How DoesTaxation Affect the Quality of Governance?(‘‘No taxation without representation.’’) There is also a more theoretical tradition of aca-demic work that links long-term

How Does Taxation Affect theQuality of Governance?

by Mick Moore

Reprinted from Tax Notes Int’l, July 2, 2007, p. 79

Volume 47, Number 1 July 2, 2007

(C)

TaxA

nalysts2007.A

llrightsreserved.

TaxA

nalystsdoes

notclaim

copyrightin

anypublic

domain

orthird

partycontent.

How Does Taxation Affect the Quality ofGovernance?

by Mick Moore

I. Beyond Tax Policy

Most contemporary governments routinely usetheir taxation systems to get their hands on,

and then spend, between 10 percent and 40 percentof national income.1 That is a lot of money. There is

little doubt that taxing is one of the most importantthings that governments do. Correspondingly, it iswidely agreed that it is important that governmentsshould get tax policy right. When people debate andargue about tax policy, they mostly address someaspect of three big questions:

How much money should government gather astax? It should be enough to meet public spendingneeds and contribute to fiscal stability, but not somuch as to encourage the government itself to bewasteful or to appropriate money that could bebetter used in private hands.

How should the tax burden be distributed amongactual or potential taxpayers? This issue may beargued either in terms of fairness in burden-sharing, or in terms of the potential instrumentaladvantages of using tax policy to help achieve otherpublic policy goals, for example, encouraging busi-nesses to locate in poorer regions or to invest inparticular sectors, or actively redistributing incomeor wealth from one group of citizens to another.

How can the potential adverse economic costs oftaxation be contained or minimized? Taxpayers tendto be alert to the costs that they directly incur,

1In 1995-2000, average tax revenues, as a proportion ofGDP, were 13 percent for low-income countries, 19 percent for

middle-income countries, and 36 percent for high-incomecountries (Baunsgaard and Keen 2005: 7).

Mick Moore is a professorial fellow at theInstitute of Development Studies and directorof the Centre for the Future State in Brighton,U.K.

The author would like to thank DeborahBrautigam, Ben Dickinson, Max Everest-Phillips, Odd-Helge Fjeldstad, AnuradhaJoshi, James Mahon, Richard Parry, LiseRakner, Michael Ross, Carlos Santiso, AaronSchneider, John Toye, Sue Unsworth, and MikeWeeding. Carlos Santiso and Julius Court pro-vided helpful comments on an earlier draft.This article was originally published by the In-stitute of Development Studies in March 2007.

Copyright©InstituteofDevelopmentStudies2007.

SR

pecialeports

(Footnote continued in next column.)

Tax Notes International July 2, 2007 • 79

(C) T

ax Analysts 2007. A

ll rights reserved. Tax A

nalysts does not claim copyright in any public dom

ain or third party content.

whether these take the less damaging form of com-plex and costly paperwork and record-keeping obli-gations, or the more malign form of harassment:arbitrariness on the part of tax collectors and theneed to pay bribes.2 If collectively well organized,taxpayers or their legislators may inquire into theefficiency of the tax administration itself. How muchof the money that it raises is absorbed in thecollection process? Economists also routinely focusour attention on the indirect costs of revenue raising— that taxing any activity almost inevitably dis-courages it. If, for example, a government choosesthe easy option of raising most of its income bytaxing coffee exports, it may be biasing the wholeeconomy, in an inefficient way, against producingcoffee for export. It is probably more efficient simplyto spread the tax burden broadly.

These tax policy questions are very important.They will continue to dominate debate about taxa-tion in most countries and to absorb the attentions oftax policy specialists. However, they are not the onlyimportant tax policy issues. In particular, there is agrowing debate — focused on poorer countries andon governments enjoying large incomes from aid orfrom oil, gas, and mineral exports — that addressesa different and in some ways more foundational setof questions: not ‘‘What is good tax policy?’’ but ‘‘Howdoes the taxation relationship between state andcitizens itself contribute to the quality of gover-nance?’’ The purpose of this article is to review thisdebate, to explain why it has arisen, and, above all,to answer the question: How does taxation affect thequality of governance?

II. The Taxation-GovernanceConnection

There is a long history to the idea that there issome strong, consistent connection between theways in which governments are financed and theways in which they govern. In particular, this isfoundational to mainstream Anglo-American inter-pretations of the emergence of representative gov-ernment and democracy in United Kingdom and theUnited States. The British narrative focuses on how,in a series of struggles with the 17th-century Stuartdynasty in particular, Parliament secured its long-term dominance over the monarchy by wrestingcontrol of public finances and restricting the sourcesof funding for the government bureaucracy, thearmed forces, and the monarchy to money raised

through tax and authorized by Parliament.3 TheAmerican story focuses on the rejection of Britishcolonial authority in the 1770s after London violateda core principle of good government and tried to taxthe American colonists without asking for theirconsent through a legislature in which they wererepresented. (‘‘No taxation without representation.’’)There is also a more theoretical tradition of aca-demic work that links long-term changes in societyand governance to changes in the ways in whichstates obtain the resources they need to govern. Itsleading exponents also see close affinities betweenthe dependence of governments on general taxation,modern capitalism, and liberal democracy.4

The notion that sources ofgovernment income mightsignificantly shape governancehas only recently entered intodevelopment policy debates.

The notion that sources of government incomemight significantly shape governance has only re-cently entered into development policy debates. It isstill on the margins. The word ‘‘taxation’’ appears inthe development literature mainly in the context ofthe kinds of economic policy discussions summa-rized in Section I. And those issues are generallybelieved to be specialized and technical and not torelate to core debates about the development pro-cess. Even those aid and development agenciesnoted for their capacity to generate and use policyresearch continue to produce major statements ongovernance that ignore revenue and taxation is-sues.5 One reason is simply the time it takes for acomplex issue to be first raised by researchers andscholars, thoroughly debated, and then entered intopolicy agendas in an authoritative way. The earliestmention I have found in a development policy docu-ment of the significance of taxation for governance isin Deborah Brautigam’s discussion paper for theWorld Bank in 1991 (Brautigam 1991). That in turnreflected a growing academic literature. Two bookshave been especially influential in putting the issueson the agenda for academic historians and political

2For recent survey information on these costs for a typicalcompany in a wide range of countries, see Pricewaterhouse-Coopers and World Bank (2006) and World Bank (2006).

3Parliament authorized a quarter of the revenues raisedby the British crown between 1560 and 1640, and 97 percentbetween 1689 and 1714 (Braddick 1996: 13).

4See especially Goldscheid (1958/1925) and Schumpeter(1918/1991).

5See, e.g., the World Development Report widely believedto have signaled the reconversion of the World Bank to theimportance of effective states (World Bank 1997), and the newaid and development policy paper issued by the U.K. govern-ment (Department for International Development 2006).

Special Reports

80 • July 2, 2007 Tax Notes International

(C) T

ax Analysts 2007. A




scientists, respectively. One is Charles Tilly’s (1992)comparative study of state formation in WesternEurope. Tilly argues that instead of simply taxingcoercively, rulers were motivated to bargain overtaxation, state financing, and public policy with thepeople who controlled large amounts of capital;states tended to become more capable (especially inwar) and more accountable and responsive to citi-zens. His final chapter on the implications for con-temporary poor countries prefigures many of thearguments in this article. The second book is KirenChaudhry’s (1997) analysis of the effects of contrast-ing sources of public revenue — oil rents in the onecase, and remittances from citizens working abroadin the other — on contemporary state formation, andespecially on the development of state capacities inSaudi Arabia and Yemen. Here was a convincingexplanation of how, over periods of years rather thanof decades or centuries, changes in sources of publicrevenue could shape the apparatus of government,the ways in which state authorities interacted withsocietal groups, and state capacity.

It was on the basis of these kinds of evidence thatsome authors began to make strong claims about thecentrality of the tax relationship to governance, withmore or less emphasis on the notion that higherlevels of taxation would bring forth more capability,accountability, or responsiveness on the part of gov-ernments.6 Like Tilly’s bold generalizations abouthistorical patterns of European state formation,these assertions have generated critique and testingby other scholars.7 As always, the specialists validlypointed out that generalizations are only generali-zations and that the detailed picture is more com-plex.8 More fundamentally, some researchers have

6See, e.g., Guyer (1992), Moore (1998), and a range ofliterature on the rentier state in the Middle East.

7The most rigorous attempts at statistical testing are byRoss (2004) and Mahon (2006).

8For some nuancing of bold claims about European histori-cal patterns, see Ertman (1997) and Rosenthal (1998).

Table 1. The Changing Importance of Sources of State Revenue

Source Frequency in Later 19thCentury

Frequency in Later 20thCentury

Extent to Which theRevenue Source TypicallyIs ‘‘Earned’’ by the State in

Terms of OrganizationalEffort

(a) (b) (c) (d)

A. Prerogative revenues:the sale of permissions,titles, and public offices

Rare Rare Low

B. Seigniorage Significant in some cases Significant in some cases, butbecoming rare by the end of

the period

Low

C. Surpluses from monopolycontrol of trade inconsumer goods

Occasional Rare Medium

D. Surpluses from stateproperty

Significant in some cases Important until recently incentrally planned economies

(state-owned industry);currently infrequent

Medium to high

E. Surpluses from monopolycontrol of agriculturalcommodity exports

Occasional Was common in many poorcountries in midcentury, and

then largely disappeared

Medium

F. Surpluses from export ofnatural resource products

Rare Very important for a limitednumber of states

Low

G. Grants from other statesor internationalorganizations

Rare Important for a large numberof poor countries

Low

H. Broad taxation ofcitizens

Important, especially inwealthier countries

Widespread and dominant inwealthier countries

High

Special Reports


(C) T

ax Analysts 2007. A




responded to the emphasis that I and others haveplaced on the potential ‘‘governance dividend’’ fromtaxation by pointing to the prevalence of more or lessdirectly coercive taxation practices in poor coun-tries.9 Recently, there has been a blending of inter-ests between researchers working on the governanceconsequences of taxation and those investigating thepolitical impacts of the ‘‘resource curse.’’ The issuesare in many respects the obverse of one another: Insome interpretations at least, the central politicalpathology of states rich in oil, gas, and mineralrevenues is that they do not need to tax (SectionV).10

In sum, there is now substantial research litera-ture about the connections between taxation andgovernance. But the issue is still only on the mar-gins of the development policy debate. This is notsolely because it takes time for scholarly insight topercolate into policy. Some other causes lie in the‘‘politics of knowledge.’’11 The reason that concernsme here is the ambiguity of the various overlappingpropositions that have been advanced about thegovernance implications of taxation. The first objec-tive of this article is to try to present the argumentsclearly. The second is to evaluate their accuracy andrelevance to policy. Before doing that, I provide abroad historical sketch of changes in sources ofgovernment income.

III. States and Revenues

What are states? Most concretely, they are sets oforganizations, which typically are complex and dif-

ferentiated.12 Those organizations need to bemanned, equipped, provisioned, and coordinated.Governments need to command the services of alarge number of people and acquire substantialmaterial resources. Historically, states have de-pended on two main kinds of material resources,extracted from the societies over which they rule:conscript manpower, mainly for military purposes,and financial resources to pay the salaries of civilianand military personnel and to meet the other capitaland recurrent costs of warfare and government (Levi1997).13 Few present-day militaries depend signifi-cantly on conscript soldiers. Trained, experiencedpersonnel are needed to operate modern weaponssystems. Contemporary military capacity depends toa high degree on the financial resources to purchasecostly equipment and to recruit and train the pro-fessionals to operate it. Money is the dominantmaterial resource required by contemporary states.

From where do states obtain their money? Plun-dering other polities has long ceased to be a routine,significant source. Eight other sources have beenimportant at different points in history:14

• Prerogative revenues, which is the sale of a widerange of permissions (to inherit or sell property,install glass windows, encroach on state for-ests, use the state’s justice system, legalizedocuments), titles (peerages, knighthoods), orpositions in the civil or military bureaucracies.

• Seigniorage, which is the revenue that govern-ments obtain from their monopoly over theissue of currencies.

• The surpluses that states obtain from monopolycontrol over wholesale or retail trade in con-sumer goods like salt, alcohol, or tobacco.

• Surpluses from state property, notably royalestates (in agrarian societies) and, in more

9See especially Fjeldstad (2001); see also Bernstein and Lu(2003) and Juul (2006).

10The literature on the resource curse and rentier states isso large and growing so fast that it is not possible even to listall the main items here. Some of the classic original state-ments of the argument were by Mahdavy (1970) and Luciani(1994) on the Middle East and by Karl (1997) on Venezuela.Ross’s (1999) excellent review of the literature is in somerespects dated. For a more recent review, see Rosser (2006).One of the more interesting recent conceptual contributions isby Collier (2006). Note that this literature is vibrant becausethe resource-curse phenomenon became widespread only re-cently, after the big oil price increases of the 1970s trans-ferred resources and rents on an unprecedented scale toproducers of this subset of primary commodities.

11All the wealthy, democratic states of the OECD havebeen almost entirely funded from broad taxation for manydecades, including most of the period following World War II.There has not been enough variation among them in sourcesof government revenue to attract the attention of socialscientists to the question whether revenue sources signifi-cantly affect governance.

12I like Theda Skocpol’s definition: ‘‘Any set of relativelydifferentiated organizations that claims sovereignty and co-ercive control over a territory and its population, defendingand perhaps extending that claim in competition with otherstates’’ (Skocpol 1992: 43). It is not very different from manyother definitions within the Weberian tradition.

13For accuracy, one might add the extraction of in-kindresources, notably agricultural products. The omission doesnot affect the argument here.

14These categories of revenue sometimes overlap at themargin. I am classifying them with an eye more to theunderlying political economy than to strict logic. Equally, onecan debate the extent to which some of them should beclassified as taxes or fees/charges.

Special Reports


(C) T

ax Analysts 2007. A




industrial societies, public utilities (railways,ports, electricity) and state-owned industrialenterprises.

• The surpluses that states obtain from monopolycontrol over the export of valuable agriculturalcommodities like cocoa, coffee, tea, cotton, andsugar.

• The surpluses that states obtain, whetherthrough taxation, direct ownership, royalties, orother charges, from the location on their terri-tory of enterprises extracting valuable oil, gas,or mineral export products.

• Aid and subsidies, which are direct transfersfrom other states or international organiza-tions.

• The broad taxation of citizens and enterprises,notably taxes on incomes, assets, property, pay-rolls, imports, exports, the internal movementof goods, persons (poll taxes), production andsale of specific commodities (excises), wealthtransfers, business turnover, and sales or valueadded business taxes.

There are wide variations, over time and space, inthe extent to which states depend on these differentrevenue sources. We can, however, make a fewuseful generalizations about the trends summarizedin columns (b) and (c) of Table 1:

• Prerogative revenues have rarely been a majorsource of state income, especially not in recentcenturies. Such devices — especially the sale ofpositions in public bureaucracies, which under-mines state authority and revenue in the longerterm — often have been a last resort in times ofcrisis. Stamp duties on contractual legal docu-ments represent one of the few remaining sig-nificant revenue sources within this generalcategory.

• Governments obtain some revenue from issuingcurrency (seigniorage). Currency represents apotential claim on the issuer. Not all claims arerealized. Seigniorage is a significant incomesource only for brief periods of time whengovernments overprint currency and cause in-flation. Over the last two decades, most govern-ments have committed themselves to respon-sible macroeconomic and monetary policies andtherefore cannot access this income source.

• Historically, governments have often obtainedsome income from monopoly control of the tradein basic consumer goods, including, classically,salt. For example, monopoly revenues fromalcohol and tobacco provided between 18 per-cent and 27 percent of the Taiwanese nationalgovernment’s revenues in the 1950s and 1960s(Republic of China 1982: 160). These incomesources shrink as economies become more di-verse, complex, and open. They are rare today.

• Royal estates were major sources of state rev-enue in the medieval era. Large state-ownedrailway and postal enterprises funded somecontinental European states in the 19th cen-tury (Ferguson 2001: 57-58; Mann 1993: 381-389). Because they are unusually small andprosperous city trading states, contemporarySingapore and Hong Kong have been able tofinance themselves largely from state propertyownership, urban development, and managingport, transport, and other public utilities. Ex-cepting these very unusual cases, own revenuestend to decline in importance as incomes in-crease and economies become more complexand interrelated: It becomes relatively easier toobtain revenue through taxation rather thanthrough direct management of productive as-sets. The main exception in the 20th centurylay in the centrally planned (communist) econo-mies, in which governments funded themselvesthrough direct control of most productive enter-prise. Some elements of these systems remainin place in China, Vietnam, and North Korea,although the Chinese and Vietnamese govern-ments are increasingly dependent on broadtaxation.

• Between the 1950s and 1980s, the governmentsof many poorer countries used their controlover the marketing and export of a wide rangeof agricultural commodities and exchange ratepolicies to extract substantial revenues andother resources from the rural sector (Bates1977; Lipton 1977). This revenue source largelydried up in the 1980s and 1990s for a combina-tion of reasons: the political economy of thosecountries that had experienced economicgrowth changed in favor of rural interests(Bates 1993; Moore 1993), the extraction under-mined itself by severely wounding the goosethat lay the golden eggs, and the widespreadstructural adjustment and economic liberaliza-tion that followed a series of international eco-nomic shocks in the 1980s deprived govern-ments of many of their extractive tools.

• Surpluses from extraction or export of naturalresource products have long been an occasionalsource of significant government revenue butbecame a major source of funding for manystates only after the first oil price shock in theearly 1970s. They have remained so since. Thatin turn reflects the combination of three maintrends: the rapid growth in the dependence ofmodern industrial economies on oil and nowgas for energy; declining international trans-port costs, fueled partly by containerization andother innovations that make it increasinglyviable to trade bulk commodities over longdistances; and continuing or growing interna-tional inequality in productive capacity and

Special Reports


(C) T

ax Analysts 2007. A




incomes. Where it exists, the natural resourcesector automatically becomes relatively impor-tant to economies that lack the capacity to adda great deal of value in manufacturing or inter-nationally competitive service industries.15

• Grants (or concessional loans) from other poli-ties have been an exceptional mode of financingstates for most of history. They have beenassociated with direct influence of the donorover the recipient state. The picture was trans-formed after the end of World War II. The bigexpansion of foreign aid was initially motivatedin part with geostrategic concerns around theCold War. Strategic motivations have dimin-ished since the end of the Cold War,16 thenumber of aid donors and aid channels hasproliferated (Acharya et al. 2006; World Bank2004: 206-207), and aid has increasingly beenconcentrated on the poorer countries. It nowconstitutes a major income source for manygovernments (Moore 1998).17

• Dependence on broad citizen taxation hassteadily become a more important revenuesource for most states recently. It is now thedominant income source for governments, ex-cept only for those that enjoy large revenueseither from natural resource exports or fromaid.

In sum, broad taxation, surpluses from naturalresource exports, and development aid are the domi-nant income sources for contemporary states. Mostgovernments depend heavily on broad taxation,some combine this with surpluses from naturalresource exports or development aid, and a few livemainly on surpluses from natural resource ex-ports.18 In column (d) of Table 1, these eight mainrevenue sources are classified in terms of the or-ganizational effort that they require of states. Howmuch energy do they need to devote to recruiting,

managing, or supervising the agencies that actuallycollect the revenue and to reshape society and theeconomy to enhance taxability?19 These simple cat-egorizations indicate a polarizing trend in recentdecades: Contemporary states depend on some mix-ture of revenue sources that typically either requirea high organizational effort (broad taxation) or a lowone (surpluses from natural resource exports anddevelopment aid). Those revenue sources requiringan intermediate level of organizational effort havebecome less significant since the later 20th century.

IV. Taxation and Governance?

In contemporary development debates, the term‘‘governance’’ is used variously to refer to outcomes(the effective provision of collective goods) and to thepolitical processes that generate these outcomes: themanner in which state elites acquire and use theirpower and authority.20 There is no professionalconsensus on the operational definition, roots, ormeasurement of good governance. There are twoimportant conceptual and theoretical underpinningsto the argument in this article.

Dependence on general taxationprovides incentives for state elitesand taxpayers to resolve theirdifferences through bargaining.

The first is a powerful and simple, but oftenimplicit, model of the political process — the notionof state-society tension and balance — that consti-tutes ‘‘an important theoretical foundation for muchpolitical science’’ (Gervasoni 2006: 5). The coreproposition is that political regimes are the outcomeof tension and conflict between (a) elites who controlthe state and wish to remain in power and toexercise that power as freely as possible, and (b)societal actors who want to place restraints on thepower of a potentially overweening state, either toprotect themselves from despotism and depredationor as a strategy for obtaining power themselves.Revenue is central to this conflict for two reasons.First, it represents a key strategic resource for stateelites. If nonstate actors can limit and control elites’access to revenue, they enjoy countervailing powerin relation to the state. Second, if state elites need to

15Conversely, even large natural resource extraction sec-tors comprise only a small proportion of national economies,such as Norway, that otherwise are very competitive inmanufacturing and services.

16With some obvious exceptions, such as Egypt, Israel, andNorth Korea.

17Very high levels of dependence generally reflect large aidinflows following natural disasters or internal conflicts. In2004, official development assistance exceeded 70 percent oftotal government expenditure in Burundi, Sierra Leone,Afghanistan, Malawi, Mozambique, the Democratic Republicof Congo, Ethiopia, and Nicaragua (Ben Dickinson, OECD,DAC, private communication).

18Governments obtain income from resource exportsthrough many channels, and official accounts are often in-complete or misleading (Mahon 2006: chapter 2). We do nottherefore have a reliable, comprehensive data series.

19This is a modification of my earlier concept of earnedincome, which referred to both the political and the organi-zational or bureaucratic efforts that states put into raisingrevenue (Moore 1998).

20I am not dealing with nonstate governance in thisarticle.

Special Reports


(C) T

ax Analysts 2007. A




depend on general taxation because they lack alter-native and easier revenue sources, they generallyhave to put considerable organizational and politicaleffort into obtaining the revenue and face strongincentives to bargain and negotiate, directly or indi-rectly, with at least some taxpayers, rather thansimply extracting revenue forcibly. In other words,dependence on general taxation provides incentivesfor state elites and taxpayers to resolve their differ-ences through bargaining.

The second conceptual-theoretical underpinningto this article is that taking into account the realitiesof patterns of governance in contemporary develop-ing countries, we can usefully think of good govern-ance as having three main operational dimensions,which tend to complement and reinforce one an-other:21

• The responsiveness of states to citizens, that is,an orientation to meeting citizens’ needs.22

• The accountability of states to citizens, wherethis implies the existence of institutionalizedmechanisms — of which electoral democracy isthe most important and thoroughgoing —through which state elites both answer to citi-zens for the ways in which they employ theirauthority and may be rewarded or sanctionedby extensions or curtailments of that authority(Schedler 1999).

• The capability of states to determine and re-spond to citizens’ needs and wants, which inturn has two complementary dimensions:— the political capability to determine needs

and to frame and nurture bargaining andcompromise among competing interests;and

— the organizational or bureaucratic capacityto settle on sensible policies, to deliverpublic services, and to enforce the authorityof the state.

For a range of contextual and historical reasons,many governments in developing countries rank21This also happens to be the analytic framework adopted

in the U.K. government’s new aid and development policypaper (Department for International Development 2006).

22The concept of responsiveness is silent on the question ofhow much power citizens might enjoy. Responsive govern-ments are concerned with citizens’ needs but have autonomyto shape those needs and to decide to which they prefer to

respond. The notion that citizens have power to express theirwants, as opposed to their needs, is central to the concept ofaccountability.

Table 2. The Effects on Governance of State Dependence on Broad Taxation

Immediate Effects Intermediate Effects Direct Governance Outcomes

A. The state becomes focused on obtainingrevenue by taxing citizens.

(i) The state is motivated to promotecitizen prosperity

More responsiveness

(ii) The state is motivated to developbureaucratic apparatuses and informationsources to collect taxes effectively

More bureaucratic capability

B. The experience of being taxed engagescitizens politically.

(Some) taxpayers mobilize to resist taxdemands or monitor the mode of taxationand the way the state uses tax revenue

More accountability

C. As a result of A and B, states andcitizens begin to bargain over revenuesand exchange willing compliance bytaxpayers for some institutionalizedinfluence over the level and form oftaxation and the uses of revenue (that is,public policy).*

(i) Taxes are more acceptable andpredictable, and the taxation process ismore efficient

More responsiveness and political andbureaucratic capability

(ii) Better public policy results fromdebate and negotiation

More responsiveness and politicalcapability

(iii) Wider and more professional scrutinyof how public money is spent

More accountability

(iv) The legislature is strengthenedrelative to the executive (assuming oneexists)

More accountability

* Bargaining is especially likely if representative institutions (legislatures) already exist.

Special Reports



(C) T

ax Analysts 2007. A




relatively low on all these dimensions of good govern-ance (Moore 2004). Increases in responsiveness,accountability, or capability are welcome, and tendto go together: More of one generally means more ofthe other.23

There is no one-line — or even single-paragraph— explanation of how state revenue sources andtaxation affect governance. Why? One set of reasonsis essentially methodological: It is hard to get theright data and sort out complex interactive politicalprocesses. In particular:

• Even the simplest models of the political pro-cesses around taxation involve two main cat-egories of actors: state (or state elites) andsocietal groups of various kinds. Some re-searchers focus more on how states respond todifferent sources or levels of revenues, andothers emphasize the response of societal actorsto attempts to tax them.24 Any realistic under-standing of the processes requires that we (a)go beyond the initial reactions to taxation ofeach type of actor and take into account theways in which they then interact, whetherconflictually, cooperatively, or in more complexways; and (b) explore the likely diversity ofinterests and behavior within our ‘‘state’’ and‘‘society’’ categories.

• Taxation and revenue are but one set of factorsthat influence the responsiveness, accountabil-ity, and capability of states. In explaining ac-tual cases, we also need to take account of theother factors, including those that tend tostrengthen or weaken democracy.

• We have neither the data on revenue sourcesnor the clear operational measurements ofbroad concepts like state responsiveness orstate capability that we would need to rigor-ously explore alternative answers to our centralquestion. It is particularly unfortunate thateven basic data on sources and uses of publicrevenue tend to be especially sparse or unreli-able for governments that depend on both for-eign aid (Moore 1998: 110-119) and naturalresource exports (Mahon 2006: chapter 2).

The second set of reasons for the relative complex-ity of the taxation-governance connection is that ithas several dimensions. We need to examine at leastfour different issues:

• From where do governments get their rev-enues? (Section V.)

• How much of their incomes do citizens pay intaxes? (Section VI.)

• Who and what is taxed? (Section VII.)

• How are taxes assessed and collected? (SectionVIII.)

V. Sources of Revenues

Let us begin with the contrast between govern-ments that depend (largely or entirely) for revenueon earned income in the form of broad taxation andthose that depend on unearned income,25 from ei-ther aid or natural resource exports. What are theconsequences for governance of being dependent ontaxes? We can build up a model of the politicalimplications step by step. It is summarized in Table2.

Let us first look at two of the more direct effectson the state itself (Section A, Table 2):

First, rulers dependent on taxes have a directstake in the prosperity of (some or most of) theircitizens because that improves rulers’ future incomeprospects. They therefore have incentives to pro-mote that prosperity.

Second, dependence on taxes promotes bureau-cratic capacity. Broad taxation, to a far greaterextent than either aid or natural resource revenues,obliges the state to invest in the creation of arelatively reliable, uncorrupt, professional careerpublic service to assess and collect dues and thenhand them over to the state treasury.26 Equally, itnecessitates the creation of reliable records of tax-payers, their activities, and assets, and it providesthe state with a wide range of information that itmight use for a variety of purposes, from improvingthe quality of policymaking (Brewer 1989: especiallychapters 4 and 8) to undercutting support for ruralguerrillas (Odon 1992).

But what about the effect on citizens of beingtaxed (Section B, Table 2)? We cannot say much onthis until we know how citizens interact with thestate over taxation. But we can assume that theexperience of being taxed will engage them politi-cally. They tend to try to organize to resist taxationor ensure their tax money is well used. Unless the

23See, e.g., Kaufmann et al. (2006).24See, e.g., Ross’s work (2004) and most of the literature he

cites.

25For the basis of this distinction, see Moore (1998).26Brewer’s (1989: 101-114) account of the organization of

the Excise Department in 18th-century Britain illustratesthis line of argument vividly. Perhaps the best contemporaryscholarly research on this issue — and on the broaderquestion of how revenue sources shape states — has beendone on Saudi Arabia and Yemen by Kiren Chaudhry (1989,1997).

Special Reports


(C) T

ax Analysts 2007. A




sole response of the state is crushing repression,these reactions tend to increase the accountability ofgovernment.

It is in Section C of Table 2 — the interactionbetween state and society over taxation issues —that we enter into some of the most interestingterritory. The primary point is simple: There areseveral reasons why rulers and taxpayers mightenter into productive bargaining around taxation, totheir mutual benefit, in ways that improve theoverall quality of governance. The potential directbenefits to this revenue bargaining are:

• If the state and taxpayers can reach agree-ments on what taxes will be levied, at whatlevels, and how, tax demands become morepredictable, taxpayers are more likely to com-ply with them, and both the economic and thepolitical costs of taxing are reduced. With betterknowledge of likely future obligations and rev-enues and protection against arbitrary levies,taxpayers could feel more secure in makingeconomic investments, and rulers could under-take long-term fiscal planning more effectively.

• Having a forum in which revenues are ex-changed for policies encourages the search forpolicies that are beneficial to both governmentsand citizen-taxpayers.

• The degree of effective public scrutiny of howpublic money is spent is likely to increase.27

• If a representative legislature already exists, itwill be strengthened by becoming a major fo-rum for the bargaining about the exchange ofrevenues for policies. If it does not exist, rulerswill be motivated to establish one to find anorganization representing taxpayers withwhich they can regularly and reliably nego-tiate.

The likely outcome of these direct effects of rev-enue bargaining is a considerable improvement inthe quality of governance in all dimensions. But howlikely is that outcome? Will the logic in the modeltranslate into reality? Below, I provide the evidencethat it does. But, at least in the contemporary world,it does not translate into reality directly, fully, andunambiguously. There is no one-to-one relationshipbetween tax dependence, revenue bargaining, andgood government. We can revert to deductive logic tounderstand the reasons. The model I have presentedabove is too simple. It is based on the history of theemergence of parliamentary government in WesternEurope, especially Britain and the Netherlands in

the 17th and 18th centuries. In three importantrespects, the institutional environment in whichcontemporary governments tax and spend is morediverse than that faced, for example, by the Stuartkings of 17th-century Britain.

In the 17th century, taxation was dominantly forpurposes of funding the preparation or fighting ofwars. Most public expenditure was on the military(Brewer 1989: 40; Mann 1993: 373). The politicsaround public expenditure were relatively uncompli-cated. They focused to a large degree on foreignpolicy issues about alliances and warfare, becausethat is what mainly shaped the size of states’ rev-enue demands. Military contractors were an interestgroup, but not a very large one, and they tended tobe very much in the parliamentary eye. In otherwords, the activities and concerns of taxpayer-basedinterest groups dominated fiscal politics. By con-trast, most contemporary states routinely spend agreat deal of money on a diverse range of activitiesless immediately pressing than war, including edu-cation, healthcare, roads, electricity and other infra-structure, pensions, social assistance, research, sta-tistics, and subsidies of various kinds to industryand agriculture. These expenditures might advan-tage many people, either in their roles as directbeneficiaries or in their roles as public-sector em-ployees dispersing these expenditures. Significant,complex interest group politics take place aroundpublic expenditure. We cannot easily predict howfar, over any public policy issue, Joan Smith willidentify with, and therefore vote or campaign in, herrole as an actual or potential taxpayer, direct ben-eficiary of public expenditure, or state employee.Fiscal politics are more complex than in the 17thcentury, and other concerns might trump taxpayers’‘‘natural’’ focus on economy and efficiency in publicspending.

Taxpayers as a category tend to be more diversein contemporary poor countries than in 17th-centuryBritain and for that reason alone are less likely toengage in collective action on behalf of all taxpayers.Most contemporary states preside over complextaxation regimes, characterized by a diversity ofdifferent taxes, a wide variety of rates, an abun-dance of exemptions, and ambiguity in legislation orassessment procedures. The tax regime itself mayinfluence the potential for taxpayers to engage incollective action in a variety of ways.28 For example,distinctively different taxes, levied on very differentbases, might lead to political conflict between one

27I like Paul Collier’s concept of sovereign rents: ‘‘that partof public revenue over which there is no effective scrutiny’’(Collier 2006: 1484).

28The underlying analytic framework is polity-centricity— the observation that the nature of public policy arenas,public programs, and the public organizations that imple-ment those programs help shape the ways in which citizensengage politically (Skocpol 1992: 47-60).

Special Reports


(C) T

ax Analysts 2007. A




group of taxpayers and another. Alternatively, acomplex taxation system with a variety of discre-tionary exemptions might motivate taxpayers topass over collective action and to engage instead inindividualized strategies to solve their problems (forexample, through bribery or employing expert taxadvisers). In many poorer developing countries, afew large transnational corporations provide a largeproportion of the more accessible potential taxbases. They have strong incentives to negotiateindividually with government and to influence bothtax legislation and its practical application by thetax administration. The political and organizationaldimensions of taxation may vary widely depending,for example, on whether the target is a few largecompanies or myriad informal petty traders.

The dependence of governmentson unearned income is likely tohave adverse effects on theiraccountability, responsiveness,and capability.

In historical Europe, governments had little scopeto resort to loans as a means of avoiding confrontingand bargaining with their own taxpayers, becausethe social contract with taxpayers was the basis oflow-cost borrowing. Financiers would lend at lowerrates of interest to governments that had goodprospects of repaying; and governments subjected toparliamentary control of finance were more likely torepay because their big taxpayers were committed tothe revenue-raising system, parliamentary scrutinywould reduce the chances that revenues would bewasted, and sometimes the people who made theloans also sat in Parliament.29 By contrast, contem-porary governments tend to borrow heavily interna-tionally, both in commercial markets and, in the caseof the poorest countries, on concessional terms frominternational organizations like the IMF and theWorld Bank. Borrowing on international marketsgives governments some autonomy from their tax-payers. By using privileged information about com-plex financing alternatives and by shaping percep-tions of what economic and financial policies areinevitable or desirable, governments have somescope to present their legislatures with interna-tional borrowing agreements that are done deals —

too complex for legislators to challenge.30 However,this partial autonomy that governments enjoy inrelation to their own taxpayers might be bought atthe expense of dependence on the IMF. The govern-ments of many poor countries need the approval ofthe IMF to access international loan markets onreasonable terms. In some cases, the IMF has as-sumed what many would see as the historic role oftaxpayers-in-assembly:

One might take the IMF to be acting as a kindof upper house or appropriations committee‘‘representing’’ Latin American taxpayers withits combination of sufficient expertise to designtax legislation and sufficient power to denycritical resources to the executive. [Mahon2004: 26]31

In sum, the material in this section points towardtwo conclusions. First, a mixture of historical anddeductive arguments suggests that the dependenceof governments on unearned income is likely to haveadverse effects on their accountability, responsive-ness, and capability. Second, there are other factorsthat influence this relationship in contemporarypoor countries: Their fiscal politics are sufficientlydiverse and complex that the causal sequences areunlikely to emerge with the same immediacy andclarity as they are sketched out in Table 2. What isthe evidence for the validity in contemporary poorstates of the classic causal sequence from depen-dence on broad taxation to better governance?

For reasons shown in Section IV, it is oftendifficult to test the key propositions in a rigorousfashion. However, the more immediate source of ourpartial ignorance is that it is only in very recentyears that much effort has been put into the testing.We are less ignorant now than we were a few yearsago and have a range of contemporary evidence.This includes:

• Country case studies of the malign governanceeffects of dependence of states on unearnedincomes from resource exports (especially Karl1997; Chaudhry 1989, 1997; Vandewalle 1998).

• Strong quantitative evidence, based on cross-national statistical analysis, that large oil and

29Some historians have seen this capacity to raise loans onthe strength of a tax base guaranteed by the central role ofparliaments as the key to the emergence, in historical Europe,of powerful, successful fiscal states (Brewer 1989).

30I am grateful to James Mahon for giving me access to hisdraft (Mahon 2006) that details this point.

31Mahon points out that as the representative of creditors,the IMF has an interest in a strong taxation system socreditors can be repaid. However, it may at times favor theinterests of bondholders if they come into conflict with thoseof taxpayers (Mahon 2004: 26) to the extent that the IMFsometimes appears responsible for squeezing poor countrytaxpayers by insisting that governments meet demandingrevenue targets.

Special Reports


(C) T

ax Analysts 2007. A




mineral revenues are associated with low levelsof democracy and with states unbound by law.32

• Somewhat more mixed and contested evidencefrom cross-country statistical analysis aboutthe impact of high levels of aid dependence.33

• Easter’s (2002) persuasive comparison of theevolution of taxation and governance in Polandand Russia after the end of the Soviet system.Russia continues to depend for revenues on arelatively small number of very large enter-prises, many of them now private or state-owned energy firms. There is little public politi-cal debate around taxation, and certainlynothing like a social contract: Most taxes areraised coercively, and ‘‘raids’’ by tax collectorsand tax police are commonplace. By contrast,the Polish government could not depend on afew big taxpaying enterprises. It had to taxemployees and small firms, and it graduallyextended the tax net to a large proportion of thepopulation. Taxation has become a standardtopic for political bargaining and debate.

Country comparisons are always hard to inter-pret; so many other things vary between any twocases. The most convincing evidence on the connec-tions between tax dependence and governance isnow emerging from comparisons between subna-tional governments within the same countries. Ger-vasoni (2006) has looked at the political history of 21of Argentina’s provinces from 1983 to 2003. Theydepend substantially — but with significant inter-province variations — on fiscal transfers from cen-tral government and, in a few cases, local revenuesfrom the oil extraction industry. Gervasoni foundthat the provinces most dependent on broad taxa-tion of their citizens had historically been moredemocratic. When provincial governments weremost generously supplied with financial transfersfrom central government or oil revenues, local po-litical leaders had been better able to buy off orsuppress competition from democratic oppositions.Hoffman and Gibson (2006) have compared 105district governments in Tanzania, a country inwhich local revenue raising is often coercive (Fjeld-stad 2001). When district populations have thegreatest potential economic mobility, and thereforethe widest scope to flee from coercive local taxation,district governments spend higher proportions oftheir revenues on providing services for their citi-zens and less on themselves. In different ways, bothstudies provide evidence for the balance of powerhypothesis: When local governments face the strong-est pressures to finance themselves by coaxingrather than extracting revenues from their citizens,

32There is continuing debate about both the possiblecausal mechanisms between enjoyment of ‘‘point’’ naturalresource revenues (oil, gas, minerals) and the quality ofgovernance, and the data and statistical methods for testingthose relationships. There is no fully adequate survey of thefast-growing research literature, although Ross (1999) andRosser (2006) cover much of the ground very well. I willrestrict myself here to comments on two main issues. The firstis that there are different underlying hypotheses about howresource wealth affects governance. For example, some theo-ries concentrate on the issues discussed in this article: theimplications of not having to tax. Others deal with a comple-mentary but distinct issue: the extent to which governmentswith large incomes from natural resources can buy off oppo-sition and thus keep themselves in power (Collier and Hoef-fler 2005). I suspect the literature has underplayed theimplications of greed and threat: Governments sitting onlarge mineral wealth want to stay in power and naturallyassume that plenty of other people are keen to displace them,probably by force. That motivates high expenditures onmilitary, police, intelligence services, the general militariza-tion of politics, and exclusionary governance. These differentcausal sequences are consistent with one another and mightall be operating simultaneously. It is difficult to test that inany rigorous manner. Part of the reason lies in the secondissue on which I comment — the problem of measuring thesize of the natural resource curse. Most studies use the onlydata series that has been available until recently — somemeasure of the value of mineral resource production orexports as a proportion of total national income or exports.But those are not the right things to measure. The theoriestell us we should be measuring rents, that is, the proportionof the sales value of a product that remains after all produc-tion and marketing costs, including some allowance for nor-mal profit, have been accounted for. Rents may be very highon some natural resource extraction activities. It may costonly a couple of dollars — a small fraction of its sales price —to obtain a barrel of oil from beneath a desert. Conversely, ahigh proportion of the sales price of a barrel of Norwegian oilrepresents real production costs out in the North Sea. Theproper measure of the likely political impact of naturalresource wealth would begin with the rents, not sales value.Collier and Hoeffler (2005) have used World Bank data tocreate just such a data base. For that reason we trust theirconclusions, from cross-national statistical analysis, thatnatural resource wealth tends consistently to depress thelevel of democracy. However, even studies using the moreeasily available measures of resource wealth point in thesame direction: Ross (2001) and Mahon (2006).

33For example, Remmer (2004) finds that aid tends todiscourage governments from taxing, and Brautigam andKnack (2004) find that high dependence on aid tends todegrade the quality of governance institutions over time. Bycontrast, Moss et al. (2006) review a range of literature on thetopic and find unproven the case for the malign effects of aidon governance. One likely reason for the ambiguity of thestatistical results is explored by Collier (2006). Aid hasmultiple and variable dimensions that are often disguised

when we measure it in aggregate value terms. In somecontexts, it is validly understood as an inflow of ‘‘free’’financial resources into government coffers. In others, itcomes in the form of specific goods or services (for example,technical assistance) or hedged with all kinds of conditions,restrictions, or injunctions about the ways in which it will beused.

Special Reports



(C) T

ax Analysts 2007. A




they are more likely to rule democratically (Argen-tina) or to spend money providing services to citi-zens (Tanzania).

VI. How Much Do Citizens Pay inTaxes?

The conclusions in Section V are based on ananalysis of the effects of alternative sources of gov-ernment income on the motivations and behavior ofthe state, taxpayers, and the two interactively. Thepotential causal interactions are many and rela-tively complex, especially because states face choicesbetween taxing broadly and thus engaging withcitizens, and exploiting unearned revenues, whichrequire them to engage with a narrow range of otheractors — aid donors or a small number of largecompanies, public or private, in the oil, gas, andminerals sector. The question of how heavily govern-ments tax citizens is distinct and conceptually lesscomplex. We need to think mainly of how the level ofrevenue demand might impact the political actionsof taxpayers. To the best of my knowledge, no onehas tried to track in detail this process for individualcountries. And we know that the political implica-tions are not straightforward. Increases in the taxtake are unlikely to occur in isolation. They willgenerally result also in increases in public expendi-ture, which might in turn intensify the processesthrough which the politics around public expendi-ture ‘‘confuses’’ the politics around revenue rais-ing.34 However, there are good, logical reasons, assummarized in Table 2, to believe that the more oftheir income they pay in tax, the more citizens arelikely to be politically engaged in and demanding ofthe government. We also have anecdotal evidencethat marked increases in the level of tax burdensmobilize taxpayers (Winters 1996: chapter 4), aswell as recent statistical evidence indicating thatincreases in revenue demands precede by only a fewyears shifts toward more democratic and liberalgovernance.35

In sum, there is compelling evidence that thedependence of states on unearned income is likely tohave adverse effects on the quality of governanceand that the overall level of taxes does help mobilizecitizens politically.36 But before we draw too-firm

policy conclusions, we need to look at our third andfourth questions: Who and what is taxed? And howare taxes assessed and collected?

VII. Who and What Is Taxed?I have so far talked in general terms, using broad

categories like ‘‘taxes,’’ ‘‘taxpayers,’’ and ‘‘citizens,’’which need to be disaggregated to understand spe-cific cases. Many contextual factors affect the extentto which citizens are engaged and politically mobi-lized by particular taxes. I deal here only with twoissues of particular importance to contemporarydeveloping countries.

First, governments are more accountable andresponsive to their citizens when dependent on themfor revenue. Does that also imply that governmentsare only accountable and responsive to taxpayers —that is, those citizens who pay taxes? Is the policyimplication that we should find ways of ensuring thepoor are taxed to prevent governments being ac-countable and responsive only to the nonpoor, at theexpense of the poor?37 The answer to the last ques-tion is no. There are two reasons. The first iscautionary: We simply do not know enough aboutthe effects of taxation on political behavior to justifythis kind of experimental social engineering. Thesecond is that there is good reason to believe that theentire polity and all social groups normally willbenefit from greater state responsiveness and ac-countability to taxpayers, even if the taxpayersthemselves are not poor. Why? Especially in coun-tries with relatively fragile public institutions, poli-tics is rarely the kind of rational public finance gamethat economists love to model, in which one particu-lar definable group — for example, farmers — willbe calculating in detail the benefits of, say, exchang-ing reduced coffee export taxes for a stronger legalcommitment to employees’ rights.38 The more that

34In addition, one should separate a tax increase thatleaves a taxpayer out of pocket from an increase that iscompensated by an improvement in the quantity or quality ofpublic services (Ross 2004).

35See the original research by Ross (2001) and somereformulations with modified data by Mahon (2006).

36One has to be careful about the kinds of quantitativeevidence used to support this argument. The work by Rossand Mahon cited in the previous footnote is valid because it

deals with changes over a few years. But one could generatespurious statistical results by looking at cross-national dataover longer time periods. We know that income levels arethemselves major determinants of the level of democracy(Diamond 1992; Heo and Tan 2001; Huntington 1991; Lon-dregan and Poole 1996; Przeworski et al. 1996) and theoverall level of the tax take (Section IX). The richer thecountries are, the more they are democratic and the higherthe proportion of gross national product their governmentsraise in taxes. Variations in the tax take are thereforeassociated statistically with variations in the level of democ-racy. But it would be wrong to interpret that as a causalrelationship when using long-term data series comparingcountries at very different income levels.

37In many poor countries, especially in Latin America, thisis an irrelevant concern because the poor already pay moretax relative to their incomes than do the rich — albeit mainlyin indirect taxes.

38Moore (2003) and Landa and Kapstein (2001).

Special Reports



(C) T

ax Analysts 2007. A




politics is like that, the more we should be concernedabout the problem of accountability only to tax-payers, not the poor. However, in most of the coun-tries with which we are concerned — the poorest,and those dependent on aid and mineral resourcewealth — the pressure for governments to be ac-countable or responsive to taxpayers is likely to havemore positive effects at a more basic level of thepolity by encouraging the creation of the kinds ofstable institutions and predictable political behav-iors that are often in deficit. Poor people will nor-mally benefit and rarely lose out when ruled bygovernments that, because they are dependent ongeneral taxes, face incentives to coax rather thansimply extract revenues from citizens, and thereforeconfront restraints on their power, are motivated toprotect human and property rights, and understandthat they and citizens share a common interest ineconomic growth.

Governments are moreaccountable and responsive totheir citizens when dependent onthem for revenue.

Second, if we perceive an intrinsic political valueto taxation through its tendency to engage taxpayersand therefore mobilize them politically to confrontand bargain with the state, should we then notrecommend an emphasis on direct taxes, and regretthe very rapid recent spread in developing countriesof indirect VAT?39 Another no. The reason is that,from a political perspective, VAT is a direct tax: Itvery perceptibly affects enterprises. Its introductionhas been resisted strongly in some developing coun-tries, mainly because of the visible bookkeepingburden it imposes on small businesses.40 Once VATis introduced, business and tax authorities tend to

be in continual negotiation over precise proceduresand coverage and, in the case of exporters, timeli-ness of repayment of VAT export rebates.41

VIII. Tax Assessment and CollectionIn Sections V and VI, I presented the positive

effects of taxation on governance, and in Section VIIsuggested that we need have no great concern thatthe main kinds of taxes employed in poorer coun-tries might be inappropriate to their role of mobiliz-ing taxpayers politically in a constructive way. Un-fortunately, we cannot leave the story here. Taxationhas its dark side: It can be destructive as well asconstructive in governance terms. The quality ofgovernance in some countries would be improved ifsome taxes were simply abolished. In other contexts,major modifications in how taxes are assessed andcollected would be desirable. The problem is theexistence of what I term ‘‘coercive taxation’’: assess-ment and collection conducted in ways that arelikely to be validly perceived by taxpayers as arbi-trary, extractive, unfair, or brutal.42

Coercive taxation takes many forms. We canusefully think of a continuum, defined in terms ofthe degree of brutality and arbitrariness involved.At the harsh end of the spectrum, local governmentemployees in rural China swoop down on a group ofhouses, drag out the household heads, demand thateach make a contribution to the development levy,the veterans’ fund, the slaughter tax, the roadconstruction fund, the land tax, the bicycle tax, thespecial products tax, the entertainment tax, theeducation fund, or any one of dozens of othercharges, and beat those who cannot or will not payand perhaps throw them in jail (Bernstein and Lu2003). At the milder end of the spectrum, revenueauthority officials turn up again at the offices of alarge brewery and explain that while they have

39At the beginning of this decade, VAT was in operation inabout 130 countries and, despite not being used in the UnitedStates, generated around a quarter of the world’s total taxrevenue (Ebrill et al. 2002).

40For example, there was a long resistance from traders inIndia, including many strikes in the form of closure of retailoutlets. It led to large-scale protests and to some riot deathsin Ghana and Uganda. Some of this resistance stems from acombination of bad political tactics on the part of governmentand a perception that is at least partly accurate that VAT isbeing introduced in poor countries at the behest of theinternational financial institutions. VAT was reintroducedrelatively smoothly in Ghana once the government paidattention to political tactics and presentation (World Bank2001). Even when it was first introduced in Europe, VAT wasrecognized to impose a heavy burden on smaller businessesbecause of the additional and relatively complex record-keeping obligations it places on them (Dosser 1981).

41The broader point is that the conceptual distinctionbetween direct and indirect taxes was originally political andhas only more recently been redefined by economists inquestionable theoretical terms: The actual burden of directtaxes is alleged to be met by the person or enterprise thathands over the money to the tax collector, while, it is claimed,the burden of indirect taxes can be passed along the chain ofeconomic transactions until it is borne by the final consumer.VAT may be the first widely used tax that is clearly direct onone criterion and indirect on the other. From that perspective,it is no surprise that in his statistical investigations of thepolitical dimensions of taxation in contemporary poor coun-tries, Mahon (2006) in Chapter 6 found no evidence that thepolitical effects of direct taxes were different from those ofindirect taxes.

42The argument in this section is made in more detail inMoore (2006). For contemporary accounts of coercive taxationin developing countries, see Bernstein and Lu (2003), Fjeld-stad (2001), Juul (2006), and Prud’homme (1992).

Special Reports


(C) T

ax Analysts 2007. A




collected corporate and payroll taxes once for thisquarter, they are behind with the collection targetsset by their bosses, that they would like some moremoney, and that it really would not be in thecompany’s interests to refuse to pay, not least be-cause it can always be shown that the company hasbeen evading some kind of environmental, health, oremployment regulations. These kinds of coercion areall too widespread. Why? There is a formal answer:Taxpayers have few rights. But why do they havefew rights? I don’t have a complete answer, but I cansee five reasons why public authorities in developingcountries are motivated to tax coercively.

Taxation is always potentiallycoercive.

First, taxation is always potentially coercive.State agents have authority to require citizens tohand over money, with no firm guarantee of reci-procity, in situations when they are perceived tohave little or no choice. For various historical rea-sons, states are often powerful relative to mostcitizens in many poorer countries (Moore 2004).State agencies may face relatively few constraints inhow they treat citizens. To some degree, it is not somuch coercion in the tax relationship that demandsexplanation but its absence. Why should tax officialsnot ask for 10,000 shillings and agree then to dropthe official demand to 6,000 shillings in exchange fora bribe of 1,000 shillings?

Second, the structure and organization of econo-mies matters. It is difficult to actually collect taxesfrom low-income agrarian economies organized insmall enterprises that lack formal, bureaucraticstructure and operate without extensive use ofbanking systems and written or electronic records ofeconomic transactions, and to collect without resort-ing to arbitrariness and coercion. In low-income,agrarian economies, tax gathering tends to be coer-cive and conflictual, and the overall tax take tends toremain low.43 Look at the logistical advantages en-joyed by tax collection agencies in wealthier contem-porary economies. Four factors facilitate their task

and help to explain the historical shifts from levyingtaxes on specific items (salt, tobacco, carriages,landed property, houses, individuals) to levyingthem according to accounting categories (especiallyincome, value added, sales, turnover, and profitsrather than simply asset values):

• Extensive written and electronic records of eco-nomic transactions help collectors to hunt downtheir quarry accurately and to create effectivechecks against misappropriation within the taxbureaucracy itself.

• The relative insulation of most economic trans-actions and incomes from seasonality or theweather makes it feasible to collect most taxesin regular installments over the course of ayear.

• The widespread use of banking and other indi-rect systems of money transfer reduces theneed for tax collectors to meet personally withmost taxpayers.

• The prevalence of bureaucratically organizedeconomic enterprises provides opportunities toplace the collection process on the impersonaland quasi-automatic basis that underpins mostcontemporary company taxation and em-ployees’ pay-as-you-earn systems.

Compare that process with collecting taxes inpoor, agrarian societies. Most taxable units aresmall, so the overhead costs of collection tend to behigh. Farm incomes tend to be seasonal and un-stable, so collections are occasional and erratic. Thedearth of records of economic transactions and thelimited use of banking systems encourage face-to-face interaction between taxpayer and tax assessoror collector and oblige the latter to make discretion-ary decisions about tax liabilities that cannot easilybe independently validated. Because a relativelylarge proportion of revenues are eaten up in collec-tion costs, tax administrations are unwilling toseparate the roles of assessing liabilities from actualcollection.44 The man who tells you what you owealso gets his hands on the money. Because it is noteasy to assess tax liabilities according to broadaccounting concepts like income, capital, turnover,

43That is partly why historians quickly tend to associatetaxation with agrarian revolt. One of the more robust statis-tical findings in social science is that both over time andcross-sectionally, official tax collectors in richer countries, andespecially in countries with a relatively small agriculturalsector, succeed in capturing higher proportions of nationalincome for the government. For recent statistical evidence onthis point, see Cheibub (1998), Fauvelle-Aymar (1999), Pian-castelli (2001), Remmer (2004), Ross (2001), and Stotsky andWoldeMariam (1997). The statistical relationship was identi-fied and labeled Wagner’s Law as long ago as 1877.

44The separation of assessment activities from collectionand the assignment of different offices or people to each havebeen central to the construction of noncoercive taxationsystems (for example, Brewer (1989: 101-114) and Daunton(2001)), and are important components of many contemporarytax reforms. In explaining why the Chilean tax administra-tion is so much more efficient than its Argentine counterpart,Bergman points out that in Chile, one department collectsand manages the money while another assesses liabilities andenforces the law. In Argentina all these powers are concen-trated in the same department of the tax agency (Bergman2003: 622).

Special Reports


(C) T

ax Analysts 2007. A




value added, or profit, taxes tend to be attached tospecific goods or services. There is then a wide rangeof potential taxes in law, increasing the scope forcollectors to squeeze taxpayers.45 These logisticalfactors endow tax collectors with considerable dis-cretionary power, facilitate corruption and perhapsextortion, increase the leakage of tax revenues intoprivate hands,46 generate resentment and tax resis-tance on the part of taxpayers, establish taxation asthe issue of choice for political rebels, and make itpractically and politically difficult for governmentsto appropriate a high proportion of national incomethrough taxation, especially direct taxation.

Third, the economic structure of poor agrariansocieties is conducive to coercive taxation for an-other and more directly political reason. One of thedominant themes in the study of the politicaleconomy of taxation is that potential tax gatherershave strong incentives to bargain with, rather thanto coerce, potential taxpayers whose main economicassets are mobile and might be either hidden, with-drawn from production, or moved to another politi-cal jurisdiction. Poor agriculturalists in relativelydensely populated localities are highly immobile inall senses: They cannot easily find land or citizen-ship elsewhere, it is not easy to sell the land theyhave, they cannot afford not to produce, and withouteducation or other capital, their prospects of earninga good living elsewhere are bleak. They are thereforeprime victims for predatory tax collectors.47

Fourth, as far as poor people are concerned, themore predatory tax collectors are often not thecentral government agencies that raise most of thetax revenue but the employees of small-scale localgovernments. Along with the police, armies, andvarious unofficial armed groups who man road-blocks, these local government employees are oftenthe only public officials who find it worthwhile to puteffort into squeezing taxes out of the poor. Some ofthe reasons have been given above — the coercivepotential inherent in the tax relationship and thepaucity of juicy tax bases in poor agrarian societies.We can add in a set of interrelated phenomenacharacteristic of many poor countries: Governmentis relatively centralized; central tax agencies targetthe more lucrative tax bases, leaving local ruraladministrations with little to work with; the recentfashion for decentralization has justified unfundedmandates — that is, devolving service delivery re-sponsibilities to local governments without provid-ing commensurate financial resources; and localgovernments are often almost entirely unregulatedand thus free to act tyrannically.

There is no clear dividing linebetween mildly coercive taxationand the firm enforcement of thelaw. Perceptions of the same eventmay differ radically.

Our fifth and final point takes us up from thelocal agrarian environment to the global level. It issometimes claimed that in some poorer countriesunder the influence of the IMF, central governmenttax administrations raise money coercively becausethey are under pressure to meet annual revenuecollection targets agreed with the IMF (Gloppen andRakner 2002). The truth is elusive. The IMF has aduty to try to ensure that governments raise rev-enue to repay loans or to maintain macroeconomicbalance. It is possible that tax administrationschoose to exaggerate IMF pressure to excuse orjustify extractive taxation, especially extractionfrom those businesses that are large enough toconstitute lucrative revenue sources but do not enjoyspecial political protection. We know that in manypoor countries, national tax administrations tend tofocus on a relatively small number of medium andlarge formal-sector businesses and largely ignorethe others, including large, growing urban informalbusiness activities.48 It is not clear how true it isthat some politically influential big companies get

45Local governments in contemporary agrarian societiesoften levy a wide range of taxes and charges (Bernstein andLu 2003). Central governments have sometimes done so inthe past. In 1952, when Taiwan still had a large agriculturalsector, a range of relatively minor taxes, including a slaughtertax, a household tax, a salt tax, an amusement tax, and afeast tax, collectively accounted for 16 percent of centralgovernment revenues. By 1981, when per capita incomes hadincreased by about 500 percent, some of these taxes had beenabolished, and the remainder accounted for only 2 percent ofrevenue (Republic of China 1982: 29 and 159-160).

46To deal with this problem, there are no satisfactorygeneral alternatives to intrabureaucratic monitoring. In somecontexts, rulers have preferred tax farming, that is, theprivatization of collection. However, the image of the greedy,oppressive tax farmer has thoroughly penetrated public con-sciousness, and the practice has almost disappeared from themodern world. In 18th- and 19th-century Britain, the use oflay assessors was one means of guarding against extortion bytax bureaucrats (Daunton 2001: 200-201), but using taxpay-ers in these roles would inevitably tend to reduce collectionlevels (Brewer 1989: 100).

47Comparing district governments in Tanzania, wherecoercive taxation is widespread (Fjeldstad 2001, Fjeldstadand Semboja 2001), Hoffman and Gibson (2006) have recentlydemonstrated that the more potentially mobile the popula-tions are, the more likely they are to receive services in returnfor their tax payments.

48In Tanzania, with a total population of more than 35million people, 286 large taxpayers pay almost 70 percent of

Special Reports

(Footnote continued on next page.)


(C) T

ax Analysts 2007. A




off lightly.49 There seems to be some kind of viciouscircle in operation in some countries at least: Taxadministrations find it hard to tax the informalsector and the politically influential wealthy, notleast because their staff are reluctant to exposethemselves to the dangers and hardships involved;50

they therefore focus on the medium and largercompanies that are on their books, but not politicallyprotected; and apparent pressure to meet IMF rev-enue targets provides a good excuse for squeezingtheir normal clients, often in the context of emer-gency campaigns to meet targets at the end of afinancial quarter.51 This dynamic effectively deflectsattention away from what should be a major rev-enue policy objective: expanding the tax base byreducing exemptions and bringing new taxpayersinto the net. Those people who do pay tax see thesystem as unfair and are more likely in consequenceto try to evade. It is not clear how big a role the IMFplays in this. But it does at least have the potentialto worsen the situation. As I explained in Section VI,in some poor countries, the IMF in some degreesubstitutes for organized taxpayers in their tradi-tional role as watchdogs over fiscal policy. The IMFhas the power to undermine governance in the longterm by encouraging, however unintentionally, tax

collection practices that worsen relations betweenstates and citizens. It is not clear how far it does thisin reality.52

There is no clear dividing line between mildlycoercive taxation and the firm enforcement of thelaw. Perceptions of the same event may differ radi-cally. Similarly, there can be genuine differencesover whether tax administrations should always tryto treat taxpayers gently for fear of generatingresentment or whether a very firm hand is not bestbecause it deters noncompliance generally, demon-strates the overall fairness of the system, and en-gages taxpayers politically and encourages them tomobilize around fiscal policy issues. One of the moresuccessful cases of improved revenue systems in thepast decade has been the South African RevenueAuthority, which has considerably increased theoverall tax take while providing a better service totaxpayers. It has used a variety of techniques,including the visible services of the ‘‘Scorpions’’ — aspecial investigatory and police force focused oncorruption and organized crime — to deal withrecalcitrant larger companies (Hlophe and Fried-man 2002; Smith 2003). The fairer a tax system isgenerally believed to be, the more it is possible forthe tax administration to take a tough approachwith problematic taxpayers without doing politicaldamage. In South Africa, there is a widespread andplausible perception that the increased revenue willbe put to good use to repay the social debts ofapartheid. Many poorer countries are not in thishappy situation; much revenue is raised coercively,for no convincing purpose, with damaging effects onrelations between states and citizens. To some de-gree, this is an inevitable feature of low-incomeeconomies. As I explained above, wealth and well-functioning markets make taxation easier. That is,however, only a part of the story. A range of policyand institutional reforms, including the simple abo-lition of some types of taxes, could reduce overalllevels of coercion in the tax relationship.

IX. ConclusionIn contemporary poor countries, there are two big

sets of policy questions about the relationship be-tween taxation and the quality of governance. Thefirst set relates to governments that do not need to

the domestic taxes. (The Tanzania Revenue Authority sup-plied Odd-Helge Fjeldstad with the data relating to June2005.) The revenue base excludes many professionals, such aslawyers, doctors, and private consultants, as well as the morestereotypical poor informal sector enterprises. Fewer than 1percent of the taxpayers pay more than 85 percent of thedirect taxes levied in Peru (Mostajo 2004).

49This seems to be part of some standard taxpayers’narrative about unfairness: We ordinary, decent people payour taxes, but the big boys with political influence get awaywith not paying them, as do people operating in the smallscale informal sector or the black market.

50We know from various anecdotal sources that tax agencystaff do not like to be given posts that take them away fromworking in lucrative positions and require them to make agreat deal of effort to raise small amounts of money from veryreluctant taxpayers. The latter posts are attractive mainly tolocal government staffs who lack alternative income earningopportunities. Staffs of central tax agencies want to staywhere the money is, preferably working in an office environ-ment. They do not want to tramp around crowded streetstrying to locate and identify informal businesses, and facehassle and threats for what are likely to be meager pickingsat the end of the day. I am especially grateful to AnuradhaJoshi and Joseph Ayee for information from their research inGhana.

51In his comparative study of Argentina and Chile, Berg-man demonstrates the long-term damage to collection capac-ity that resulted from repeated ‘‘emergency’’ revenue-raisingcampaigns in Argentina (2003: 623).

52It is difficult to research these issues. There is almost nopublished research on the work practices and strategies of taxadministrations. What does exist relates almost entirely tofrontline tax collectors, not to their managers, and to localrather than to central revenue agencies. For some insightfulrecent exceptions, see Jairaj (2005) and Blundo (2006).

Special Reports


(C) T

ax Analysts 2007. A




make much tax effort because they have large non-tax incomes from oil, gas, and mineral exports orfrom foreign aid. The consequent lack of dependenceon citizens tends to consistently undermine thequality of governance. The second set relates tocoercive taxation. This is widespread, and it under-mines the quality of governance and leads to somerather brutal treatment of some poor people byothers almost equally poor.

Fortunately, there are many policy changes thatwould improve the situation, and most could bemade in small steps.53 The greatest scope for imme-diate change lies with aid donors, especially whenthey promise to greatly increase the level of aidgiven to the poorest countries, most of them alreadyhighly aid-dependent. The evidence on the effect ofhigh levels of aid on the quality of governance isambiguous (Section IV). However, providing largeaid volumes without taking steps to encourage theexpansion of efficient domestic revenue-raising ca-pacity in the long term is irresponsible. It would behard for aid donors collectively to acquit themselvesof such a charge. While many do support revenuereform projects, they collectively spend very littlemoney on this activity.54 None make clear state-ments linking aid to improvements in domesticrevenue performance. Most prefer instead to empha-size the need for improved public expenditure man-agement. One does not have to be a conspiracytheorist to realize that this emphasis on expenditurerather than revenue reform is in the institutionalinterests of donor agencies themselves. Especially insituations when many donors give general budgetsupport, aid agency managers worry that corruptionscandals in aid recipient countries will underminepolitical support for aid programs at home. Focusingtheir resources on scrutiny of public expenditures inrecipient countries is a way of reducing this risk. Anemphasis on revenue generation has no such politi-cal side benefits and might indeed invite awkward,skeptical questions about the need for aid in the longterm. Similarly, despite the efforts of some intergov-ernment organizations, aid donors have been suc-cessful at keeping off the policy agenda the fact thataid inflows into developing countries enjoy exactlythe tax exemptions that are decried by the IMF and

tax specialists. If donors trust recipient govern-ments enough to give them general budget support,why do they not suggest that tax exemptions fortheir aid programs be withdrawn? This would notchange the amount of money going to the recipient,but it would help develop the local apparatus forrevenue raising and provide a fine example of prac-ticing what one preaches.55

Many senior tax administrators in developingcountries see themselves as having a professionalmission to free their countries from humiliating aiddependence. Value added taxes, which are now inplace in most countries, are good instruments forraising more revenue fairly.56 A range of other recentreforms provides the basis for making the changes topolicy and practice required so that taxation cancontribute more effectively to the quality of gover-nance. Stimulated partly by the increasing influenceof international associations of tax administratorsand related professions, many tax administrationshave been reorganizing and reorienting themselvesto become more effective and more user-friendly, atleast as far as larger taxpayers are concerned. Theautonomous revenue authorities that have beenestablished in many countries have, like many simi-lar organizations, not lived up to the original, highlyinflated expectations. They have generally contrib-uted to creating more effective revenue systems. Ithas been orthodoxy for some time that tax policyshould focus on widening the tax net, whether byreducing the number and variety of exemptions orby extending registration to a wider range of taxpay-ing businesses and individuals. There is already asubstantial global expert consensus around possibleand desirable tax administration and policy.

There are no great intellectual or organizationalwars that need to be fought and won before govern-ments and aid donors are in a position to change taxpolicies in ways that will improve governance: abol-ish some taxes, replace them with more modern andeffective alternatives that can be levied less coer-cively, use widely known techniques to make the taxpaying experience less coercive, and find nationalrevenue sources to replace aid in the long term.

53These issues are discussed in more detail in Fjeldstadand Moore (2006), in which we make some criticisms ofconventional tax reform policies from a governance perspec-tive.

54Ben Dickinson of the OECD Development AssistanceCommittee has recently calculated that in 2004, official aiddonors, excluding the IMF, spent US $6.6 billion on support toimprove government administration, economic policy, andpublic-sector financial management, and that only 2.7 per-cent of this went to projects that had a taxation component(Ben Dickinson, private communication).

55The case for removing some of the tax exemptions for aidwas recently detailed by International Tax Dialogue (2006).

56VAT remains controversial, partly because it is widelybelieved to be regressive. However, the implementation of aVAT is so partial in many poor countries that the evidencesuggests no general regressive impact (Gemmell and Morris-sey 2005). A more complete and better-managed VAT can bemade nonregressive by zero (or low) rating for consumeritems that account for a high proportion of poor people’sbudgets.

Special Reports


(C) T

ax Analysts 2007. A




ReferencesAcharya, A., A. Fuzzo de Lima, and M. Moore. 2006.

‘‘Proliferation and Fragmentation: TransactionsCosts and the Value of Aid,’’ Journal of DevelopmentStudies 42.1:1-21.

Bates, R. 1993. ‘‘Urban Bias: A Fresh Look,’’ in A.Varshney (ed.), Beyond Urban Bias, London: FrankCass.

—— 1977. Markets and States in Tropical Africa. ThePolitical Basis of Agricultural Policies, Berkeley:University of California Press.

Baunsgaard, T. and M. Keen. 2005. ‘‘Tax Revenue and(or?) Trade Liberalization,’’ Working Paper WP/05/112, Washington: IMF.

Bergman, M.S. 2003. ‘‘Tax Reforms and Tax Compli-ance: The Divergent Paths of Chile and Argentina,’’Journal of Latin American Studies 35.3:593-624.

Bernstein, T.P. and X. Lu. 2003. Taxation WithoutRepresentation in Contemporary China, Cambridge,Mass., and New York: Cambridge University Press.

Blundo, G. 2006. ‘‘Dealing with the Local State: TheInformal Privatization of Street-Level Bureaucra-cies in Senegal,’’ Development and Change 37.4:799-819.

Braddick, M.J. 1996. The Nerves of State. Taxation andthe Financing of the English State, 1588-1714,Manchester, U.K., and New York: Manchester Uni-versity Press.

Brautigam, D. 1991. ‘‘Governance and Economy: AReview,’’ Policy Research Working Papers WPS 815,World Bank.

Brautigam, D. and S. Knack. 2004. ‘‘Foreign Aid, Insti-tutions, and Governance in Sub-Saharan Africa,’’Economic Development and Cultural Change52.2:255-86.

Brewer, J. 1989. The Sinews of Power. War, Money andthe English State, 1688-1783, London: Unwin Hy-man.

Chaudhry, K.A. 1997. The Price of Wealth: Economiesand Institutions in the Middle East, Ithaca, N.Y.,and London: Cornell University Press.

—— 1989. ‘‘The Price of Wealth: Business and State inLabor Remittance and Oil Economies,’’ InternationalOrganization 43.1:101-45.

Cheibub, J.A. 1998. ‘‘Political Regimes and the Extrac-tive Capacity of Governments: Taxation in Democ-racies and Dictatorships,’’ World Politics 50.3:349-76.

Collier, P. 2006. ‘‘Is Aid Oil? An Analysis of WhetherAfrica Can Absorb More Aid,’’ World Development34.9:1482-97.

Collier, P. and A. Hoeffler. 2005. Democracy and Natu-ral Resource Rents, Department of Economics, Ox-ford, U.K.: University of Oxford.

Daunton, M. 2001. Trusting Leviathan. The Politics ofTaxation in Britain 1799-1914, Cambridge, Mass.:Cambridge University Press.

Department for International Development. 2006.Eliminating World Poverty. Making GovernanceWork for the Poor, London: DFID.

Diamond, L. 1992. ‘‘Economic Development and De-mocracy Reconsidered,’’ American Behavioral Scien-tist 35.4/5:450-99.

Dosser, D. 1981. ‘‘The Value Added Tax in the UK andthe EEC,’’ in A. Peacock and F. Forte (eds.), ThePolitical Economy of Taxation, Oxford, U.K.: BasilBlackwell.

Easter, G.M. 2002. ‘‘The Politics of Revenue Extractionin Post-Communist States: Poland and Russia Com-pared,’’ Politics and Society 30.4:599-627.

Ebrill, L., M. Keen, J.-P. Bodin, and V. Summers. 2002.‘‘The Allure of the Value-Added Tax,’’ Finance andDevelopment 39.2.

Ertman, T. 1997. Birth of the Leviathan. BuildingStates and Regimes in Mediaeval and Early ModernEurope, Cambridge, Mass.: Cambridge UniversityPress.

Fauvelle-Aymar, C. 1999. ‘‘The Political and Tax Ca-pacity of Government in Developing Countries,’’Kyklos 52.3:391-413.

Ferguson, N. 2001. The Cash Nexus. Money and Powerin the Modern World 1700-2000, London: Allen Lane,The Penguin Press.

Fjeldstad, O.-H. 2001. ‘‘Local Government Tax Enforce-mentinTanzania,’’JournalofModernAfricanStudies39.2:289-306.

Fjeldstad, O.-H. and M. Moore. 2006. Tax Reform andState Building in a Globalized World, Brighton,U.K.: IDS.

Fjeldstad, O.-H. and J. Semboja. 2001. ‘‘Why PeoplePay Taxes: The Case of the Development Levy inTanzania,’’ World Development 29.12:2059-74.

Gemmell, N. and O. Morrissey. 2005. ‘‘Distribution andPoverty Impacts of Tax Structure Reform in Devel-oping Countries: How Little We Know,’’ DevelopmentPolicy Review 23.2:131-44.

Gervasoni, C. 2006. ‘‘A Rentier Theory of SubnationalAuthoritarian Enclaves: The Politically RegressiveEffects of Progressive Federal Revenue Distribu-tion,’’ paper presented at the annual meeting of theAmerican Political Science Association, Philadel-phia.

Gloppen, S. and L. Rakner. 2002. ‘‘AccountabilityThrough Tax Reform? Reflections From Sub-Saharan Africa,’’ IDS Bulletin 33.3:30-40.

Goldscheid, R. 1958 (1925). ‘‘A Sociological Approach toProblems of Public Finance,’’ in R.A. Musgrave andA.T. Peacock (eds.), Classics in the Theory of PublicFinance, London and New York: Macmillan (firstpublished in Tubingen in 1925).

Guyer, J.I. 1992. ‘‘Representation Without Taxation: AnEssay on Democracy in Rural Nigeria, 1952-1990,’’African Studies Review 35.1:41-80.

Special Reports


(C) T

ax Analysts 2007. A




Heo, U. and A.C. Tan. 2001. ‘‘Democracy and EconomicGrowth: A Causal Analysis,’’ Comparative Politics33.4:463-74.

Hlophe, D. and S. Friedman. 2002. ‘‘And Their Heartsand Minds Will Follow? Tax Collection, Authorityand Legitimacy in Democratic South Africa,’’ IDSBulletin 33.3:67-76.

Hoffman, B.D. and C.C. Gibson. 2006. ‘‘Mobility andthe Political Economy of Taxation in Tanzania,’’paper presented at the annual meeting of the Ameri-can Political Science Association, Philadelphia.

Huntington, S.P. 1991. The Third Wave: Democratiza-tion in the Late Twentieth Century, Norman, Okla.,and London: University of Oklahoma Press.

International Tax Dialogue. 2006. ‘‘Tax Treatment ofDonor-Financed Projects,’’ discussion paper pre-pared by the International Tax Dialogue, Paris:OECD Development Assistance Committee.

Jairaj, A. 2005. ‘‘The Political Economy of CommodityTaxation in Tamil Nadu,’’ Masters in PhilosophyThesis, Queen Elizabeth House, Oxford University.

Juul, K. 2006. ‘‘Decentralization, Local Taxation andCitizenship in Senegal,’’ Development and Change37.4:821-46.

Karl, T.L. 1997. The Paradox of Plenty. Oil Booms andPetro-States, Berkeley, Calif., Los Angeles, and Lon-don: California University Press.

Kaufmann, D., A. Kraay, and M. Mastruzzi. 2006.Governance Matters V: Aggregate and InternationalGovernance Indicators for 1996-2005, World Bank.

Landa, D. and E.B. Kapstein. 2001. ‘‘Review Article:Inequality, Growth, and Democracy,’’ World Politics53.1:264-96.

Levi, M. 1997. Consent, Dissent, and Patriotism, Cam-bridge, Mass., and New York: Cambridge UniversityPress.

Lipton, M. 1977. Why Poor People Stay Poor. UrbanBias in World Development, London: Temple Smith.

Londregan, J.B. and K.T. Poole. 1996. ‘‘Does HighIncome Promote Democracy?’’ World Politics 49.1:1-30.

Luciani, G. 1994. ‘‘The Oil Rent, the Fiscal Crisis of theState and Democratization,’’ in G. Salame (ed.),Democracy Without Democrats? The Renewal of Poli-tics in the Arab World, London and New York: I.B.Tauris.

Mahdavy, H. 1970. ‘‘The Patterns and Problems ofEconomic Development in Rentier States: The Caseof Iran,’’ in M.A. Cook (ed.), Studies in the EconomicHistory of the Middle East from the Rise of Islam tothe Present Day, London: Oxford University Press.

Mahon, J.E. 2006. ‘‘Revenue and Regimes,’’ mimeo,Washington: Williams College.

—— 2004. ‘‘Causes of Tax Reform in Latin America,1977-95,’’ Latin American Research Review 39.1:3-30.

Mann, M. 1993. The Sources of Social Power, Vol. 2: TheRise of Classes and Nation-States, 1760-1914, Cam-bridge, Mass., and New York: Cambridge UniversityPress.

Moore, M. 2006. ‘‘Between Coercion and Contract:Competing Narratives around Taxation and Gover-nance,’’ mimeo, IDS, University of Sussex.

—— 2004. ‘‘The Development of Political Underdevel-opment,’’ in G. Harrison (ed.), Global Encounters:IPE, Development and Global Governance, London:Palgrave Macmillan.

—— 2003. ‘‘Arguing the Politics of Inclusion,’’ in P.Houtzager and M. Moore (eds.), Changing Paths:International Development and the New Politics ofInclusion, Ann Arbor, Mich.: University of MichiganPress.

—— 1998. ‘‘Death Without Taxes: Democracy, StateCapacity, and Aid Dependence in the Fourth World,’’in M. Robinson and G. White (eds.), The DemocraticDevelopmental State: Politics and Institutional De-sign, Oxford, U.K.: Oxford University Press.

—— 1993. ‘‘Economic Structure and the Politics ofSectoral Bias,’’ Journal of Development Studies29.4:79-128.

Moss, T., G. Petterson, and N. Van de Walle. 2006. ‘‘AnAid-Institutions Paradox? A Review Essay on AidDependency and State Building in Sub-SaharanAfrica,’’ Working Paper 74, Washington: Center forGlobal Development.

Mostajo, R. 2004. Prospects for an Integral Tax ReformWith Equity: Towards a Fiscal Covenant for Peru,Lima: DFID.

Odon, W.E. 1992. On Internal War. American andSoviet Approaches to Third World Clients and Insur-gents, Durham, N.C., and London: Duke UniversityPress.

Piancastelli, M. 2001. ‘‘Measuring the Tax Effort ofDeveloped and Developing Countries. Cross-Country Panel Data Analysis — 1985/95,’’ Discus-sion Paper 818, Rio de Janeiro, Brazil: Instituto dePesquisa Economica Aplicada.

PricewaterhouseCoopers and World Bank. 2006. ‘‘Pay-ing Taxes. The Global Picture,’’ Washington: WorldBank.

Prud’homme, R. 1992. ‘‘Informal Local Taxation inDeveloping Countries,’’ Environment and PlanningC: Government and Policy 10:1-17.

Przeworski, A., M. Alvarez, J.A. Cheibub, and F. Li-mongi. 1996. ‘‘What Makes Democracies Endure?’’Journal of Democracy 7.1:39-55.

Remmer, K.L. 2004. ‘‘Does Foreign Aid Promote theExpansion of Government?’’ American Journal ofPolitical Science 48.1:77-92.

Republic of China. 1982. Taiwan Statistical Data Book,Taipei: Council of Economic Planning and Develop-ment.

Rosenthal, J.-L. 1998. ‘‘The Political Economy of Abso-lutism Reconsidered,’’ in Robert H. Bates, AvnerGreif, Margaret Levi, Barry R. Weingast, and

Special Reports


(C) T

ax Analysts 2007. A




Jean-Laurent Rosenthal (eds.), Analytic Narratives,Princeton, N.J.: Princeton University Press.

Ross, M.L. 2004. ‘‘Does Taxation Lead to Representa-tion?’’ British Journal of Political Science 34.2:229-49.

—— 2001. ‘‘Does Oil Hurt Democracy?’’ World Politics53.3:325-61.

—— 1999. ‘‘The Political Economy of the ResourceCurse,’’ World Politics 51.1:297-322.

Rosser, A. 2006. ‘‘The Political Economy of the ResourceCurse: A Literature Survey,’’ IDS Working Paper268, Brighton, U.K.: IDS.

Schedler, A. 1999. ‘‘Conceptualizing Accountability,’’ inA. Schedler et al. (eds.), The Self-Restraining State.Power and Accountability in New Democracies,Boulder, Colo., and London: Lynne Rienner.

Schumpeter, J.A. 1918/1991. ‘‘The Crisis of the TaxState,’’ in R.A. Swedberg (ed.), Joseph A. Schum-peter. The Economics and Sociology of Capitalism,Princeton, N.J.: Princeton University Press.

Skocpol, T. 1992. Protecting Soldiers and Mothers. ThePolitical Origins of Social Policy in the UnitedStates, Cambridge, Mass., and London: The BelknapPress of Harvard University Press.

Smith, L. 2003. ‘‘The Power of Politics. The Perform-ance of the South African Revenue Service and Someof Its Implications,’’ Policy: Issues and Actors, 16.2,Johannesburg: Centre for Policy Studies.

Stotsky, J.G., and A. WoldeMariam. 1997. ‘‘Tax Effortin Sub-Saharan Africa,’’ Working Paper WP/97/107,Washington: International Monetary Fund

Tilly, C. 1992. Coercion, Capital and European States,AD 990-1992, Cambridge, Mass., and Oxford, U.K.:Blackwell.

Vandewalle, D. 1998. Libya Since Independence. Oiland State-Building, Ithaca, N.Y., and London: Cor-nell University Press.

Winters, J.A. 1996. Power in Motion. Capital and theIndonesian State, Ithaca, N.Y., and London: CornellUniversity Press.

World Bank. 2006. Doing Business 2007: How to Re-form, Washington: World Bank.

—— 2004. World Development Report 2004: MakingServices Work for Poor People, Washington: WorldBank.

—— 2001. Introducing a Value Added Tax: LessonsFrom Ghana, PREM Notes 61, Washington.

—— 1997. World Development Report 1997: The Statein a Changing World, New York: Oxford UniversityPress. ◆

Special Reports


(C) T

ax Analysts 2007. A




Documents

How DoesTaxation Affect the Quality of Governance?(‘‘No taxation without representation.’’) There is also a more theoretical tradition of aca-demic work that links long-term