lower-dimensional vector to simplify notation

16 Since the deviation does not affect any payoff-relevant variable in the continuation game theMPE involves return to the same strategy and since we are focusing on pure strategy equilibriathis is simply (t i

e T ie)


s9 cedil S As we will see later in this case the ruler can reduce taxes andincrease transfers so as to maintain power We will also see that in this case(t1

e t2e) (t t) that is the ruler will necessarily be forced to reduce taxes

and policy will be less distortionary

To characterize S let us start with the subgame in which group j hasproposed to replace the ruler and denote the policies initially chosen by the rulerby (t 1t

e t2te T 1t

e T 2te ) If the ruler responds with (t it

r T itr) for i THORN j such that

V iC(t it

r T itr ut i

e T ie) V i

D then he will be replaced This shows that the ruler mustensure (15)

To analyze how and when the ruler can do so let us rst de ne

V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)

subject to (14) If V iC[ti

e Tie] V i

D for i 5 1 or 2 then group j THORN i anticipatingthat its proposal will be accepted will propose to replace the ruler and the rulerwill be deposed Therefore the ruler must guarantee that V i

C[tie Ti

e] $ V iD for

i 5 1 and 2 to remain in powerConsequently we rst need to nd V i

C[t ie Ti

e] the maximum utility that theruler can give to the proposed group off the equilibrium path To do this weneed to maximize (16) subject (14) and set the consumption of the ruler to 0 thatis CKt 5 0 Straightforward differentiation establishes that

t ir 5 0 and tj

r 51

1 1 h

Therefore in ghting off a challenge from group j the ruler will set therevenue-maximizing tax rate on this group and set zero taxes on the proposedgroup i In addition the ruler will clearly give the minimum possible amount tothe proposer group thus T j

r 5 0 Then the government budget constraint (8)implies

Tir 5 R~t i

r t jr 1 Z 1 F

Using these expressions we can derive the maximum off-the-equilibrium-path payoff of the proposed group as a function of the MPE policy vector(t1t

e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)

This expression is the maximum utility that the ruler can give to group ifollowing a proposal by group j as a function of the equilibrium tax and transferrates on group i


Given this analysis the problem of nding the MPE is equivalent to ndinga solution to the following maximization problem of the ruler

maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)

subject to the constraint set

V iCt i

e T ie $ V i

D for i 5 1 2 (19)

We now characterize the solution to this constrained maximization problemFirst notice that combining (11) (12) and (17) the constraint set (19) can berewritten as

vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)

for i 5 1 2 and j THORN i Then exploiting the fact that v1 5 v(1 1 x) and v2 5v(1 2 x) we can write the constraint set as

C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)

Moreover in the case where there is no inequality between the two groups iewhen x 5 0 the constraint set is simply

C~t i Ti C~t i T i x 5 0 $ Z S b 21

2D 2 ~1 2 b F (24)

It is already possible to see why the divide-and-rule strategy can arise inequilibrium The constraint set characterized by (21) and (22) will be satis edwhen off-the-equilibrium path the ruler can shift enough resources to the


proposed group In other words a very inef cient set of policies can besupported when each group knows that if it proposes to replace the ruler theruler will bribe the other group successfully and remain in power Recognizingthis off-the-equilibrium path threat no group will challenge the ruler who willthen be able to pursue kleptocratic policies along the equilibrium path

34 Equilibrium without Inequality

Let us start with the case in which x 5 0 and there is no inequality between thetwo groups First note that whenever he can the ruler would like to set the taxrates that maximize (18) that is (t1t

e t2te T1t

e T2te ) 5 (t t 0 0) Therefore the

rst step is to characterize the set of parameters S such that these best tax rates(from the point of view of the kleptocratic ruler) can be supported as equilibriaUsing (24) and substituting (t1t

e t2te T1t

e T2te ) 5 (t t 0 0) we immediately

obtain the set S as the set of parameters such that C(111h 0) $ Z(b 2 12) 2(1 2 b)F or more explicitly

S 5 H s 5 ~h b Z F v v11h

1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)

If s 5 (h b Z F v) [ S then the MPE involves (t1te t2t

e T1te T2t

e ) 5 (t t0 0) What happens if s cedil S Then (t1t

e t2te T1t

e T2te) 5 (te te T e T e) will

be chosen such that C(te T e) 5 Z(b 2 12) 2 (1 2 b)F (given the symmetrybetween the two groups the ruler will choose the same taxes and transfers forboth groups) Moreover inspection of (23) establishes that as long as C(t T 50) 5 Z(b 2 12) 2 (1 2 b)F for some t [ [0 t] the ruler will reduce taxesto t and sets 0 lump-sum transfers (this is intuitive since taxes are distortion-ary) The important point to note is that the ruler can always satisfy (25) andtherefore remain in power17 This highlights the importance of the underlyingpolitical institutions in this context by allowing the ruler to use divide-and-rulethe current set of political institutions make sure that he always remains inpower Nevertheless the extent to which he can transfer rents to himself anddistort the allocation of resources will depend on parameter values as we willsee next

17 To see this suppose that for some parameter values the kleptocrat is removed from powerThen it must be the case that Vi

C[tie Ti

e] V iD for i 5 1 or 2 But in this case it is credible for

the kleptocrat to reduce tie and increase Ti

e and by construction ViC[ti

e 5 0 Tie 5 Z 2] Vi

Dproviding a contradiction and establishing that the kleptocrat will not be removed from poweralong the equilibrium path This argument would not necessarily work when F 0 however

In a more general setup with stochastic productivity or costs of removing the ruler there wouldalso be parameter values for which dictatorship cannot survive similar to and for the same reasonsas the models of transition to democracy in Acemoglu and Robinson (2000a 2004)


This discussion establishes the following proposition

PROPOSITION 1 Let S be given by (25) Then we have

1 If s [ S then the unique MPE is an unconstrained kleptocratic regimewhere (t1t

e t2te T 1t

e T 2te ) 5 (t t 0 0) for all t and i 5 1 2

2 If s cedil S then the unique MPE is a constrained kleptocratic regime wherethe equilibrium policies are (t1t

e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)

for some t [ [0 t] and (t1te t2t

e T1te T2t

e ) 5 (0 0 T T) whereC(t 5 0 T) 5 Z(b 2 12) 2 (1 2 b)F otherwise

In both cases a challenge from group j that is pjt 5 1 is met by (tjtr T jt

r) 5(t 0) and (tit

r Titr) for i THORN j such that V i

C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti

e Tie) is given by (13) and Vi

D is given by (12)

The discussion above establishes this proposition The only part that mayneed more comment is the uniqueness of equilibrium Recall that if group jmakes a proposal to remove the ruler from power the ruler will respond withtjt

r 5 t and when s [ S we also have tjte 5 t It may therefore appear that

we can construct equilibria where there are challenges along the equilibriumpath when s [ S and thus the equilibrium described in part 1 of Proposition1 is not unique This is not the case however Any combination of strategieswhere pjt 5 1 cannot be an equilibrium If it were a deviation to (tjt

e T jte) 5 (t

laquo) for laquo 0 would be a best response for the ruler and the strategy of pjt 5 1would then cost group j an amount laquo 0 Since a smaller laquo is always preferredby the ruler the only combination of best response strategies is when laquo iexcl 0which is the one described in the proposition

This proposition therefore formalizes how the ruler remains in power and isable to transfer resources to himself thanks to the divide-and-rule strategy Heachieves this as follows when threatened by the proposer group he can alwaysgain the allegiance of the other proposed group by shifting resources to themBecause the proposed group is pivotal the ruler can remain in power if he cansuccessfully buy off the proposed group If this is the case anticipating thisoutcome neither group will attempt to remove the ruler from power and he willbe able to establish a kleptocratic regime transferring resources to himself at theexpense of the productive groups in society

The proposition also highlights the notion of constrained kleptocratic re-gime where the ruler is able to pursue kleptocratic policies transferring re-sources to himself but in this endeavor he is constrained by the threat that thetwo groups will coordinate and remove him from power To avoid this possi-


bility the ruler reduces the equilibrium taxes (or sometimes sets zero taxes andmakes positive transfers) to the two groups

Notice that when s [ S the equilibrium does not feature the notion ofpunishment which was discussed in the introduction According to this notionkleptocrats are in power because they can threaten to punish challengers andreward loyal groups In this case both on and off the equilibrium path the groupthat challenges the ruler is taxed at the rate t and receives 0 transfers Incontrast when s cedil S tj

e t and if group j challenges the ruler not only willgroup i THORN j be bribed to cooperate with the ruler but also group j will bepunished with the tax increased to t

Next we turn to a discussion of a number of natural comparative statics inthis model Most of those are immediate from the inspection of (25) and (26)

1 Greater F makes s [ S more likely and when s cedil S greater Fincreases taxes This comparative static is intuitive greater F that is greaterforeign aid relaxes the budget constraint of the ruler and provides him withmore resources to buy off the pivotal group off the equilibrium path Thereforegreater F makes the kleptocratic regime easier to sustain This comparativestatic result suggests that the foreign aid given to many African regimes by theUnited States and the United Nations during the Cold War period may have hadthe unforeseen consequence of consolidating kleptocratic regimes As we dis-cussed in the introduction this comparative static result may help us understandwhy in the postwar period foreign aid appears to have had no positive effect oneconomic growth on average and in fact it may have had a negative effect oneconomic outcomes in certain non-democratic countries

2 Greater b makes s [ S less likely and when s cedil S greater breduces taxes Greater b means that both groups are more patient Since thebene t of replacing the rulermdashgreater returns in democracymdashaccrues in thefuture greater patience makes it less likely that the ruler will be able to maintainhis kleptocratic regime This comparative static suggests that kleptocratic re-gimes are more likely to emerge in societies where citizens or their politicalrepresentatives value the future less

3 If b 12 then greater Z makes s [ S more likely and when s cedil Sit increases taxes If b 12 the opposite comparative statics apply Inspectionof (25) shows that greater natural resource rents create two opposing forcesFirst like foreign aid greater Z relaxes the budget constraint of the ruler andenables him to sustain his kleptocratic regime by buying off pivotal groupswhen challenged Second greater Z increases the value of democracy Whenb 12 the two groups are suf ciently short-sighted that the rst effectdominates When b 12 the second effect dominates The reason why therelevant threshold is 12 is that in democracy natural resource rents will bedivided between the two groups whereas off the equilibrium path the ruler canpay all the rents to the proposed group Casual empiricism suggests that the case


with b 12 appears more relevant here and suggests that natural-resource-richcountries may be more prone to kleptocratic regimes18 In fact the comparativestatic that greater rents from natural resources make kleptocracy more likely(ie the case with b 12) may help us explain why kleptocratic andneopatrimonial regimes have often emerged in resource-rich countries such asthe Democratic Republic of the Congo (Zaire) Sierra Leone Liberia andNigeria

4 If b (h (1 1 h))h then greater v makes s [ S less likely and whens cedil S it reduces taxes If b (h (1 1 h))h the opposite comparative staticsapply This result can be obtained by differentiating (25) The intuition for theresult may be better understood by considering condition (20) which separatelyshows the effect of the productivity of the proposed and the proposer groupsHigher productivity of the proposed group vi makes the condition less likelyto hold (and kleptocracy more likely to survive) because the proposed group hasmore to gain from democracy On the other hand a greater level of productivityof the proposer group vj makes kleptocracy more likely to survive because itimplies greater tax revenues that the ruler can use to bribe the proposed groupConsequently higher average productivity v creates two opposing forces Whenthe discount factor b is suf ciently large that is b (h(1 1 h))h the owneffect dominates and greater productivity makes kleptocracy less likely19 Thiscomparative static suggests that as long as b is not very low societies that areotherwise less productive are also more likely to suffer from kleptocraticregimes and distortionary policies of rulers

35 Equilibrium with Inequality

Let us now return to the case where the two groups do not necessarily have thesame productivity that is x [ (0 1] First recall that despite the differences inproductivity between the two groups the analysis in Section 33 established thatthe most preferred (unconstrained) policy for the ruler is still (t1t

e t2te T 1t

e T 2t

e ) 5 (t t 0 0) The question is when this policy vector will be possible forthe ruler

To answer this question recall that s9 5 (v h b Z F x) is the vector ofparameters and let S1 the the set of parameters such that C(t T 5 0 x) $ Z(b 2 12) 2 (1 2 b)F and S2 the set of parameters such that C(t T 5 02x) $ Z (b 2 12) 2 (1 2 b)F In other words s9 [ S1 implies that if Group2 makes a proposal to remove the ruler from power the ruler can make a

18 The relevant time period here is not a year so b 12 may not be unreasonable19 It can be veri ed that (h (1 1 h))h is monotonically decreasing in h with (h (1 1 h))h iexcle21 037 as h iexcl ` Therefore for the set of parameters where greater natural resources rentsmake kleptocratic policies more likely while simultaneously greater average productivity makethem less likely we need b [ (037 05)


counteroffer that coopts Group 1 even when along the MPE Group 1rsquos membersare taxed at the rate t and receive no transfers S2 is the corresponding set forGroup 2

Formally these two sets are de ned by

S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)

Since C(t T 5 0 x) is a decreasing function of x we have that S1 S2 Inother words when the producers in one of the groups become more productivethey also become more willing to oust the ruler Consequently the tighterconstraint faced by the ruler is to satisfy the more productive group off theequilibrium path This result re ects the fact that the more productive group hasmore to gain from democracy where its members will not be taxed (or moregenerally where they will be taxed more lightly) The logic of the political gameabove therefore implies that everything else equal the constraints that the rulerhas to worry about is Group 2 making an offer and Group 1 the more productivegroup accepting this proposal

This observation and a similar analysis to the one in the previous subsectionlead to the following proposition

PROPOSITION 2 Let S1 and S2 be given by (27) and (28) Then we have

1 When s9 [ S1 then the unique MPE is an unconstrained kleptocraticregime where (t1t

e t2te T 1t


2 When s9 cedil S1 but s9 [ S2 then the unique MPE is a partially constrainedkleptocratic regime where the equilibrium policy combination is (t1t

e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)

3 When s9 cedil S2 then the unique MPE is a fully constrained kleptocratic


regime where (t1te t2t

e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)

This proposition extends Proposition 1 to a situation with potential heter-ogeneity in the productivities of the two groups It also introduces the notion ofpartially and fully constrained kleptocratic regimes as before an unconstrainedkleptocratic regime pursues the policy most preferred by the ruler When it ispartially constrained the ruler has to reduce the tax rate on the more productivegroup but can tax the less productive group as heavily as he wishes When theregime is fully constrained the tax rates on both groups are constrained

The comparative static results discussed in the previous subsection continueto apply in this extended model The new result here is with respect to x thedegree of inequality between the two producer groups A greater xmdashgreaterinequality between the producer groupsmdashmakes the unconstrained kleptocraticregime less likely (ie it makes it less likely that s9 [ S1) Intuitively the morebinding constraint from the point of view of the ruler is to satisfy the moreproductive group when this group becomes even more productive democracybecomes more attractive for the producers in this group and therefore itbecomes more dif cult for the ruler to buy them off when challenged Thiscomparative static captures the notion that when there is a strong producer groupin society the ruler has less room to maneuver and therefore the unconstrainedkleptocratic regime is less likely to emerge Loosely speaking we can say thata highly productive producer group creates a balance of power and this balancebetween one of the major producer groups and the ruler prevents the mostegregious kleptocratic policies This comparative static result might help usunderstand why kleptocracies are rare in African countries with powerfulproducer groups such as the cattleowners in Botswana or the sugar planters inMauritius20

4 Concluding Remarks

Most current political economy research focuses on the study of politics inenvironments where formal political institutions place effective constraints bothon politicians and on political strategies This approach does not seem suf cient

20 Note however that a different timing of events might lead to the opposite comparative staticwith respect to inequality For example if after the two groups agree the ruler can make acounteroffer to one of the groups then he will always choose the ldquoweakest linkrdquo the group morelikely to agree which will be the less productive group In that case greater inequality may makekleptocracy more likely We believe that the timing of moves in the political game in the text ismore plausible


to understand the nature of politics in weakly institutionalized societies such asLiberia Haiti the Dominican Republic or the Democratic Republic of theCongo In these polities differences in formal political institutions cannot be themain determinant of differences in policies For example even though theCongo had several constitutions during the rule of Mobutu it appears that thishad little impact on his behavior For example Gould (1980 p 63) shows thatthe 1974 Zairian Constitution put Mobutu ldquoabove the constitution giving himthe right to run for president as often as he wishes to modify the conditionsgoverning elections to change the number of seats on the Political Bureau [thegoverning body of Mobutursquos political party and effectively of the state] todismiss as many political commissioners as he would like and indeed toabrogate the constitution as he sees trdquo Indeed Mpinga Kasenda a professor oflaw and subsequently Zairian prime minister noted in 1975 before an audienceat the Institut Royal des Relacions Internacionales in Brussels that (quoted inTurner and Young 1985 p 70) ldquoZairians have no need to refer to Montesquieuto determine what form of government they needrdquo

Such a situation is perhaps extreme Zimbabwe currently has a much moreconventional constitution written during the Lancaster House negotiations of1979 with the British government before the end of the Rhodesian stateNevertheless it is clear that the Mugabe regime acts in agrant disregard for theformal structure of the constitution

Motivated by these observations and as a rst step on the road to buildinga framework for analysis of weakly institutionalized polities we developed amodel of how kleptocratic rulers that expropriate the wealth and incomes oftheir citizens remain in power without a signi cant base of support in societyThe success of kleptocrats rests on their ability to use a particular type ofpolitical strategymdashdivide-and-rule Members of society (producer groups) needto cooperate in order to depose a kleptocrat The kleptocrat may undermine suchcooperation however by using tax revenues the rents from natural resourcesand the funds from foreign aid to bribe other groups to maintain his positionThus kleptocrats can intensify the collective action problem by threats whichremain off the equilibrium path In equilibrium all are exploited and no onechallenges the kleptocrat because of the threat of divide-and-rule We argue thatsuch a model provides a good stylistic description of the behavior of manykleptocratic regimes including that of Mobutu Sese Seko in the DemocraticRepublic of the Congo (Zaire) and Rafael Trujillo in the Dominican Republic

Our model not only formalizes divide-and-rule but yields a number ofcomparative static results useful in thinking about the emergence longevity andimplications of kleptocracy We showed that foreign aid and rents from naturalresources typically help kleptocratic rulers by providing them with greaterresources to buy off opponents Kleptocratic policies are also more likely toarise when producer groups or their political representatives are shortsighted


and when the average productivity in the economy is low Also interestinglygreater inequality between producer groups may constrain kleptocratic policiesbecause more productive groups are more dif cult to buy off We discussed howthese comparative static results are useful in interpreting a number of cases ofkleptocratic rule

We view this paper as only a rst tentative step towards a systematicanalysis of equilibrium politics in weakly institutionalized polities To under-stand the political economy of less developed societies we need a theory ofbehavior and change in weakly institutionalized societies We developed asimple model to study a speci c aspect of political behavior that seems to beendemic in such polities but did not attempt to analyze how weakly institu-tionalized societies may develop their institutions and what makes a societypossess weak institutions in the rst place These are not only important butalso exciting areas for future research

References

Abreu Dilip David G Pearce and Ennio Stacchetti (1990) ldquoToward a Theory of DiscountedRepeated Games with Imperfect Monitoringrdquo Econometrica 58 1041ndash1063

Acemoglu Daron (2003) ldquoWhy Not a Political Coase Theoremrdquo forthcoming in the Journalof Comparative Economics 31 620ndash652

Acemoglu Daron Simon Johnson and James A Robinson (2003) ldquoAn African SuccessStory Botswanardquo In In Search of Prosperity Analytic Narratives on Economic Growthedited by Dani Rodrik Princeton University Press

Acemoglu Daron and James A Robinson (2000a) ldquoWhy did the West Extend the FranchiseDemocracy Inequality and Growth in Historical Perspectiverdquo Quarterly Journal ofEconomics 115 1167ndash1199

Acemoglu Daron and James A Robinson (2000b) ldquoPolitical Losers as Barriers to EconomicDevelopmentrdquo American Economic Review 90 126ndash130

Acemoglu Daron and James A Robinson (2004) Economic Origins of Dictatorship andDemocracy forthcoming

Bates Robert H (1981) Markets and States in Tropical Africa University of CaliforniaPress

Bates Robert E (2001) Prosperity and Violence The Political Economy of DevelopmentW W Norton amp Co

Bowman Larry W (1991) Mauritius Democracy and Development in the Indian OceanBoulder CO Westview Press

Bratton Michael and Nicolas van der Walle (1997) Democratic Experiments in AfricaRegime Transitions in Comparative Perspective Cambridge University Press

Brautigam Deborah (2000) Aid Dependence and Governance Stockholm Almqvist andWicksell

Bueno de Mesquita Bruce D James D Morrow Randolph M Siverson and Alastair Smith(2003) The Logic of Political Survival MIT Press

Burnside Craig and David Dollar (2000) ldquoAid Policies and Growthrdquo American EconomicReview 90 847ndash 868

Cox Gary W (1997) Making Votes Count Cambridge University PressCrassweller Robert D (1966) Trujillo The Life and Times of a Caribbean Dictator

Macmillan


Dollar David and Lant Pritchett (1998)Assessing Aid What Works What Doesnrsquot and WhyOxford University Press

Easterly William (2001) The Elusive Quest for Growth Economistsrsquo Adventures andMisadventures in the Tropics MIT Press

Easterly William and Ross Levine (1997) ldquoAfricarsquos Growth Tragedy Policies and EthnicDivisionsrdquo Quarterly Journal of Economics 112 1203ndash1250

Easterly William Ross Levine and David Roodman (2003) ldquoNew Data New DoubtsRevisiting lsquoAid Policies and Growthrsquordquo Working Paper No 26 Center for GlobalDevelopment Available at ^httpwwwcgdevorgwpcgd_wp026pdfamp

Gould David J (1980) Bureaucratic Corruption and Underdevelopmentin the Third WorldThe Case of Zaire New York Pergamon Press

Green Edward J and Robert H Porter (1984) ldquoNoncooperative Collusion under ImperfectPrice Informationrdquo Econometrica 52 87ndash100

Grossman Herschel I (1991) ldquoA General Equilibrium Model of Insurrectionsrdquo AmericanEconomic Review 81 912ndash921

Grossman Herschel I and Suk Jae Noh (1994) ldquoProprietary Public Finance and EconomicWelfarerdquo Journal of Public Economics 53 187ndash204

Hartlyn Jonathan (1998) ldquoThe Trujillo Regime in the Dominican Republicrdquo In SultanisticRegimes edited by H E Chehabi and Juan J Linz Johns Hopkins University Press

Herbst Jeffrey I (2000) States and Power in Africa Comparative Lessons in Authority andControl Princeton University Press

Jackson Robert H and Carl G Rosberg (1982) Personal Rule in Black Africa University ofCalifornia Press

Karl-i-Bond Nguza (1982) Mobutu ou Lrsquoincarnationdu Mal Zairois London Rex CollinsKnack Stephen (2000) ldquoAid Dependence and the Quality of Governancerdquo Policy Research

working paper no 2396 World BankLa Ferrara Eliana and Robert H Bates (2001) ldquoPolitical Competition in Weak Statesrdquo

Economics and Politics 13 159ndash184Leith J Clark (2000) Why Botswana Prospered Working paper Department of Economics

University of Western OntarioLeslie Winsome J (1987) The World Bank and Structural Transformation in Developing

Countries The Case of Zaire Boulder CO Lynne Rienner PublishersLeslie Winsome J (1993) Zaire Continuity and Political Change in an Oppressive State

Boulder CO Westview PressLevi Margaret (1988) Of Rule and Revenue Berkeley University of California PressLinz Juan J and Alfred Stepan (1996) Problems of Democratic Transition and Consolida-

tion Southern Europe South America and Post-Communist Europe Johns HopkinsUniversity Press

Maddison Angus (1995) Monitoring the World Economy 1820ndash1992 Paris OECD Devel-opment Centre

McGuire Martin C and Mancur C Olson (1996) ldquoThe Economics of Autocracy andMajority Rulerdquo Journal of Economic Literature 34 72ndash96

Mehlum Halvar Karl Ove Moene and Ragnar Torvik (2001) ldquoInstitutions and the ResourceCurserdquo Working paper Department of Economics University of Oslo Available at^httpeconwpawustleduepsgepapers02100210004pdfamp

Moya Pons Frank (1995) The Dominican Republic A National History Princeton NJMarkus Weiner

North Douglass C (1981) Structure and Change in Economic History NortonNorth Douglass C and Barry R Weingast (1989) ldquoConstitutions and Commitment The

Evolution of Institutions Governing Public Choice in Seventeenth-Century EnglandrdquoJournal of Economic History 49 803ndash832


Olson Mancur C (1965) The Logic of Collective Action Public Goods and the Theory ofGroups Harvard University Press

Olson Mancur C (2000) Power and Prosperity Basic BooksPersson Torsten and Guido Tabellini (2000) Political Economics Explaining Economic

Policy MIT PressPersson Torsten and Guido Tabellini (2003) The Economic Effects of Constitutions What

Do the Data Say MIT PressRobinson James A (1998) ldquoTheories of Bad Policyrdquo Journal of Policy Reform 3 pp 1ndash46Robinson James A Ragnar Torvik and Thierry Verdier (2002) ldquoPolitical Foundations of

the Resource Curserdquo CEPR Discussion Paper No 3422Ross Michael (1999) ldquoThe Political Economy of the Resource Curserdquo World Politics 51

297ndash322Ross Michael (2001) ldquoDoes Oil Hinder Democracyrdquo World Politics 53 325ndash361Sachs Jeffery D and Andrew Warner (1995) ldquoNatural Resource Abundance and Economic

Growthrdquo Development Discussion Paper No 517 Harvard Institute for InternationalDevelopment

Sandbrook Richard (1985) The Politics of Africarsquos Economic Stagnation Cambridge Uni-versity Press

Shepsle Kenneth A and Barry R Weingast (1995) Positive Theories of CongressionalInstitutions University of Michigan Press

Turits Richard Lee (2003) Foundations of Despotism Peasants The Trujillo Regime andModernity in Dominican History Stanford University Press

Turner Thomas and Crawford Young (1985) The Rise and Decline of the Zairian StateUniversity of Wisconsin Press

van der Walle Nicolas (2001) African Economies and the Politics of Permanent Crisis1979ndash1999 Cambridge University Press

Wiarda Howard J (1968) Dictatorship and Development The Methodrsquos of Control inTrujillorsquos Dominican Republic University of Florida Press

Wintrobe Ronald (1998) The Political Economy of Dictatorship Cambridge UniversityPress

Young Crawford (1965) Politics in the Congo Princeton University Press


under Ferdinand Marcos In all these cases kleptocratic regimes appear to havebeen disastrous for economic performance and caused the impoverishment ofthe citizens

A study of the political economy of such regimes must depart from thestandard presumptions of most research in economics and political sciencewhich assume that rulers make choices within strongly institutionalized poli-ties1 In these polities formal political institutions such as the constitution thestructure of the legislature or electoral rules place constraints on the behaviorof politicians and political elites and directly in uence political outcomes Incontrast kleptocracy emerges in weakly institutionalized polities where formalinstitutions neither place signi cant restrictions on politiciansrsquo actions nor makethem accountable to citizens

While the academic study of strongly institutionalized polities is welladvanced (eg Shepsle and Weingast 1995 Cox 1997 Persson and Tabellini2000 2003) there are few studies and less of a consensus on the nature ofweakly institutionalized polities What determines corruption rent extractionand bad policies when institutions are weak Indeed the qualitative nature ofpolitics appears to differ markedly between strongly and weakly institutional-ized polities when institutions are strong citizens punish politicians by votingthem out of power when institutions are weak politicians punish citizens whofail to support them When institutions are strong politicians vie for the supportand endorsement of interest groups when institutions are weak politicianscreate and control interest groups When institutions are strong citizens demandrights when institutions are weak citizens beg for favors

The research program proposed in this paper is a systematic study ofpolicymaking in weakly institutionalized societies and ultimately a study of theprocess by which strongly institutionalized societies emerge (see North andWeingast (1989) for a classic account of such a process) To take a rst step inthis program we construct a model to study kleptocratic politics Perhaps themost puzzling feature of kleptocracies illustrated by the examples from theCongo the Dominican Republic Nicaragua or Haiti is their longevity despitethe disastrous policies pursued by the rulers2 This longevity is made even moreparadoxical by the fact that such regimes apparently lacked a political base (acore constituency) that supported them Despite the absence of formal institu-tional mechanisms for deposing unpopular rulers constraints on the behavior ofrulers exist even in weakly-institutionalized societies (eg the threat of revo-

1 We owe this terminology and the distinction between strongly and weakly institutionalizedpolities to Robert Powell2 Despite the appealing intuition perhaps derived from many political analyses of stronglyinstitutionalized societies that rulers who reduce the utility of citizens ought to be replaced thisappears not to be the case in weakly-institutionalized societies Many kleptocrats rule for longperiods Mobutu for 32 years Trujillo for 31 years and the Somozas for 42 years






































































3 A Model


31 The Environment


Z t 5 Z





Os5t

`


`

bsS y is 2h

1 1 hl is~11hhD (1)



q it 5 vilit (2)


v1 5 v~1 1 x and v2 5 v~1 2 x (3)












U i~t it T it 51

1 1 hvi~1 2 t it

11h 1 Tit (6)



11h 1 t2t~1 2 t2thv2

11h (7)




Os5t

`

bKs CKs








UiD 5

vi11h

1 1 h1

Z

2 (9)


t1t 5 t2t

5 t 1

1 1 h (10)




vi11h

1 1 h S h

1 1 hD 11h

(11)









r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan















































































































3 A Model


31 The Environment


Z t 5 Z





Os5t

`


`

bsS y is 2h

1 1 hl is~11hhD (1)



q it 5 vilit (2)


v1 5 v~1 1 x and v2 5 v~1 2 x (3)












U i~t it T it 51

1 1 hvi~1 2 t it

11h 1 Tit (6)



11h 1 t2t~1 2 t2thv2

11h (7)




Os5t

`

bKs CKs








UiD 5

vi11h

1 1 h1

Z

2 (9)


t1t 5 t2t

5 t 1

1 1 h (10)




vi11h

1 1 h S h

1 1 hD 11h

(11)









r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan













































Os5t

`


`

bsS y is 2h

1 1 hl is~11hhD (1)



q it 5 vilit (2)


v1 5 v~1 1 x and v2 5 v~1 2 x (3)












U i~t it T it 51

1 1 hvi~1 2 t it

11h 1 Tit (6)



11h 1 t2t~1 2 t2thv2

11h (7)




Os5t

`

bKs CKs








UiD 5

vi11h

1 1 h1

Z

2 (9)


t1t 5 t2t

5 t 1

1 1 h (10)




vi11h

1 1 h S h

1 1 hD 11h

(11)









r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan













































U i~t it T it 51

1 1 hvi~1 2 t it

11h 1 Tit (6)



11h 1 t2t~1 2 t2thv2

11h (7)




Os5t

`

bKs CKs








UiD 5

vi11h

1 1 h1

Z

2 (9)


t1t 5 t2t

5 t 1

1 1 h (10)




vi11h

1 1 h S h

1 1 hD 11h

(11)









r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan













































UiD 5

vi11h

1 1 h1

Z

2 (9)


t1t 5 t2t

5 t 1

1 1 h (10)




vi11h

1 1 h S h

1 1 hD 11h

(11)









r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan














































r t 2tr T 1t



r t 2tr T 1t



r T 1tr T 2t









33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan












































33 Analysis


V iD 5

U iD

1 2 b (12)





V iC~t it

r T itr ut i

e T ie 5 U i~t it

r T itr 1

bUi~t ie T i

e

1 2 b (13)





e)16


T 1tr 1 T 2t

r 2 CKt R~t1tr t2t

r 1 Z 1 F (14)


V iC~t it

r T itr ut i

e T ie $ V i

D (15)





e t2te T 1t

e T 2te ) 5 (t t 0 0)



r T 1tr T 2t




e T ie)






e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan















































e t2te T 1t



V iC(t it

r T itr ut i

e T ie) V i



V iCti

e T ie 5 max

t1tr t2t

r T1tr T2t

r

V iC~t it

r T itr ut i

e Tie (16)


e Tie] V i


C[tie Ti

e] $ V iD for


C[t ie Ti


t ir 5 0 and tj

r 51

1 1 h



Tir 5 R~t i

r t jr 1 Z 1 F


e t2te T1t

e T2te ) This is

V iCti

e Tie 5

vi11h

1 1 h1

vj11h

1 1 h S h

1 1 hDh

1 Z 1 F 1bUi~ti

e Tie

1 2 b (17)




maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan












































maxt1t

e t2te T1t

e T2te

1

1 2 bKR~t1

e t2e 1 Z 1 F (18)


V iCt i

e T ie $ V i

D for i 5 1 2 (19)


vi11h

1 1 h1

v j11h

1 1 h S h

1 1 hD h

1 Z 1 F

1b

1 2 b S 1

1 1 hv i~1 2 t i

11h 1 T iD $1

1 2 b S vi11h

1 1 h1

Z

2D

(20)


C~t1 T1 x $ Z S b 21

2D 2 ~1 2 b F and (21)

C~t2 T2 2x $ Z Sb 21

2D 2 ~1 2 bF (22)

where

C~t T x ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hD h

2bv11h~1 1 x11h

1 1 h~1 2 ~1 2 t11h 1 bT

(23)



2D 2 ~1 2 b F (24)






e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan














































e t2te T1t



e t2te T1t




1 1 h S h

1 1 hDh

2bv11h

1 1 h$ Z S b 2

1

2D 2 ~1 2 b FJ

(25)


e T1te T2t


e t2te T1t




C[tie Ti




e 5 0 Tie 5 Z 2] Vi







e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan














































e t2te T 1t



e t2te T 1t

e T 2te ) 5 (t t 0 0) if

C~t T 5 0 5 Z S b 21

2D 2 ~1 2 b F (26)


e T1te T2t



r) 5(t 0) and (tit


C (titr Tit

rutie Ti

e) 5 V iD where Vi

C (titr

Titruti


D is given by (12)




e T jte) 5 (t

















e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan























































e t2te T 1t

e T 2t







S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan













































S1 5 H s9 ~1 2 bv11h~1 2 x11h

1 1 h S h

1 1 hDh

2bv11h~1 1 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ(27)

and

S2 5 H s9 ~1 2 bv11h~1 1 x11h

1 1 h S h

1 1 hDh

2bv11h~1 2 x11h

1 1 h F 1 2 S h

1 1 hD 11hG $ Z S b 21

2D 2 ~1 2 b FJ

(28)





e t2te T 1t



e t2te

T 1te T 2t

e ) 5 (t1 t T1 0) with

C~t1 T1 x 5 Z S b 21

2D 2 ~1 2 b F (29)




e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan












































e T 1te T 2t

e ) 5 (t1 t2 T1 T2) with

C~t1 T1 x 5 C~t2 T22x 5 ZSb 21

2D 2 ~1 2 bF (30)














References















Macmillan


















































References















Macmillan











































KLEPTOCRACY AND DIVIDE- AND- RULE: A MODEL OF …KLEPTOCRACY AND DIVIDE-AND-RULE: A MODEL OF...

Documents

Transcript of KLEPTOCRACY AND DIVIDE- AND- RULE: A MODEL OF …KLEPTOCRACY AND DIVIDE-AND-RULE: A MODEL OF...