ANRV377-PL12-07 ARI 7 April 2009 16:12

Variation in InstitutionalStrengthSteven Levitsky1 and Marıa Victoria Murillo2

1Weatherhead Center for International Affairs, Harvard University, Cambridge,Massachusetts 02138; email: Levitsky@wcfia.harvard.edu2Department of Political Science, Columbia University, New York, New York 10027;email: Mm2140@columbia.edu

Annu. Rev. Polit. Sci. 2009. 12:115–33

The Annual Review of Political Science is online atpolisci.annualreviews.org

This article’s doi:10.1146/annurev.polisci.11.091106.121756

Copyright c© 2009 by Annual Reviews.All rights reserved

1094-2939/09/0615-0115$20.00

Key Words

political institutions, institutional development, weak institutions,enforcement, stability, developing countries

AbstractComparative research on political institutions has begun to turn fromissues of formal institutional design to issues of institutional strength.Rather than assuming a tight fit between formal rules and politicalbehavior, these studies examine how variation in the stability and/orenforcement of formal rules shapes actors’ expectations and behav-ior. This article explores the emerging research agenda on institutionalstrength. It disaggregates the concept of institutional strength into twodimensions—enforcement and stability—and it argues that institutionsvary widely on both dimensions. The article then examines the sourcesof this variation and its implications for comparative research. It showshow recent research on weak institutions may be used to refine existingtheories of institutional effects, design, and development, which shouldbroaden the comparative scope of these theories. The conclusion exam-ines ways of developing comparative measures of institutional strength.

115

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

INTRODUCTION

The institutionalist turn in comparative pol-itics focused considerable attention on issuesof formal institutional design. Given the ex-tent of democratization across the globe dur-ing the last part of the twentieth century, thisfocus was not surprising. The proliferation ofnew constitutions, electoral systems, constitu-tional courts, central banks, and other institu-tions created an unprecedented opportunity forscholars to study the causes and consequencesof institutional design. Moreover, the spread ofdemocracy meant that constitutional arrange-ments in the developing and postcommunistworlds increasingly resembled those in the ad-vanced industrialized countries, which seem-ingly presented scholars with an opportunityto test existing theories—nearly all of whichwere based on studies of the United States andWestern Europe—on a new set of cases.

More recently, however, research on polit-ical institutions in the developing world hasturned to another, heretofore understudied, di-mension: institutional strength. Debates aboutconstitutional design during the 1990s oftentook for granted that the rules being de-signed would be enforced and minimally sta-ble. In effect, parchment rules (Carey 2000)were equated with institutions, implying thatthe rules that were written down on papereither reflected or generated shared expecta-tions about how others would behave. Althoughthese assumptions often hold up well in stud-ies of advanced democracies, where underlyingrule-making frameworks are stable and effectivestates ensure that most rules are regularly en-forced, they do not travel as well to the develop-ing world. An emerging body of research showsthat many formal institutions in Africa, LatinAmerica, Asia, and the former Soviet Unionare neither minimally stable nor regularly en-forced. In other words, instead of taking rootand generating shared behavioral expectations,formal rules are widely contested, routinely vi-olated, and frequently changed.

For example, as this article was being writ-ten, Bolivia and Ecuador were drafting new

constitutions aimed at “refounding the repub-lic.” Yet both countries had already written newconstitutions during the 1990s; in fact, eachhad written at least 16 different constitutions—about one per decade—since independence. InArgentina, lifetime tenure security for SupremeCourt justices has been enshrined in the consti-tution since 1853. Nevertheless, virtually everyincoming government since the 1940s, includ-ing three of the first four presidents elected inthe post-1983 democratic era, packed the courtwith impunity (Helmke 2004). As Huntington(1968) argued four decades ago, a major prob-lem in much of the developing world is that in-stitutions of all types have consistently failed totake root. Such institutional weakness has far-reaching implications. In the absence of stableand effective rules of the game, both democ-racy and economic growth are difficult to sus-tain (O’Donnell 1993, 1994; Spiller & Tommasi2007).

This article examines recent research on thecauses and consequences of institutional weak-ness, seeking to bring a large but disparatebody of work together into a more coherentresearch agenda. Recent studies of political in-stitutions in the developing and postcommu-nist world show that formal institutions varywidely along two dimensions: enforcement andstability. This variation has important theoret-ical implications. Where actors do not expectformal institutions to endure or be enforced,their behavior will differ, often markedly. Ex-isting theories about the design and effects offormal institutions may thus need to be refinedif they are to be usefully applied to cases of in-stitutional weakness. By treating institutionalstrength as a variable, rather than a taken-for-granted assumption, recent scholarship has be-gun to refine institutionalist theories in waysthat enhance their utility in contexts—such asmuch of the developing world—in which strongformal institutions are the exception, not therule.

The article is organized into three mainsections. The first section disaggregates theconcept of institutional strength into two

116 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

dimensions, enforcement and stability. It thenshows how real-world institutions vary alongboth dimensions. The second section exam-ines some potential causes of variation in in-stitutional enforcement and stability. The thirdsection discusses the implications of variationin institutional strength for theory building incomparative politics.

THE CONCEPT OFINSTITUTIONAL STRENGTH:ENFORCEMENT AND STABILITY

Political institutions may be defined as hu-manly devised rules and procedures—both for-mal and informal—that constrain and enablepolitical behavior. Notwithstanding the impor-tance of informal rules and procedures (March& Olsen 1989, North 1990a, O’Donnell 1996,Helmke & Levitsky 2006), this article focuseson formal or written rules, or what Carey(2000) calls “parchment” institutions. Follow-ing North (1990a, pp. 4–5), we distinguishbetween institutions (the “rules of the game”)and organizations (the “players”).

Institutional strength may be conceptual-ized along two dimensions: enforcement andstability. Enforcement is the degree to whichparchment rules are complied with in practice.Where all relevant actors in a given territoryroutinely comply with parchment rules or facea high risk of punishment, enforcement is high.1

By stability, we mean durability. Institutionsare stable to the degree that they survive notonly the passage of time but also changes in theconditions—i.e., underlying power and pref-erence distributions—under which they wereinitially created and reproduced.2 Institutional

1We do not equate enforcement with compliance. Althoughenforcement raises the cost of noncompliance, actors oftencomply with rules for other reasons, including a perceptionof the rule’s fairness, ethical reciprocity in response to others’compliance, ideological agreement, or sheer habit (Levi 1988,1997).2This conceptualization is rooted in the historical institu-tionalist tradition. Scholars from a rational choice traditiongenerally expect institutions to change in response to changes

Enforcement High Low

Highstrong formal institutions stable (but weakly enforced)

formal institutionsStability

Lowunstable (but enforced)

formal institutions

weak formal institutions

Figure 1Two dimensions of institutional strength.

instability must be distinguished from “normal”institutional change. Because even the mostrobust rules evolve, one or more instances ofinstitutional change cannot necessarily be takenas an indicator of instability. Rather, institu-tional instability is best understood as a pat-tern in which, given a common environment(i.e., exogenous shocks, changes in power andpreference distributions), a particular institu-tional arrangement changes with greater fre-quency than other similarly designed ones.For example, if Bolivia and Ecuador have hadsix different constitutions since World WarII, whereas Mexico has had only one (de-spite considerable change in regime coalitions,power distributions, and elite preferences), thenthe Bolivian and Ecuadorian constitutions maybe considered less stable than the Mexicanone.

Enforcement and stability do not neces-sarily vary together. As Figure 1 shows, thetwo dimensions may be combined to producefour distinct types. The upper left quadrant ofFigure 1 corresponds to strong institutions, orthose that are both stable and enforced. It isthese “pure” cases of institutional strength—characterized by a tight coupling betweenformal rules and actual behavior—that predom-inate in most of the theories of institutionsgenerated by studies of advanced industrial-ized democracies. Institutional strength is notexclusive to established democracies, however.

in power and preference distributions (see Knight 1992). Inour conceptualization, institutions may be stable even thoughtheir goals change (see Huntington 1968). Here we differfrom Streeck & Thelen (2005), who treat goal adaptation asa form of institutional change.

www.annualreviews.org • Variation in Institutional Strength 117

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Mexico sustained and enforced a ban on presi-dential and legislative re-election throughoutmost of the twentieth century. Similarly, Ar-gentina’s provincial list-based proportional rep-resentation (PR) electoral system, Brazil’s open-list PR system, and Chile’s binomial electoralsystem remained stable and enforced during the1990s and 2000s, despite widespread criticismand demands for change.

The lower left quadrant corresponds to in-stitutions that are enforced but unstable. In suchcases, parchment rules are widely followed, butthey are modified with comparatively high fre-quency, such that actors are unable to developshared expectations based on past behavior. Anexample is electoral rules in Poland. Every Pol-ish election in the first postcommunist decadewas “preceded by a contest among political par-ties to modify the electoral law, resulting in foursubstantively different electoral laws” (Benoit &Hayden 2004, p. 397). Thus, the electoral sys-tem shifted from majority runoff in the 1989Sejm elections to a two-tiered Hare PR systemin 1991, to a D’Hondt PR system with higherthresholds in 1993, to a single-tiered modi-fied Sainte-Lague PR system in 2001 (Benoit& Hayden 2004). Likewise, Madagascar under-went three major constitutional reforms duringthe 1990s: In 1992, it replaced a superpresiden-tialist constitution with French-style semipres-identialism; in 1995, it adopted a modifiedsemipresidential system in which presidentscould dismiss prime ministers; and in 1998, itreverted back to a presidential system with adominant executive (Marcus 2005).

The upper right quadrant correspondsto formal institutions that are stable but notenforced. In other words, rules remain on thebooks for long periods of time but are routinelyignored, serving, in effect, as “window dress-ing.” Frequently in such cases, actors adhereto established informal rules of the game.For example, many clauses of Mexico’s 1917constitution—including those prescribingfair elections, judicial tenure security, and aweak presidency—were ignored for decades.Although the political rules of the game were

relatively stable after the 1930s, many of theserules—including the well-known dedazo, inwhich presidents single-handedly selectedtheir successors—were informal (Weldon1997, Langston 2006).

Socialist institutions in contemporary Chinaand Vietnam provide another example. In bothcountries, socialist institutions remained for-mally in place during the 1990s—perhaps be-cause their elimination would threaten elitecohesion or regime legitimacy—but were in-creasingly abandoned in practice. In China,laws banning private property remained on thebooks, but economic agents—with the collu-sion of local state officials— developed “quasi-legal” arrangements to operate private firms,obtain credit, and raise capital (Tsai 2007). InVietnam, socialist laws restricting foreign in-vestment were ignored during the 1990s aslocal state officials carried out “fence break-ing” liberalization measures (Malesky 2005,p. 12). In both cases, the failure to enforceexisting laws, while keeping them on thebooks, transformed once-effective institutionsinto window-dressing institutions.

Finally, the lower right quadrant ofFigure 1 corresponds to institutions that areneither stable nor regularly enforced. Here thecoupling between formal rules and actual be-havior is very loose. Rules exist on parchment,but in practice, they do little to constrain ac-tors or shape their expectations.3 For exam-ple, Bolivia’s first constitution, written in 1826,was “never fully implemented” (Loveman 1993,p. 238) and was abolished within three years.Successor constitutions in 1831 and 1834

3Even these “pure” cases of institutional weakness differ fromthe absence of an institution. Because they exist on paper, evenunstable or weakly enforced institutions may softly guide ac-tors’ expectations by providing at least some cues about oth-ers’ behavior. Even when parchment rules are unlikely to beenforced or complied with, actors must at least take them intoaccount, incorporating the low probability of their predictedeffects into their strategic calculation. Indeed, rules that arewidely subverted may nevertheless serve as a point of refer-ence in the menu of available options, with a slightly higherprobability of occurrence than other alternatives.

118 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

were also quickly discarded, and new consti-tutions were promulgated in 1839, 1843, 1851,and 1861 (Loveman 1993, pp. 239–60). Gen-eral Mariano Melgarejo “canceled” the 1861constitution after seizing power in 1864, andhe subsequently promulgated a new one “ona whim” (Loveman 1993, pp. 251–52). At abanquet celebrating the new charter, Melgarejodeclared:

I want the [gentleman] who just spoke to know,and all the honorable congressmen also, thatthe 1861 Constitution, which was a very goodone, I stuck in this [left] pocket, and that of1868, which is even better according to these[gentlemen], I stuck in this other pocket, andthat no one but me rules Bolivia. (Loveman1993, p. 252)

Many contemporary political institutionscombine instability and low enforcement. Forexample, Pakistan’s postcolonial constitutionalstructure, which prescribed a strong parliamentand a weak executive, was neither stable nor en-forced during the country’s first four decadesof independence. At independence in 1947,Pakistan was a “British-style parliamentarypolity” in which presidents formally had “nodirect role in government except in the gravestemergencies” (Rose & Evans 1997, p. 95).However, founding leader Mohammad AliJinnah established a pattern of de facto ex-ecutive authority that greatly exceeded thatprescribed by either the colonial charter orthe postcolonial (1956) constitution (Sherwani1962; Rose & Evans 1997, p. 95). The par-liamentary constitution was dissolved in 1958following a military coup, and the constitu-tion promulgated by military ruler MuhammadAyub Khan in 1962 briefly established a formalpresidential system (Sherwani 1962). However,Abu Khan’s constitution did not survive his fallfrom power in 1969, and in 1973, Pakistan againadopted a parliamentary constitution. Between1973 and 1985, formal limits on presidentialpower were again weakly enforced, and in 1985,a constitutional amendment formally strength-

ened the presidency (Rose & Evans 1997,p. 95).

Whereas the literature on political institu-tions has focused primarily on the upper leftquadrant of Figure 1, all four of these insti-tutional patterns are common in the develop-ing and postcommunist world. Moreover, allfour patterns may be enduring. For example,some formal rules endure precisely because theyare not enforced. Take the case of postrevo-lutionary Mexico. Had the 1917 constitutionbeen rigorously enforced, such that electionswere truly clean and competitive, the resultmight well have been intense political conflictthat could have threatened the stability of thePRI regime. In other words, the stability of theregime rested, in part, on a systematic failureto comply with core elements of the 1917 con-stitution. At the same time, enforcement may,in some cases, undermine institutional stabil-ity. Take Argentina: Between 1930 and 1943,Argentine conservatives maintained power viafraud-ridden elections. After 1943, however,electoral rules were followed, in that therewas little fraud or manipulation of the vote-counting process. A consequence of this com-pliance, however, was increased institutionalinstability. Powerful actors who opposed par-ticular electoral outcomes, usually Peronist vic-tories, repeatedly undermined (via coups) andmodified electoral institutions—shifting froma plurality system (1958, 1963) to a two-roundpresidential run-off system (1973) and back toplurality (1983)—in an effort to prevent thoseoutcomes.

EXPLAINING INSTITUTIONALWEAKNESS

What explains variation in the strength ofparchment institutions? Institutions themselvesvary in stability and enforcement. For exam-ple, laws against prostitution, euthanasia, andthe hiring of undocumented immigrants tendto be weakly enforced even in countries witheffective legal systems (Van Oenen 2001). Ourprimary interest, however, lies in cross-national

www.annualreviews.org • Variation in Institutional Strength 119

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

variation in institutional strength. Why wascentral bank autonomy respected in post-Pinochet Chile, while it has been repeatedlyviolated in Argentina? Why did presidentialterm limits remain intact for nearly a century inMexico, while they were revised or scrappedonly a few years after being enshrined in con-stitutions in Belarus, Cameroon, the Domini-can Republic, Gabon, Kazakhstan, Namibia,Uganda, and elsewhere? During the 1990s, asthe global diffusion of formal democratic andmarket institutions focused scholars’ attentionon issues of institutional design, these questionswere often neglected. Indeed, debates over in-stitutional design often took for granted thatparchment rules would be stable and effective.In other words, it was assumed that the actorswho wrote the rules would have the will and ca-pacity to enforce them. Yet formal institutionsare often “born weak,” either because the actorswho create them lack an interest in enforcing(or sustaining) them or because they lack thecapacity to enforce (or sustain) them. In thissection, we explore some of the conditions thatunderlie such institutional weakness.

Explaining Weakly EnforcedInstitutions

Weakly enforced institutions may be createdintentionally or unintentionally. Many institu-tions are weak because the actors who createthem do not intend to enforce them. Thus, oneimpetus for the emergence of weakly enforcedinstitutions is a divergence between political ac-tors’ real and publicly stated goals. Formal in-stitutions may be so widely viewed as “appro-priate” that power holders adopt them not inpursuit of the ends for which they are designed,but rather in search of international or domesticlegitimacy—the phenomenon that sociologicalinstitutionalists call isomorphism (DiMaggio &Powell 1991, Meyer & Rowan 1991). In theirextreme form, such institutions may effectivelyserve as window dressing, in that power holdershave an interest in keeping them on the booksbut no interest in enforcing them. These insti-

tutions may be stable, but stability is based ona shared expectation that the rules will not beenforced.

In the developing world, window-dressinginstitutions are often a response to internationaldemands or expectations. Dependence on inter-national assistance creates incentives for stateelites to borrow the form—though not neces-sarily the substance—of Western-style institu-tions ( Joseph 1999, Van de Walle 2001). Takeelections. During the 1990s, the unprecedentedpower of Western capitalist democracies andabsence of internationally legitimate regimealternatives created incentives for elites in pe-ripheral states to hold multiparty elections.However, in much of the former Soviet Unionand Africa, ruling elites adopted the formalarchitecture of democracy without any inten-tion of allowing those institutions to loosentheir grip on power ( Joseph 1999, Levitsky &Way 2002). In many cases, what emerged was a“virtual democracy,” a regime whose formal in-stitutions had more to do with “ritual and sym-bol, appearance and presentability” than withpolitical competition ( Joseph 1999, p. 13).

Window-dressing institutions may also becreated in pursuit of domestic legitimacy. Somelaws are written (or remain on the books) be-cause large numbers of citizens view them asmorally appropriate, often because they arerooted in religious codes or widely shared con-ceptions of basic rights. Thus, even when stateofficials are unable or unwilling to system-atically enforce these laws, their removal ispolitically unthinkable. Examples of such de-liberate nonenforcement include laws againstprostitution, soft drug use, euthanasia, and insome countries, abortion (van Oenen 2001). Al-though punitive anti-divorce and anti-abortionstatutes were on the books in Argentina, Brazil,and Chile throughout most of the twenti-eth century, they were “hardly ever enforced”(Htun 2003, pp. 153–54). According to Htun,the laws were designed primarily to achievemoral or symbolic goals rather than to shapebehavior. For example, after a divorce ban waswritten into Brazil’s 1946 constitution, one

120 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

commentator remarked that “the dispositionwas maintained for love and hypocrisy. Noone who voted for it denied that the [rule]is. . .incapable of deterring what is apparent inBrazilian society: de facto divorcing and result-ing de facto marriages” (Htun 2003, p. 86).

Weakly enforced institutions may also beunintended. Actors who create formal rulesand procedures may lack the power to makethem binding on all actors. Indeed, a majorcause of weak enforcement is limited state ca-pacity to monitor and sanction. Low bureau-cratic capacity—for example, the absence ofa professional civil service or an infrastruc-ture that effectively penetrates the nationalterritory—limits state actors’ ability to monitorrule violations (Gonzales & King 2004). Wherepublic agencies and officials responsible for en-forcement are weak or absent, actors’ incen-tives to comply with official rules—regardlessof their original preferences—will be weak.Recent studies have linked low enforcement,or the “unrule of law,” to the limited reachof the state (O’Donnell 1993, 1999; Herbst2000; Gonzales & King 2004; Yashar 2005;Soifer 2006). For example, Soifer (2006) showshow repeated government failure to enforce tax,conscription, and other laws in Peru was rootedin the state’s limited infrastructural power in thecountryside. Likewise, the nonenforcement ofcivil rights legislation in the U.S. South priorto the 1950s has been attributed to the limitedreach of the state (Gonzales & King 2004).

Weakly enforced institutions may also berooted in a disjuncture between those actorswith the formal authority to make rules andthose with “real” power. In most establisheddemocracies, formal rule-making authorities(e.g., parliaments, supreme courts) are fullyvested with power, and as a result, the rules theyproduce are generally made binding on all ac-tors in society. In many polities, however, for-mal and de facto powers are not so neatly fused.Those who write the rules do not always havethe power to systematically enforce them. Forexample, monarchic, military, or religious au-thorities may exert informal “tutelary” power

over elected governments, limiting their capac-ity to enforce the law. Although Guatemala’s1985 constitution declared the army to be “apo-litical, obedient, and nondeliberative,” there ex-isted “few countervailing forces to offset thepower of the armed forces” during the 1980s(Fitch 1998, p. 54), and as a result, the army “re-tain[ed] control of the state from its barracks”(Torres-Rivas 1996, p. 52). Likewise, althoughthe constitution enshrined human rights guar-antees, President Vinicio Cerezo acknowledgedthat he was unable to enforce these rights (Fitch1998, pp. 54–55), and indeed, “human rightswere violated in the same manner as in the eraof military governments” (Torres-Rivas 1996,p. 53).

In some cases, weak enforcement is unin-tended but not unexpected. Reformers mayadopt ambitious laws that are unlikely to beevenly enforced initially but that potentiallyempower weaker actors to pursue their imple-mentation via the courts. In other words, rulewriters are aware of their inability to make for-mal rules binding, but they nevertheless writethose rules in an effort to strengthen their handin future bargaining vis-a-vis powerful actors.For example, constitutions in Brazil (1988) andSouth Africa (1996) included a variety of so-cial or “third-generation” rights—such as racialand gender equality and rights to shelter, healthcare, and a clean environment—that the statecould not guarantee at the time but that ac-tivists pushed for in the hope of making themeffective through subsequent legislation andjudicial action (Klug 2000, pp. 91–92; Htun2003, p. 126).4 Although enforcement of theserights proved uneven, subsequent legal actiondid place issues of compliance on the agenda(Pieterse 2004; Htun 2003, p. 126).

Weak enforcement may also be rooted invarying levels of societal compliance. As Levihas argued, functioning rules generally elicit

4This strategy was influenced by the U.S. experience, whereconstitutional civil rights were unevenly enforced (acrossracial groups and regions) for decades, and the courts ulti-mately played a central role in increasing enforcement.

www.annualreviews.org • Variation in Institutional Strength 121

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

high levels of “quasi-voluntary compliance,”which, in turn, is rooted in perceptions that therules are fair and that others will comply withthe rules or be punished (Levi 1988, pp. 52–70; 1997, pp. 19–25). Where levels of quasi-voluntary compliance are high, monitoring andenforcement costs are reduced, as state agenciescan focus scarce resources on those actors whoare most likely to be noncompliant (Levi 1988,p. 54). Where quasi-voluntary complianceis low, states must rely almost entirely on coer-cion, which can be costly and ineffective (Levi1988, p. 50).

Finally, weak enforcement may be rooted ininequality. The coexistence of political democ-racy and high levels of social, economic, orethnoreligious inequalities creates conditionsunder which certain rules are less likely to be en-forced. On the one hand, privileged elites havean incentive to evade laws or policies that allowdisadvantaged groups to translate their politicalrights into claims for greater substantive equal-ity. Lieberman (2003) has shown that in Brazil,where nominal political equality (via multira-cial citizenship) coexisted with extreme racialand socioeconomic inequality, the wealthy sys-tematically failed to comply with tax laws that,if enforced, would have permitted a redistribu-tion of wealth. On the other hand, enormouswealth disparities make it harder for those wholack resources to demand effective equality be-fore the law, as they are often poorly informedabout their rights and lack both access to thejustice system and the capacity to monitor thebehavior of state authorities.

Explaining Variation inInstitutional Stability

Recent studies have found considerable cross-national (and within-country) variation in thestability of formal political institutions (Steinet al. 2006, Spiller & Tommasi 2007). Never-theless, scholars are only beginning to explorethe sources of this variation (see Clemens &Cook 1999, Helmke 2007). Like weak enforce-ment, instability is often rooted in a disjuncturebetween rule writers and power holders. Such

disjunctures often emerge when the balanceof forces is uncertain or rapidly shifting. Al-though an uncertain balance of forces may yielda compromise arrangement that protects the in-terests of all major players (Przeworski 1991,pp. 87–88), it may just as easily lead to miscal-culation and error. Those in temporary controlof the rule-writing process may fail to take realpower distributions into account, leaving newlydesigned institutions vulnerable to assault bypowerful actors who remain outside the pro-cess. In Gabon, for example, when a pluralistNational Conference was authorized to drawup a new multiparty constitution in 1990, “po-litically intoxicated” opposition forces createda range of formal checks on presidential au-thority (e.g., independent electoral authority,term limits), but they failed to strip PresidentOmar Bongo of control over state resources orthe coercive apparatus (Messone & Gros 1998,p. 138). With his real power intact, Bongo dis-mantled most of these institutions in the yearsthat followed.

In transitional settings, new institutional ar-rangements are most likely to endure whererule writers either (a) gain the acceptance ofpowerful actors and groups who remain outsidethe rule-writing process or (b) decisively defeatmajor opponents, thereby destroying their ca-pacity to overturn the rules in the future (seePrzeworski 1991, pp. 81–88). Where neitheroccurs, powerful actors who lose out underthe new institutional arrangements are likelyto work to overturn them as soon as theyare in a position to do so. For example, thePeron government used its electoral majority toimpose a constitution in 1949 without incor-porating opposition Radicals, who maintainedconsiderable power resources despite their elec-toral weakness. These groups never acceptedthe new rules of the game, and when Peron wastoppled six years later, his constitution was dis-carded. By contrast, Argentina’s 1853 constitu-tion (which was imposed on all provinces by thewinners of a civil war) and its 1994 constitution(the product of a negotiated agreement betweenPeronist and Radical leaders) proved morestable.

122 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Institutional stability is also rooted in time.Time matters in at least two senses. First, asGrzymala-Busse (2006) has argued, the pace ofinstitutional design may affect stability. Whereinstitutions are created slowly, actors have moretime to evaluate their (often unintended) con-sequences, calculate how the rules affect theirinterests, and organize for the collective defenseof (or opposition to) the rules. Rules that sur-vive a slow process of formation are thus morelikely to enjoy organized support and othermeans of institutional reproduction. By con-trast, where rules are designed quickly, actorsare more likely to miscalculate their potentialconsequences and/or how their interests are af-fected, and mechanisms of reproduction haveless time to emerge (Grzymala-Busse 2006).

Another factor underlying institutional sta-bility is the passage of time (Hacker 2004,Pierson 2004). When institutional arrange-ments persist (and are enforced) over time,surviving repeated crises and changes ofgovernment, actors develop expectations ofstability and consequently invest in skills, tech-nologies, and organizations that are appro-priate to those institutions (North 1990a,b;Pierson 2000a) . As these investments accumu-late, existing arrangements grow increasinglyattractive relative to their alternatives, therebyraising the cost of institutional replacement.

Such virtuous cycles are not inevitable, how-ever. Where formal institutions are repeatedlyoverturned or rendered ineffective, actors maydevelop expectations of instability (Przeworski1991, p. 82; Grzymala-Busse 2006). Conse-quently, they will be less likely to invest in thoseinstitutions or develop skills and technologiesappropriate to them, thereby keeping the costof overturning the rules low. Patterns of in-stitutional weakness may be reinforced by ac-tors’ investment in the skills and technologiesappropriate to an unstable institutional envi-ronment or to alternative informal rules of thegame. In a context of persistent regime instabil-ity, for example, actors may opt not to invest inparty building or in legislative skills—which areonly useful under democratic institutions—andinstead to develop skills (e.g., insurrectionary

capacity), resources (e.g., clandestine networksand other nonparty organizations), and rela-tionships (e.g., with religious authorities, themilitary, or foreign powers) that enhance theircapacity to operate in multiple regime settings.These investments may give actors a stake innoninstitutional politics, thereby reinforcingexisting patterns of institutional weakness. Fi-nally, as Helmke argues in her analysis of insti-tutional crises in Latin America, repeated boutsof instability may undermine the legitimacy of(or public support for) particular institutions.Their decreased legitimacy lowers the cost offuture attacks on them, thereby “paving the wayfor more crises” (Helmke 2007, p. 28).

Institutional instability may thus followa path-dependent logic (see Grzymala-Busse2006). An initial period of institutional failureor instability, which may be a product of histori-cally contingent circumstances (including sheerbad luck), may effectively lock a polity into apath of institutional weakness, or what Helmke(2007, p. 28) calls an “institutional instabilitytrap.” In such a context, actors develop expecta-tions that institutions will not endure and, con-sequently, do not invest in them. Institutionsmay also fail to generate public support or le-gitimacy. As a result, the cost of institutional re-placement will remain low, which increases thelikelihood of further rounds of change—and re-inforces expectations of institutional weakness.

Argentina followed a path of institutional in-stability after 1930 (Levitsky & Murillo 2005).Decades of regime instability, in which bothrules and rule-making processes were repeat-edly overturned, had a powerful effect on actorexpectations (Spiller & Tommasi 2007). Owingto the frequent collapse, suspension, or purgeof institutions as important as the constitu-tion, Congress, and Supreme Court, Argen-tines came to expect instability. Betting thatinstitutions would be replaced or purged witheach change of government or regime, polit-ical and economic actors did not make long-term investments in them. Few developed astake in any particular institutional arrange-ment. In fact, many actors—including Pero-nist unions, economic elites, and conservative

www.annualreviews.org • Variation in Institutional Strength 123

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

politicians—invested little in electoral, legisla-tive, and other democratic institutions andinstead developed skills, organizations, and re-lationships (particularly with the military) thathelped them survive in a context of regime in-stability (Cavarozzi 1987, McGuire 1997). As aresult, the cost of institutional change remainedlow.

Path-dependent analyses tend to focus oninstitutions that endure rather than those thatfail. However, the tools of comparative histor-ical analysis may also be employed to analyzethe roots of institutional weakness. Such anal-yses may help identify critical junctures thattrigger patterns of institutional instability, aswell as the mechanisms—such as changes in ac-tor expectations—by which those patterns arereproduced.

In summary, although weak institutions ex-ist everywhere, there are several reasons whythey are particularly widespread in the develop-ing world. One is countries’ position in the in-ternational system. Lower- and middle-incomecountries are more likely to engage in institu-tional borrowing from abroad (Weyland 2004).Borrowing may give rise to window-dressinginstitutions that governments do not intend toenforce (van de Walle 2001). Second, politiesin the developing world are more frequentlycharacterized by a disjuncture between rulewriters and actual power holders. Most de-veloping countries continue to be governedby dictatorships, hybrid regimes, or weaklyinstitutionalized democracies characterized byextensive informality and abuse of formal pro-cedure (O’Donnell 1994, 1996; Levitsky & Way2002). In few of these regimes is binding powerconsistently vested in the hands of rule-writingauthorities. Third, developing countries are of-ten characterized by states with limited bureau-cratic capacity and territorial reach (Evans &Rauch 1999). Finally, many developing coun-tries are characterized by high levels of socialand economic inequality. Tensions generatedby the coexistence of political equality and vastsocioeconomic inequality are likely to be re-solved via uneven enforcement or continuousinstitutional re-engineering when existing rules

generate outcomes that powerful players deemundesirable.5

IMPLICATIONS OFINSTITUTIONAL WEAKNESS

Variation in institutional strength has impor-tant implications for our understanding ofhow institutions shape—and are shaped by—politics. Different levels of enforcement andstability have important effects on actors’ ex-pectations and behavior. When institutions arestable and enforced, actors can assume that(a) others will play by the rules and (b) therules will endure in the foreseeable future. Be-cause the menu of behavioral options is lim-ited to those prescribed or permitted by theformal rules, actors’ choices are constrained,which reduces uncertainty. And because actorsbelieve that others are similarly constrained bythe rules, and that the rules will endure, theydevelop stable expectations about others’ be-havior, which lengthens time horizons and fa-cilitates cooperation (O’Donnell 1994). In sucha context, actors tend to take institutions se-riously. Able to define their own institutionalpreferences (and calculate those of others) withrelative ease, they invest time and energy inquestions of institutional design, and once in-stitutions are in place, they invest in the skillsand organizations necessary to achieve theirgoals through those institutions (North 1990b,pp. 364–65, Pierson 2000a).

Institutional weakness creates a different setof expectations. Where institutions are weaklyenforced, for example, actors cannot assumethat others will comply with the rules. In sucha context, actors enjoy greater discretion: Theymay select strategies that are prescribed by therules, but they may also choose among vari-ous extrainstitutional options. As the numberof strategic options increases, so does uncer-tainty. Actors are less able to form stable expec-tations about others’ behavior, which narrowstheir time horizons and erodes trust. In this way,

5On the relationship between socioeconomic inequality andpolitical instability, see Alesina & Perotti (1996).

124 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

low enforcement limits cooperation and collec-tive action, which can undermine the quality—and in some cases, the stability—of new democ-racies (Nino 1992; O’Donnell 1993, 1994; Steinet al. 2006).

The implications of institutional instabil-ity have only recently attracted scholarly at-tention (Stein et al. 2006, Spiller & Tommasi2007). Although it may be argued that frequentchanges in the rules matter little, as long asthe rules in place at a given time continue toguide behavior, Spiller & Tommasi (2007) makea compelling case to the contrary. They arguethat repeated change undermines institutionalcredibility and narrows actors’ time horizons,which limits their capacity to sustain intertem-poral agreements. The result is a volatile policy-making process and, in most cases, low-qualitypolicies (Spiller & Tommasi 2007; also Steinet al. 2006).

In recent years, scholars have begun to re-fine existing institutionalist theories in light ofobserved variation in enforcement and stability.This section examines some of these efforts inthree areas of comparative research: (a) analy-ses of institutional effects, (b) theories of insti-tutional design, and (c) theories of institutionaldevelopment and change.

Assessing Institutional Effects

Much of the comparative literature on politi-cal institutions focuses on how variation in in-stitutional design shapes political and/or eco-nomic outcomes. During the initial wave ofresearch on comparative institutional designin the 1980s and 1990s, studies frequently as-sumed a tight coupling between formal rulesand actor behavior. Where these assumptionsheld, comparative institutional analyses oftengenerated fruitful results. This was clearly thecase, for example, in the comparative literatureon electoral rules (Taagepera & Shugart 1989,Carey & Shugart 1995, Cox 1997). To the ex-tent that parchment rules are not systematicallyobeyed, however, the assumption that formalconstraints can be mapped directly onto polit-ical behavior—and consequently, the assump-

tion that institutional design has a significanteffect on political outcomes—becomes moreproblematic.

An example is the comparative study ofpresidentialism. In evaluating how presidentialsystems work, scholars have focused on bothformal rules (Shugart & Carey 1992, Carey &Shugart 1998) and informal factors such as par-tisan strength and the character of party systems(Mainwaring 1993, Mainwaring & Shugart1997). Yet recent studies show that variation inenforcement also has a significant impact onhow presidential regimes function. Take thecases of Chile and Mexico during the 1990s.From a constitutional standpoint, the Chileanpresidency was among the most powerful inthe world, whereas the Mexican presidencywas relatively weak (Shugart & Carey 1992).In practice, however, Mexican presidentsenjoyed “an extraordinary range of powers”that extended well beyond their constitutionalprerogatives (Weldon 1997, p. 225) while thelegislature was little more than a “rubber stampfor presidential decrees” (Eisenstadt 2004,p. 40). In Chile, by contrast, presidents of thegoverning Concertacion systematically underuti-lized their vast legislative and agenda-settingpowers, preferring to negotiate legislationwith coalition partners, opposition parties, andleading interest groups (Siavelis 1997). Hence,presidential dominance in Mexico under thePRI and effective horizontal accountability inpost-Pinochet Chile cannot be easily tracedto constitutional design. Indeed, constitutionaldesign leads one to predict the opposite.

Similar issues arise in analyses of judicial in-dependence. An established body of literaturehas linked judicial independence to the rulesgoverning the appointment and removal of jus-tices (see Dahl 1957, Verner 1989). Yet recentresearch has shown that enforcement of theserules is a critical factor in determining judicialbehavior. Although many Latin American con-stitutions provide formal guarantees of tenuresecurity—including lifetime appointment—forSupreme Court justices, variation in compli-ance with these guarantees has produced verydifferent patterns of judicial behavior (Bill

www.annualreviews.org • Variation in Institutional Strength 125

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Chavez 2004, Helmke 2004). In postrevolu-tionary Mexico, for example, Supreme Courtappointments rarely endured beyond a singlepresidential term (Domingo 2000, pp. 722–25),and as a result, the Court was characterized byits “subservience to the will of the executive”(Domingo 2000, p. 706). In Argentina, repeatedcourt packing by incoming governments cre-ated a logic of “strategic defection,” in whichSupreme Court justices ruled in favor of thegovernment as long as it was solidly entrenchedin power, but then abandoned it in favor of alikely successor when it was expected to leavepower (Helmke 2004).

Taking enforcement into account also helpsclarify the relationship between judicial inde-pendence and economic growth. Analyses thatfocus on the formal rules underlying judicialindependence—without taking levels of en-forcement into account—have found little re-lationship between judicial independence andeconomic growth (Glaeser et al. 2004). How-ever, studies that measure de facto, or enforced,judicial independence find a positive correla-tion between judicial independence and GDPgrowth (Feld & Voigt 2003).

Enforcement issues are also critical to com-parative research on central banks. Compara-tive political economy analyses of central bankindependence in advanced democracies gener-ally assume that rules regulating bank inde-pendence are enforced and thereby produceexpectations of low inflation (Franzese & Hall1998, Iversen 1998, Franzese 2002). As in thecase of judicial independence, however, stud-ies of central banks in developing countrieshave found a weak relationship between dejure central bank independence and economicoutcomes (Cukierman et al. 1992). Indeed,Cukierman et al. (1992) show that whereas dejure central bank independence is a good pre-dictor of inflation rates in advanced industrial-ized countries, the level of turnover of centralbank presidents—a measure of degree of en-forcement of central bank independence—is abetter predictor of inflation rates in developingcountries.

In summary, the utility of comparative anal-yses of parchment rules hinges on how wellthose rules are enforced. Hence, comparativeresearch must take into account not only whatthe rules say, but also the extent to which ac-tors comply with them in practice. Rules ofsimilar design but distinct levels of enforce-ment may generate dramatically different ex-pectations, behavior, and outcomes.

Understanding Institutional Design

Variation in enforcement and stability is alsorelevant to research on institutional design.Theories of institutional design often assumethat actors expect the rules they create to en-dure and to constrain future behavior (for a cri-tique, see Pierson 2000a). For example, scholarshave shown that politicians facing a high prob-ability of losing power may create self-bindinginstitutions in order to constrain their rivals inthe future (Moe 1990, Geddes 1994, Boix 1999,Boylan 2001). Where institutions are stable andenforced, such assumptions of far-sighted ex-pectations have proven theoretically fruitful.For example, in her study of the origins ofcentral bank autonomy in Chile and Mexico,Boylan (2001) shows how looming democratictransitions created incentives for autocrats toinsulate monetary policy making from theirdemocratic successors via the creation of inde-pendent central banks.

Yet recent studies have shown that actors’preferences regarding institutional design of-ten hinge on their expectations regarding stabil-ity and enforcement. For example, in her workon redistributive social insurance, Mares (2005,pp. 626–27) argues that workers’ support forthe creation of redistributive social policy insti-tutions hinges not only on calculations of howsuch policies would affect their material inter-ests but also on their expectations about the en-forcement of those policies. Thus, workers inhigh-risk sectors, who would ordinarily supportinstitutions of social protection, may not do soin a context of weak enforcement (Mares 2005,p. 630). In effect, workers’ preferences for risk

126 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

redistribution change in line with expectationsof enforcement. Indeed, if actors do not expectrules to be enforced, they may support formalinstitutional arrangements that—on paper—seem to run counter to their interests.

Expectations about enforcement also haveimplications for research on policy diffusion.Much of the early literature on policy dif-fusion implicitly assumes that borrowed in-stitutions will be enforced, such that similarrules produce similar behavior across cases. [Onthe spread of market institutions in postcom-munist states, see Aslund (1995). On centralbank independence in developing countries, seeMaxfield (1997).] As recent research has shown,however, policy makers in developing coun-tries borrow institutions for diverse reasons (seeWeyland 2008), and in some cases, their pref-erences regarding borrowed institutions hingeon whether or not they expect those institu-tions to be enforced. Because international fi-nancial institutions—often lacking the capac-ity to monitor enforcement—tend to base theirevaluations on parchment institutions ratherthan actual implementation (van de Walle 2001,p. 232), governments may adopt formal rulesas window dressing. Such institutions can beexpected to function differently. Indeed, inher analysis of state institution building inCentral Europe, Grzymala-Busse (2007) findsthat institutions imposed by the EuropeanUnion were more weakly enforced than thosewhose origins lay in pressure by domesticopposition.

Variation in stability also shapes institutionaldesign. For example, where politicians do notexpect formal rules to endure, they may be lesslikely to build self-binding institutions aimedat constraining their successors. Although thePinochet government took institutional designvery seriously, crafting a range of institutions(e.g., the 1980 constitution, the electoral sys-tem, an independent central bank) to limitthe power and policy-making discretion of itsdemocratic successors (Boylan 2001), Chile is,by Latin American standards, a case of unusualinstitutional strength. In countries with weaker

institutions, military dictatorships tended to in-vest less in constitutional design. For example,Bolivian dictator Hugo Banzer (1971–1978) en-gaged in virtually no institution building, de-spite being the country’s longest-serving presi-dent in more than a century (Whitehead 1986,pp. 54–58).

In summary, actors’ expectations about en-forcement and stability shape how they ap-proach institutional design. Most theories ofinstitutional design emerged from studies ofcountries with strong institutions; scholars arejust beginning to understand how and why ac-tors create institutions in contexts in whichthe rules of the game are unstable or weaklyenforced.

Analyzing Institutional Developmentand Change

Variation in institutional strength also has im-portant implications for research on institu-tional development and change. The historicalinstitutionalist literature on institutional devel-opment in the advanced industrialized coun-tries focuses more on continuity than change(Thelen 2004, Streeck & Thelen 2005). AsThelen (1999) notes, much of the histori-cal institutionalist literature views institutionalchange—explicitly or implicitly—through thelens of a punctuated equilibrium model, inwhich institutional development is character-ized by “long continuities,” periodically inter-rupted by “radical shifts” (Pempel 1998, p. 1;also Krasner 1988). Recent work by Streeck &Thelen (2005, pp. 18–30) points to a varietyother ways in which institutions may evolve,including changes in function (“conversion”),the grafting of new institutions onto old ones(“layering”), and gradual atrophy (“drift”) orloss of relevance (“displacement”). However,nearly all of the patterns of change they iden-tify are gradual and incremental. In a con-text of generalized institutional strength, suchcharacterizations make sense. By narrowing ac-tors’ choice sets, stabilizing expectations, rou-tinizing behavior, and generating asset-specific

www.annualreviews.org • Variation in Institutional Strength 127

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

investment linked to existing institutional ar-rangements, strong institutions take radical al-ternatives off the table and virtually ensurethat—outside of critical junctures—change isincremental.

In a context of institutional weakness, how-ever, patterns of change may be very different.Weak institutions are more open to repeatedand radical change. Where the rules are notsystematically enforced, actors’ choices are notconfined to a set of stable and narrowly definedoptions; actors have a wider menu of strate-gic options available to them. This includesextraprocedural change, or change that occursoutside existing legal-institutional frameworksand/or in violation of officially sanctioned pro-cedures. Moreover, in a context of low institu-tional durability, the cost of wholesale replace-ment may be relatively low, as vested interestsand other mechanisms of reproduction have lesstime to develop (Hacker 2004, Grzymala-Busse2006). Consequently, institutional change ismore likely to take the form of “breakdownand replacement.” For example, whereas highlyinstitutionalized European social democraticparties adapted slowly and incrementally tosocioeconomic change during the 1980s and1990s (Kitschelt 1994), Levitsky (2003) showshow Argentine Peronism, whose internal ruleswere unstable and weakly enforced, underwentrapid and far-reaching programmatic changeduring the 1990s. In less than a decade, Per-onism transformed itself from a labor-basedparty into a patronage-based machine capableof undertaking radical market-oriented reform.(The party underwent another transformationfollowing the 2001–2002 economic collapse,embracing a more statist project.)

Finally, attention to issues of institutionalstrength allows us to see an additional patternof institutional evolution: changing levels ofenforcement. Increased or decreased enforce-ment can produce substantial change in institu-tional effects—altering actors’ expectations andbehavior—without modification to parchmentrules. Mexican democratization provides anexample of institutional change via increased

enforcement. The transformation from a hege-monic regime into a democracy involved rela-tively little change in the form of Mexico’s con-stitution but dramatic change in enforcementin key areas. Postsocialist China and Vietnamprovide examples of change via decreased en-forcement. As noted above, initial steps towardeconomic liberalization in China and Vietnamtook the form of relaxed enforcement, ratherthan dismantling, of state socialist institutions.Socialist institutions remained on the books,but a decline in enforcement allowed both stateofficials and emerging entrepreneurs to oper-ate according to informal market institutions(Malesky 2005, Tsai 2007). Thus, actors’ expec-tations and behavior changed dramatically.

In summary, relaxing assumptions of formalinstitutional strength can generate new insightsinto why politicians create institutions, how in-stitutions shape political outcomes, and howpolitical institutions evolve or change. Emerg-ing research in these areas is critically importantif the new institutionalism is to travel to the de-veloping world.

CONCLUSION: ISSUES FORFUTURE RESEARCH

Variation in the enforcement and stability offormal rules matters. If the new institution-alism in political science is to generate theo-ries that are truly comprehensive, in that theyare applicable not only to advanced capitalistcountries but also to developing ones, then thestrength of formal institutions cannot be takenfor granted. Rather, enforcement and stabilityshould be treated as dimensions along whichparticular institutions vary. As we have shown,recent studies have begun to explore this varia-tion, and their findings allow scholars to refineexisting theories of institutional design, institu-tional effects, and institutional change.

A crucial remaining task is to develop mea-sures of enforcement and stability that cantravel across diverse cases. With respect toenforcement, an obvious first step is to sep-arate measurement of the formal rules from

128 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

measurement of actual enforcement. In someareas, unsanctioned violations are either highlyvisible (e.g., Supreme Court packing) or can bereliably counted (e.g., nonpayment of taxes). Insuch cases, observed violations may be a validand useful measure. In other areas, however,rule infractions are not as visible or as easilycounted (e.g., campaign finance laws, labor reg-ulation). In these cases, the use of observed vio-lations as a measure of enforcement may be bi-ased by the level of monitoring. In other words,violations may be more easily observable in acontext of high enforcement, because state of-ficials, the media, and other public “fire alarms”are doing their job more effectively. Hence,measures of violation must control for the levelof monitoring. This can be done by assessingthe resources, quality, and incentives of moni-toring agents [see, e.g., Schrank’s (2006) workon labor inspectors and monitoring in the Do-minican Republic].

Where direct observation is difficult, surveysmay provide an alternative means of measuringenforcement. Several major cross-national sur-veys (e.g., Transparency International, Latino-barometer, Afrobarometer) include questionsabout legal and regulatory enforcement. How-ever, these surveys’ reliance on respondents’perceptions, rather than actual experience, lim-its their validity. Studies that directly surveyindividuals about their own compliance, suchas Gingerich’s (2007) recent work on politi-cal corruption across state agencies in Bolivia,Brazil, and Chile, are promising. Such studiesrequire techniques to minimize bias rooted inactors’ reluctance to respond truthfully aboutimproper or illicit behavior, and where theyinvolve cross-national surveys, they must beable to correct for culturally rooted differ-ences in perceptions of (and expectations about)enforcement.6

6See, for example, King et al.’s (2003) proposed use of “an-choring vignettes” to help correct for cultural differences incross-national survey research or Calvo & Murillo’s (2008)work on clientelistic networks, which treats survey respon-dents as observers of behavior.

Instability is easier to observe than enforce-ment, but it can be tricky to interpret. For exam-ple, the mere passage of time fails to capture theextent to which an institution has actually beentested, or whether actors have come to expect itto survive. A better way of measuring stabilityis to examine whether an institution has sur-vived changes in the distribution of power andpreferences among the actors who designed it.For example, changes in government or regimemay be a useful indicator of institutional stabil-ity. The fact that Pinochet’s market institutionssurvived Chile’s democratic transition and thesubsequent election of center-left governmentswas a clear indicator of their durability—better,perhaps, than the number of years in which theysurvived under Pinochet. Survival through cri-sis may also be a useful indicator of durability.For example, the survival of Argentina’s 1994constitution through the 2001–2002 economiccollapse is an indicator of durability. An insti-tution’s survival through a period of crisis maybe considered an indicator that relevant actorshave come to expect the institution to endure.

Good measures of enforcement andstability—and particularly measures that cantravel across cases—are critical to our ability toexplain variation in institutional strength acrosscases and over time. They are also critical toour ability to study the effects of institutionalweakness. Developing such measures maybe the biggest challenge facing comparativeresearch in this area.

We conclude with a plea for further re-search on institutional enforcement and sta-bility. Taking institutional strength seriously—treating it as a variable, rather than a taken-for-granted assumption—will help us to buildtheories of institutions that apply to both theindustrialized and developing worlds. Increas-ing our stock of knowledge about the causes andconsequences of institutional weakness will notonly improve our understanding of institutionsand their effects in theory, but also may ulti-mately contribute to effective institution build-ing in practice—a process of great importancein many new democracies.

www.annualreviews.org • Variation in Institutional Strength 129

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

We are thankful for comments and suggestions from Carlos Acuna, David Collier, Javier Corrales,Jorge Domınguez, Anna Grzymala-Busse, Peter Hall, Gretchen Helmke, Pauline Jones-Luong,Ira Katznelson, Jorge Leon, Margaret Levi, Scott Mainwaring, Alison Post, James Robinson,Catalina Smulovitz, Hillel Soifer, Kathleen Thelen, Kurt Weyland, and an anonymous reviewer,as well as participants in the conference “Argentine Democracy: The Politics of InstitutionalWeakness,” Universidad Torcuato Di Tella, June 29, 2006.

LITERATURE CITED

Alesina A, Perotti R. 1996. Income distribution, political instability and investment. Eur. Econ. Rev. 40:1202–29Aslund A. 1995. The case for radical reform. In Economic Reform and Democracy, ed. L Diamond, MF Plattner,

pp. 74–85. Baltimore: Johns Hopkins Univ. PressBenoit K, Hayden J. 2004. Institutional change and persistence: the evolution of Poland’s electoral system.

1989–2001. J. Polit. 66(2):396–427Bill Chavez R. 2004. The Rule of Law in Nascent Democracies: Judicial Politics in Argentina. Stanford, CA: Stanford

Univ. PressBoix C. 1999. Setting the rules of the game: the choice of electoral systems in advanced democracies.

Am. Polit. Sci. Rev. 93(3):609–24Boylan D. 2001. Defusing Democracy: Central Bank Autonomy and the Transition from Authoritarian Rule. Ann

Arbor: Univ. Mich. PressCalvo E, Murillo MV. 2008. When parties meet voters: partisan networks and distributive expectations in Argentina

and Chile. Presented at Annu. Meet. Am. Polit. Sci. Assoc., BostonCarey JM. 2000. Parchment, equilibria, and institutions. Comp. Polit. Stud. 33(6/7):735–61Carey JM, Shugart MS. 1995. Incentives to cultivate a personal vote: a rank ordering of electoral formulas.

Elect. Stud. 14:417–39Carey JM, Shugart MS, eds. 1998. Executive Decree Authority. New York: Cambridge Univ. PressCavarozzi, Marcelo. 1987. Autoritarismo y democracia. Buenos Aires: Centro Editor de America Latina.Clemens E, Cook J. 1999. Politics and institutionalism: explaining durability and change. Annu. Rev. Sociol.

25:441–66Cox GW. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. New York:

Cambridge Univ. PressCukierman A, Webb S, Neyapti B. 1992. Measuring the independence of central banks and its effects on policy

outcomes. World Bank Econ. Rev. 6:353–98Dahl R. 1957. Decision-making in a democracy: the Supreme Court as a national policymaker. J. Public Law

6:279–95Dimaggio PJ, Powell WW, eds. 1991. The New Institutionalism in Organizational Analysis. Chicago: Univ.

Chicago PressDomingo P. 2000. Judicial independence: the politics of the Supreme Court in Mexico. J. Latin Am. Stud.

32:705–35Eisenstadt TA. 2004. Courting Democracy in Mexico: Party Strategies and Electoral Institutions. New York:

Cambridge Univ. PressEvans P, Rauch R. 1999. Bureaucracy and growth: a cross-national analysis of the effect of “Weberian” state

structures on economic growth. Am. Sociol. Rev. 64:748–65Feld L, Voigt S. 2003. Economic growth and judicial independence: cross-country evidence using a new set

of indicators. Eur. J. Polit. Econ. 19:497–527

130 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Fitch JS. 1998. The Armed Forces and Democracy in Latin America. Baltimore: Johns Hopkins Univ. PressFranzese RJ. 2002. Macroeconomic Policies of Developed Democracies. Cambridge Univ. PressFranzese R, Hall P. 1998. Mixed signals: central bank independence, coordinated wage bargaining, and

European monetary union. Int. Org. 52(3):505–36Geddes B. 1994. Politicians’ Dilemma. Berkeley: Univ. Calif. PressGingerich D. 2007. Corruption in general equilibrium: political institutions and bureaucratic performance in South

America. Ph.D. thesis, Dep. Gov., Harvard Univ.Glaeser E, LaPorta R, Lopez de Silanes F, Shleifer A. 2004. Do institutions cause growth? J. Econ. Growth

9:271–303Gonzales FE, King D. 2004. The state and democratization: the United States in comparative perspective.

Br. J. Polit. Sci. 34:193–210Greif A, Laitin DD. 2004. A theory of endogenous institutional change. Am. Polit. Sci. Rev. 98:613–32Grzymala-Busse A. 2006. Disaggregating temporal effects. Presented at Annu. Meet. Am. Polit. Sci. Assoc.,

PhiladelphiaGrzymala-Busse A. 2007. Rebuilding Leviathan: Party Competition and State Exploitation in Post-Communist

Democracies. New York: Cambridge Univ. PressHacker J. 2004. Privatizing risk without privatizing the welfare state. Am. Pol. Sci. Rev. 98(2):243–60Helmke G. 2004. Courts under Constraints: Judges, Generals, and Presidents in Argentina. New York: Cambridge

Univ. PressHelmke G. 2007. The origins of institutional crises in Latin America: a unified strategic model and test. Presented at

Annu. Meet. Midwest Polit. Sci. Assoc., ChicagoHelmke G, Levitsky S, eds. 2006. Informal Institutions and Democracy: Lessons from Latin America. Baltimore:

Johns Hopkins Univ. PressHerbst J. 2000. States and Power in Africa: Comparative Lessons in Authority and Control. Princeton, NJ: Princeton

Univ. PressHtun M. 2003. Sex and the State. New York: Cambridge Univ. PressHuntington S. 1968. Political Order in Changing Societies. New Haven, CT: Yale Univ. PressIversen T. 1998. Wage bargaining, hard money and economic performance. Br. J. Polit. Sci. 28:31–61Joseph R. 1999. Africa, 1990–1997: from Abertura to closure. In Democratization in Africa, ed. L Diamond,

MF Plattner, pp. 3–18. Baltimore: Johns Hopkins Univ. PressKing G, Murray CJL, Salomon JA, Tandon A. 2003. Enhancing the validity and cross-cultural comparability

of survey research. Am. Polit. Sci. Rev. 97(4):567–83Kitschelt H. 1994. The Transformation of European Social Democracy. Cambridge, UK: Cambridge Univ. PressKlug H. 2000. Constituting Democracy: Law, Globalism, and South Africa’s Political Reconstruction. Cambridge,

UK: Cambridge Univ. PressKnight J. 1992. Institutions and Social Conflict. New York: Cambridge Univ. PressKrasner SD. 1988. Sovereignty: an institutional perspective. Comp. Polit. Stud. 21(1):66–94Langston J. 2006. The birth and transformation of the dedazo in Mexico. In Informal Institutions and Democracy:

Lessons from Latin America, ed. G Helmke, S Levitsky, pp. 143–59. Baltimore: Johns Hopkins Univ. PressLevi M. 1988. Of Rule and Revenue. Berkeley: Univ. Calif. PressLevi M. 1997. Consent, Dissent, and Patriotism. New York: Cambridge Univ. PressLevitsky S. 2003. Transforming Labor-Based Parties in Latin America: Argentine Peronism in Comparative Perspec-

tive. New York: Cambridge Univ. PressLevitsky S, Murillo MV, eds. 2005. Argentine Democracy: The Politics of Institutional Weakness. University Park:

Penn. State Univ. PressLevitsky S, Way LA. 2002. The rise of competitive authoritarianism. J. Democracy 13(2):51–65Lieberman E. 2003. Race and Regionalism in the Politics of Taxation in Brazil and South Africa. Cambridge, UK:

Cambridge Univ. PressLoveman, B. 1993. The Constitution of Tyranny: Regimes of Exception in Spanish America. Pittsburgh, PA: Univ.

Pittsburgh PressMainwaring S. 1993. Presidentialism, multipartism, and democracy: the difficult combination. Comp. Polit.

Stud. 26(2):198–228

www.annualreviews.org • Variation in Institutional Strength 131

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Mainwaring S, Shugart MS, eds. 1997. Presidentialism and Democracy in Latin America. New York: CambridgeUniv. Press

Malesky E. 2005. Straight ahead on red: the impact of foreign direct investment on local autonomy in Vietnam.Presented at Annu. Meet. Midwest Polit. Sci. Assoc., Chicago

March JG, Olsen JP. 1989. Rediscovering Institutions: The Organizational Basis of Politics. New York: Free PressMarcus RR. 2005. The fate of Madagascar’s democracy: following the rules while eroding the substance. In

The Fate of Africa’s Democratic Experiments: Elites and Institutions, ed. LA Villalon, P VonDoepp, pp. 153–74Bloomington: Indiana Univ. Press

Mares I. 2005. Social protection around the world: external insecurity, state capacity, and domestic politicalcleavages. Comp. Polit. Stud. 38(6):623–51

Maxfield S. 1997. Gatekeepers of Growth: The International Political Economy of Central Banking in DevelopingCountries. Princeton, NJ: Princeton Univ. Press

McGuire J. 1997. Peronism without Peron: Unions, Parties, and Democracy in Argentina. Stanford, CA: StanfordUniv. Press

Messone N, Gros JG. 1998. The irony of wealth: democratization in Gabon. In Democratization in LateTwentieth-Century Africa: Coping with Uncertainty, ed. JG Gros, pp. 129–47. Westport, CT: Greenwood

Meyer J, Rowan B. 1991. Institutionalized organizations: formal structure as myth and ceremony. In TheNew Institutionalism in Organizational Analysis, ed. WW Powell, PJ DiMaggio, pp. 41–62. Chicago: Univ.Chicago Press

Moe T. 1990. The politics of structural choice: toward a theory of public bureaucracy. In Organization Theory:From Chester Barnard to Present and Beyond, ed. O Williamson, pp. 116–53. New York: Oxford Univ. Press

Nino C. 1992. Un paıs al margen de la ley. Buenos Aires: EmeceNorth DC. 1990a. Institutions, Institutional Change, and Economic Performance. New York: Cambridge Univ.

PressNorth DC. 1990b. A transaction cost theory of politics. J. Theor. Polit. 2(4):355–67O’Donnell G. 1993. On the state, democratization, and some conceptual problems: a Latin American view

with some postcommunist countries. World Dev. 21(8):1355–69O’Donnell G. 1994. Delegative democracy. J. Democracy 5(1):55–69O’Donnell G. 1996. Illusions about consolidation. J. Democracy 7(2):34–51O’Donnell G. 1999. Polyarchies and the (un)rule of law in Latin America: a partial conclusion. In The (Un)rule

of Law and the Underprivileged in Latin America, ed. J Mendez, G O’Donnell, PS Pinheiro, pp. 303–337.Notre Dame, IN: Univ. Notre Dame Press

Pempel TJ. 1998. Regime Shift: Comparative Dynamics of the Japanese Political Economy. Ithaca, NY: CornellUniv. Press

Pierson P. 2000a. The limits of institutional design: explaining institutional origins and change. Governance13(4):475–99

Pierson P. 2000b. Increasing returns, path dependence, and the study of politics. Am. Polit. Sci. Rev. 94(2):251–67

Pierson P. 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Princeton Univ. PressPieterse M. 2004. Coming to terms with judicial enforcement of socio-economic rights. S. Afr. J. Hum. Rights

20(3):383–417Przeworski A. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America.

New York: Cambridge Univ. PressRose LF, Evans DH. 1997. Pakistan’s enduring experiment. J. Democracy 8(1):83–96Schrank A. 2006. Professionalization and probity in the patrimonial state: labor law enforcement in the Dominican

Republic. Presented at Inst. Work and Employment Res., Mass. Inst. Technol., Cambridge, MASherwani LA. 1962. The constitutional experiment in Pakistan. Asian Surv. 2(6):9–14Shugart MS, Carey JM. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge,

UK: Cambridge Univ. PressSiavelis P. 1997. Executive-legislative relations in post-Pinochet Chile: a preliminary assessment. See

Mainwaring & Shugart 1997, pp. 321–62Soifer H. 2006. Authority over distance: explaining variation in state infrastructural power in Latin America. Ph.D.

thesis, Dep. Gov., Harvard Univ.

132 Levitsky · Murillo

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

ANRV377-PL12-07 ARI 7 April 2009 16:12

Spiller P, Tommasi M. 2007. The Institutional Foundations of Public Policy: A Transactions Theory and an Applicationto Argentina. Cambridge, UK: Cambridge Univ. Press

Stein E, Tommasi M, Echebarrıa K, Lora E, Payne M. 2006. The politics of policies: economic and social progressin Latin America, 2006 report. Cambridge, UK: Inter-Am. Dev. Bank and David Rockefeller Cent. LatinAm. Stud./Harvard Univ. Press

Streeck W, Thelen K. 2005. Introduction: institutional change in advanced political economies. In BeyondContinuity: Institutional Change in Advanced Political Economies, ed. W Streeck, K Thelen, pp. 1–39.New York: Oxford Univ. Press

Taagepera R, Shugart MS. 1989. Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven,CT: Yale Univ. Press

Thelen K. 1999. Historical institutionalism in comparative politics. Annu. Rev. Polit. Sci. 2:369–404Thelen K. 2004. How Institutions Evolve: The Political Economy of Skills in Germany, Britain, the United States,

and Japan. New York: Cambridge Univ PressTorres-Rivas E. 1996. Guatemala: democratic governability. In Constructing Democratic Governance: Mexico,

Central America, and the Caribbean in the 1990s, ed. JI Dominguez, AF Lowenthal, pp. 50–63. Baltimore:Johns Hopkins Univ. Press

Tsai KS. 2007. Adaptive informal institutions and endogenous institutional change in China. World Polit.59(1):116–41

van de Walle N. 2001. African Economies and the Politics of Permanent Crisis: 1979–1999. Cambridge, UK:Cambridge Univ. Press

Van Oenen G. 2001. Citizenship and “informal rule of law.” How lax enforcement can sustain the rule of law.Presented at Annu. Meet. Latin Am. Stud. Assoc., Washington, DC

Verner JG. 1989. The independence of supreme courts in Latin America: a review of the literature. J. LatinAm. Stud. 16:463–506

Weldon J. 1997. Political sources of presidencialismo in Mexico. See Mainwaring & Shugart 1997, pp. 225–58Weyland K. 2004. Learning from foreign models in Latin American policy reform: an introduction. In Learning

From Foreign Models in Latin American Policy Reform, ed. K Weyland, pp. 1–34. Baltimore: Johns HopkinsUniv. Press

Weyland K. 2008. Toward a new theory of institutional change. World Polit. 60:281–314Whitehead L. 1986. Bolivia’s failed democratization, 1977–1980. In Transitions from Authoritarian Rule: Latin

America, ed. G O’Donnell, PC Schmitter, L Whitehead, pp. 49-71. Baltimore: Johns Hopkins Univ. PressYashar D. 2005. Contesting Citizenship in Latin America. New York: Cambridge Univ. Press

www.annualreviews.org • Variation in Institutional Strength 133

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

AR377-FM ARI 7 April 2009 11:25

Annual Review ofPolitical Science

Volume 12, 2009Contents

A Conversation with Robert A. DahlRobert A. Dahl and Margaret Levi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Neorepublicanism: A Normative and Institutional Research ProgramFrank Lovett and Philip Pettit � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �11

Domestic Terrorism: The Hidden Side of Political ViolenceIgnacio Sánchez-Cuenca and Luis de la Calle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �31

Women in Parliaments: Descriptive and Substantive RepresentationLena Wängnerud � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �51

Self-Government in Our TimesAdam Przeworski � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �71

Social Policy in Developing CountriesIsabela Mares and Matthew E. Carnes � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �93

Variation in Institutional StrengthSteven Levitsky and María Victoria Murillo � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 115

Quality of Government: What You GetSören Holmberg, Bo Rothstein, and Naghmeh Nasiritousi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 135

Democratization and Economic GlobalizationHelen V. Milner and Bumba Mukherjee � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 163

Has the Study of Global Politics Found Religion?Daniel Philpott � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 183

Redistricting: Reading Between the LinesRaymond La Raja � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 203

Does Efficiency Shape the Territorial Structure of Government?Liesbet Hooghe and Gary Marks � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 225

Bargaining Failures and Civil WarBarbara F. Walter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 243

Hobbesian Hierarchy: The Political Economy of PoliticalOrganizationDavid A. Lake � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 263

v

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.

AR377-FM ARI 7 April 2009 11:25

Negative CampaigningRichard R. Lau and Ivy Brown Rovner � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 285

The Institutional Origins of Inequality in Sub-Saharan AfricaNicolas van de Walle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 307

RiotsSteven I. Wilkinson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 329

Regimes and the Rule of Law: Judicial Independence in ComparativePerspectiveGretchen Helmke and Frances Rosenbluth � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 345

Field Experiments and the Political Economy of DevelopmentMacartan Humphreys and Jeremy M. Weinstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 367

Laboratory Experiments in Political EconomyThomas R. Palfrey � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 379

Field Experiments on Political Behavior and Collective ActionEline A. de Rooij, Donald P. Green, and Alan S. Gerber � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 389

Experiments on Racial Priming in Political CampaignsVincent L. Hutchings and Ashley E. Jardina � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 397

Elections Under AuthoritarianismJennifer Gandhi and Ellen Lust-Okar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 403

On Assessing the Political Effects of Racial PrejudiceLeonie Huddy and Stanley Feldman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 423

A “Second Coming”? The Return of German Political TheoryDana Villa � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 449

Group Membership, Group Identity, and Group Consciousness:Measures of Racial Identity in American Politics?Paula D. McClain, Jessica D. Johnson Carew, Eugene Walton, Jr.,and Candis S. Watts � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 471

Opiates for the Matches: Matching Methods for Causal InferenceJasjeet Sekhon � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 487

Indexes

Cumulative Index of Contributing Authors, Volumes 8–12 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 509

Cumulative Index of Chapter Titles, Volumes 8–12 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 511

Errata

An online log of corrections to Annual Review of Political Science articles may be foundat http://polisci.annualreviews.org/

vi Contents

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:115

-133

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Har

vard

Uni

vers

ity o

n 07

/16/

12. F

or p

erso

nal u

se o

nly.