Abstract
Leading theories of transitions from single-party dominant systems begin with economic crisis, the party's loss of patronage resources, and elite-level defections. The multiparty elections that are then held exert no independent effect, but instead register neutrally the party's decline and the democratization of politics. This paper, however, shifts attention from the dominant party to citizens and elections in non-crisis conditions. It argues that on key dimensions citizens assess the dominant party's legitimacy or worthiness of support. Further, where they grow critical of its policy outputs, they scrutinize more closely its conformity to procedures. And as they anticipate that their voting preferences will be thwarted by electoral manipulations, they vote in protest, perhaps producing a “liberalizing electoral outcome.” Elections, then, do not simply indicate the dominant party's decline. By deepening alienation, they help citizens to cause it. Analysis is set in Malaysia, long an exemplar of single-party dominance, but recently a case in which the government was dealt a striking electoral setback.
After having studied the durability of authoritarian regimes, comparativists have begun returning to earlier questions over authoritarian breakdown and democratic transitions. But they have mostly addressed a particular kind of authoritarianism, one wherein a single dominant party holds multiparty elections, but then limits competitiveness. Labeled by some scholars as electoral authoritarianism (Schedler 2006), this marks the spot where much of the third wave has equilibrated.
But electoral authoritarianism can also break down, prompting comparativists also to study the tensions in single-party dominance. Hence, they have traced the decline of dominant parties through a weakened grip on public sector resources, a loss of state patronage, and elite-level defections, at last ceding space in which opposition parties might mobilize constituencies. In this research agenda, conducted from “inside-out,” elections are assumed to have been lost by dominant parties well before they have even been waged.
This paper, 1 however, in exploring recent politics in Malaysia – long an exemplar of electoral authoritarianism, yet perhaps poised today for democratic change – takes a different tack. It shifts attention from the dominant party, cascading defections, and vanishing resources and patronage. It turns instead to critical evaluations made by citizens, the deepening intensity of which can usefully be understood as deficits in legitimacy and support. It shows also how these evaluations can be worsened by elections themselves. Failing to stanch or even neutrally to record a government's diminishing legitimacy and support, the conduct of elections, heavy with manipulations, can exacerbate popular discontents. Accordingly, citizens approach electoral contests in new ways, using the limited competitiveness that had once helped to placate them now to cast their ballots in concerted protest.
A slightly different and earlier version of this paper was published at the beginning of this year by the Journal of East Asian Studies, based in Seoul, Vol 10, #1. Copyright © 2010 by the East Asia Institute. Used with permission by Lynne Rienner Publishers. I would like to thank Ms. Chow Wing Yin for her assistance in researching, calculating, and graphing statistical data.
In these conditions, while elections had earlier given a boost to evaluations of the dominant party's legitimacy, they now help to run these assessments down, a metaphor for which can be found in a mechanical flywheel that has been thrown into reverse. And lest the dominant party react by stealing the contest and jamming the flywheel, democratic change may take place. In this trajectory, change does not originate in the failure of the dominant party to generate patronage and manage elite relations. And further, elections do not merely signpost the transitional pathway. Change commences instead with the failure of the dominant party to perform in ways that citizens assess as legitimate or at least worthy of support. Elections then deepen this appraisal, their manipulations adding to the grievances of citizens, yet their residual competitiveness still offering a mechanism by which change may begin.
Introduction
It was not democracy's falling back into harder forms of authoritarian rule that drew comparativists away from their study of its consolidation. Since the 1970s outright breakdowns have been few. Rather, governments have more typically ceased or slowed their democratizing progress, seeking unbroken incumbency and political equilibrium in various kinds of hybrid regimes (Karl 1995; Diamond 2002). In the most effective approach, governments form a single dominant party, then fuse it with the state apparatus. And while regularly holding multiparty elections, they dampen, though do not extinguish, the competitiveness of these contests by monopolizing patronage and manipulating procedures. In this way, a type of hybridity sets in that Levitzky and Way (2002) conceptualize as electoral authoritarianism. Further, in adopting this form, governments that seek interminably to perpetuate their tenures may find new efficiencies, avoiding the costs of coercion imposed by hard autocracy, while reducing the uncertainties of democratic politics. As Bunce and Wolchik (2009: 97) recount, “the norm, even when the economy is failing, is for leaders in electoral authoritarian regimes to win one election after another.”
Even so, governments that operate electoral authoritarian regimes are occasionally stunned by the results of the contests that they wage. Levitzky and Way record a lengthening roster of cases in Eastern Europe, Africa, and Central America where governments have manipulated multiparty elections, yet met with defeat, leading to their ouster and democratic change. A new research agenda has thus emerged over when, under conditions of electoral authoritarianism, elections might be regime-sustaining or regime-subverting. The latter event amounts to what Howard and Roessler (2006) have recently labeled a “liberalizing electoral outcome,” perhaps presaging what is sometimes understood as “democratization by elections” (Schedler 2002, 2006; Lindberg 2009).
However, in casting doubt on the thrust of this agenda, Jason Brownlee (2007: 30-32), in a four-country study, has argued that elections in themselves do little either to sustain or to subvert single-party dominance and electoral authoritarianism. In his “institutional” theory, electoral outcomes are merely a “symptom,” passively reflecting the extent to which governments have maintained their dominant party apparatuses and prevented elite-level defections beforehand. Kenneth Greene (2007: 14, 63, 306), in an influential analysis of contemporary politics in Mexico, suggests similarly that contests are won or lost long before election day. In his “resource” theory of “hyper-incumbency” advantage, elections show only the extent to which dominant parties have kept control over public sector resources and state patronage, forcing the opposition parties that it confronts to the ideological fringes. Benjamin Smith (2005: 431), in measuring authoritarian durability across four cases, contends too that the “crucial task” of dominant parties is to maintain the loyalties of “in-groups” by guaranteeing their “long-term interests.” And in another study of Mexico, Beatriz Magaloni (2006: 18) also highlights the importance of patronage and elite-level cohesion, writing that “hegemonic parties must distribute ample spoils […] so as to deter elites from splitting.”
But Magaloni gives equal weight to mass-level attitudes and elections. Though she eschews any analytical usage of legitimacy (Magaloni 2006: 12-13), she addresses the ways in which a government's performance is appraised by citizens, as well as the chances of its electoral defeat. And the account of Mexico that she offers, with voters finally ousting the country's long dominant Institutional Revolutionary Party (PRI) in 2002, anticipates in some ways the setback suffered by its counterpart in Malaysia, the United Malays National Organization (UMNO), in 2008. Like the PRI once did, UMNO has long delivered patronage, therein perpetuating elite-level cohesion. Yet in the recent election, though the coalition led by UMNO, the Barisan Nasional (National Front), was returned to power, it was gravely weakened. Indeed, so startling was its setback that the contest was popularly interpreted as an “historic ‘victory’” for the opposition (Weiss, forthcoming). Thus, while a transition to democracy has not been completed in Malaysia, the contest in 2008 amounted to liberalizing electoral outcome, offering the prospect of a “new beginning”. 2
Howard and Roessler 2006: 366. In Freedom House's (2009) most recent country index, Malaysia's political rights score fell short of the full point in improvement that Howard and Roessler (2006: 369) cite as indicating an LEO. However, a trend arrow indicates “positive movement” within Malaysia's already “partly free” rating.
In other ways, though, the Mexican case departs sharply from Malaysia's experience. Magaloni (2006: 194, 207) tells us that the assessments made by citizens of the PRI's performance were darkened in 2002 by a “dismal longer-term economic record” and expectations of an “end-of-term crisis.” Citizens were also encouraged to act on their discontents by the government's having earlier formed an independent election commission, rendering contests so fair that the PRI was, after seven decades in power, seen as beatable (Magaloni 2006: 217-22). Doubts thus emerge at this stage over whether Mexico's politics should even be conceptualized in terms of electoral authoritarianism.
In Malaysia, by contrast, as elections approached in 2008, citizens were enjoying their sixth year of economic recovery (see figure 1 below). And like citizens everywhere, they had little inkling of the global financial crisis about to strike. In addition, the country's electoral commission remained firmly under the government's thumb, leaving citizens with little confidence that the UMNO-led Barisan could be defeated. Rather, in Malaysia, single-party dominance seemed utterly intact.
Malaysian Economic Growth and Inflation Rates 2003-2008
Thus, in examining recent political events in Malaysia, I argue that contrary to the findings of Brownlee, Greene, and Smith dwindling patronage and elite-level defections can be less crucial for the fates of electoral authoritarian regimes than are the critical evaluations made by citizens. However, unlike in the Mexican case analyzed by Magaloni, few citizens in Malaysia thought the government could be beaten, suggesting that their reasons for voting against it were different. Specifically, they were less tantalized by institutional reforms and possibilities of turnover than they were embittered by seemingly unstoppable corruption, iterated electoral manipulations, and the yawning deficits in legitimacy that resulted. Accordingly, far from shrinking before the electoral manipulations that Bunce and Wolchik (2009: 97) contend so demoralize voters, citizens in Malaysia were galvanized by them. In casting their ballots in 2008, then, they gave more thought to protesting against the government than they did to bringing the opposition to power. Citizens would thus be as stunned as the government was by the electoral results that they would produce.
In making this argument, questions over methods and evidence must be addressed. A claim that citizens were driven by deficits in legitimacy and support to vote in new and potentially transformative ways could best be substantiated, of course, with complete sets of public opinion data over time. In Malaysia, however, systematic polling has only recently been introduced. But some limited data available for the period just prior to the election in 2008 will be presented, suggesting the extent to which grievances had set in. Documentation will also be provided of the extraordinary measures with which UMNO politicians reacted in hopes of reenergizing Malay loyalties, as well as the greater intensity and new dynamics of protest activity that followed. Thus, cumulative evidence enables us reasonably to conclude that evaluations of weakening legitimacy among the Malays and calculations over support among the non-Malays instigated a liberalizing electoral outcome, bringing electoral authoritarianism under significant new pressures for change.
Legitimacy Deficits and Flywheel Reversals
To be sure, transitions from single-party dominance can sometimes begin within the party itself, with the defections of elites and the diminution of patronage paving the way for opposition parties to gain coherence and mobilize constituencies. In this trajectory, democratic change is neutrally registered, rather than accelerated, by elections. But in presenting recent data from Malaysia, this paper maps a second course. In Malaysia, pressures for change have not originated in elite-level divisions. They have emanated instead from citizens, alienated by precisely the ways by which elites have perpetuated their own cohesion and the dominance of their party. Further, the grievances of citizens have extended to the conduct of elections, adding separately to the impetus to use what competitiveness has remained to vote in protest. In this trajectory, then, democratic change is separately advanced, rather than simply revealed, by elections.
As analysis tilts from elite-level relations to mass-level sentiments, a fruitful way in which to assess the alienation of citizens involves deficits in legitimacy and support. To be sure, the notion of legitimacy, while once so cherished, is dismissed today as “unfortunately ambiguous” (Buchanan 2002: 689), leaving it a faded concept in political science. But the rest of this section demonstrates that in cases like Malaysia, it is able still to generate some analytical mileage.
One reason that legitimacy has fallen from favor is that it seems difficult to distinguish from pedestrian support. Yet where we are able to, it is precisely in making this distinction that the analytical value of legitimacy becomes clear. Specifically, where citizens evaluate the ways in which their government exercises and renews its state power as legitimate, their loyalties grow sticky. And thus, they identify closely with the government across multiple elections. But where they calculate that their government is worthy of no more than support, their affiliations remain shaky, sometimes shifting abruptly across contests. By understanding, then, the different intensities of legitimacy and support that underlie voter preferences, we can better account for political continuity or looming transition.
But how can we make this distinction between legitimacy and support? In a starkly divided or “plural” society like Malaysia (Furnivall 1956), scored nearly in half by “indigenous” Malays and “immigrant” non-Malays (the latter a negative residual for ethnic Chinese, Indians, and other social minorities), the difference between legitimacy and support is readily seen. In making judgments about political rightness, citizens in divided societies typically gather in competing communities through their valorization of kinship and the vilification of rivals (Horowitz 1993, 2000). And in appraising the government, they are then guided by a sense of birthright and indigenous entitlement or resentments over “second class citizenship” and wrongful exclusion. Favored communities thus find the government legitimate. Excluded communities, of course, find this same government to lack legitimacy – but may still view it as worthy of support if it is more accommodative than alternatives. In any given election in Malaysia, ethnic Malays and non-Malays in a particular district may vote the same way. Yet by recognizing the differences in legitimacy and support that underlie their respective preferences, we can gauge the probability and significance of change across contests. With evaluations of legitimacy far more viscous than calculations over support, any meaningful swing among Malay voters is unexpected and hence momentous. A far larger shift in non-Malay voting is routine and predictable.
Another reason that comparativists avoid using legitimacy for analysis is that on the highly differentiated institutional terrain that most modern polities possess, citizens make evaluations across a great multitude of arenas and processes. And thus, the broad assessments made by citizens are often shallow, varied, even incoherent, leaving them difficult to disentangle and measure. But in a single-party dominant system like Malaysia's, the object of study itself aids investigation. With party organizations, state apparatuses, and even business conglomerates fused in a tight amalgam of power, evaluations of legitimacy made by citizens gain focus. Put simply, throughout Malaysia's political economy, UMNO is omnipresent.
Thus, in Malaysia the criteria by which legitimacy are evaluated are primarily ethnic. And the object of evaluation is mainly single-party dominance. But though examining the dominant UMNO and the Barisan coalition that it leads through the prism of ethnicity, what behaviors specifically do citizens seize upon in making their assessments? On this count, analysis can be advanced by making still another distinction, this time between a government's substantive policy outputs and its “conformity to procedures” (Beetham 1991: 16; see table 1 below).
Popular Election Results and Distributions Parliamentary Seats
Note: 1969 transition year from democratic to electoral authoritarian regime; 1974-2004 electoral authoritarianism prevails; 2008 liberalizing electoral outcome.
Source: Institute of Southeast Asian Studies 2000; Government of Malaysia 2004; Moten 2009. Balance for the 2008 election consists of independent candidates and unreturned or spoilt ballots.
On the policy dimension, citizens apply ethnically framed criteria by which to evaluate whether the government distributes and renews public resources, mainly material, but also cultural, in ways that they regard as fair. Further, where citizens doubt their government's legitimacy on this score, they begin next to scrutinize more closely the procedural dimension. In particular, they assess the extent to which the government avoids corrupt practices while holding office, as well as electoral manipulations when seeking return to office. Thus, it is when evaluations grow critical on this second procedural dimension too that pressures for democratic change start to mount.
To be sure, where deficits spread across dimensions, it can be difficult to disaggregate them, further problematizing the conceptualization and analytical utility of legitimacy. Citizens may loathe cronyism both because its exclusivity disadvantages them materially, while its jobbery offends them morally. However, though the deficits in legitimacy that result may come to overlap, a distinction appears as we observe their sequencing. In Malaysia, when the government has delivered adequate policy outputs, citizens have mostly tolerated its corruption. It has only been after its policies have faltered that conformity to procedures has been questioned. Similarly, in the developed world, so long as the securitization of mortgages inflated the pensions of small-time investors, few citizens assessed the rightfulness of investment bank practices. It was only after the bubble burst and taxpayer-funded bailouts became necessary that citizens were appalled by subprime lending, over-leveraging, and bonus-based incentives. Thus, as grievances mount over failed policies, they may gradually extend to shoddy procedures. A reverse sequencing rarely occurs, with citizens clamoring to fix what does not yet appear to be broken.
Accordingly, under conditions of single-party dominance, while a government's policy outputs are regarded as rightful, the manipulations of elections that take place, as well as the corruption that persists, remain under-examined by citizens. Flywheel effects thus top up the legitimacy that the government has earned beforehand, ensuring that elections are regime-sustaining. But where a government's policies suffer legitimacy deficits, the manipulations that had earlier been ignored attract new scrutiny, thereby adding to fast-rising grievances. The electoral flywheel thus stutters, then goes into reverse, now helping to run evaluations of legitimacy back down. Moreover, to the extent that competitiveness remains, citizens find in elections an outlet through which to declare the increasingly critical evaluations that they make. At this crook in the pathway, then, elections may become regime-subverting, therein producing a liberalizing electoral outcome.
But why, when policies are evaluated by majorities of citizens as legitimate, would a government resort to electoral authoritarianism, therein acquiring the record of manipulations that can come later to haunt it? In Malaysia's divided society, policies and procedures that historically have been regarded as rightful by the Malays have been excoriated by the non-Malays. UMNO has strongly favored the Malays through a comprehensive affirmative action program, the New Economic Policy (NEP) (see Faaland, Parkinson, and Saniman 1990). Most Malays, then, in evaluating the government's distributive policies, have looked upon the quotas of public resources by which they are benefited as rightfully commensurate with their indigenous, indeed “sovereign” status. They have similarly viewed the government's developmental strategies, characterized by a reliance on state-owned enterprises staffed almost exclusively by Malay personnel, as well as an historical preference for foreign investors over local Chinese entrepreneurs (Jesudason 1989). The Malays have traditionally asked few questions on the procedural dimension, then, about UMNO's corrupt practices and electoral manipulations. Indeed, at election time, Malay villagers have long been advantaged on this count too by the weighting that heavily favors rural constituencies, the on-the-spot development grants that UMNO politicians dispense when campaigning, and the single-member district “winner-takes-all” approach that magnifies popular majorities into overwhelming representation in parliament. 3
The UMNO-led Barisan‘s practiced electoral manipulations are well-known. For an overview, see Gomez 1998.
Meanwhile, the non-Malays have chafed under the NEP's skewed distributions. They have regularly disparaged UMNO's developmental strategies too, unfavorably comparing the results with those achieved in largely Chinese Singapore. But though many non-Malays have then also been driven to pass harsh judgment on the procedural dimension, they mostly supported the Barisan in elections held during 1995-2004. During this decade, with UMNO at least mildly restrained by the non-Malay component parties in its ruling coalition, the government has never so rigorously imposed the NEP's quotas that the livelihoods of the Chinese were seriously crimped. And it never took up the full Islamist agenda of its chief rival, the Islamic Party of Malaysia (Parti Islam se-Malaysia, PAS), leaving space in which non-Malay cultures might be practiced.
The UMNO-led Barisan has thus held elections because it has so reliably won them. And in so doing, it has found that elections do much more than Brownlee and Greene acknowledge. With the government's policy outputs having earned legitimacy and support beforehand, its procedural abuses are mostly overlooked, enabling even manipulated elections to boost perceptions of rightness. But further, with the government having narrowed politics into electoral authoritarianism (after nearly losing a contest four decades ago), its sundry manipulations (especially the use of the plurality system), have enabled it in every election except the most recent one to bolster its simple majorities into two-thirds parliamentary majorities (see table below), constitutionally necessary for its freely amending the charter. Indeed, Magaloni (2006: 32-42) suggests that it is this, the capacity unilaterally to alter institutional rules, that most motivates governments to manipulate elections, even when they could win fairly.
But for reasons that we will explore, the legitimacy that the UMNO-led Barisan has enjoyed among the Malays has recently been strained. And the support that it has drawn from the non-Malays has been sorely diminished. In this situation, the separate impact that elections can have has grown more observable still. Briefly, as the government's substantive policy outputs began to sag, citizens scrutinized its conformity to procedures more closely. Public opinion data will be presented in the next section showing that citizens came to view corrupt practices with greater irritation. And an analysis of changing patterns of protest activity will show that as elections approached, citizens no longer ignored the manipulations that they expected to thwart their revised preferences. In these circumstances, the electoral flywheel slipped into reverse, helping now to blacken the assessments that citizens made about the government's legitimacy and worthiness of support.
UMNO under Abdullah
In late 2003, UMNO's long-serving president and Malaysia's prime minister, Mahathir Mohamad, stepped down. He was succeeded in both posts by his deputy, Abdullah Badawi, through an orderly transfer of power, again evincing the durability that single-party dominant systems can attain. Further, in seeking legitimacy among the Malays, Abdullah promised new attention to policy outputs and procedures. But he extended a hand also to the non-Malays, pledging to be “a leader of all Malaysians” (Lee 2008: 187). Thus, barely six months after Abdullah had ascended to the prime ministership, the UMNO-led Barisan won its greatest electoral victory ever, refreshing assessments of legitimacy and support (Moten 2006). However, though off to a strong start, Abdullah would soon disillusion citizens on multiple fronts.
Policy Outputs
To maintain distributive fairness in the eyes of the Malays, Abdullah continued the NEP's quotas. And to refresh their perceptions of rightness in developmental strategies, he shifted emphasis from the megaprojects that so puff up “Malay millionaires” and the Chinese tycoons with whom they partner to more modest agrarian and village-level pursuits (Case 2005). As one example, opposite the Iskandar Development Region in Peninsula Malaysia's industrialized south, geared principally to Singaporean investment in medium-tech services, Abdullah unveiled the Northern Corridor Economic Region, dedicated to new agricultural and biotech programs in the upcountry “Malay states.” Further, while during Abdullah's tenure Malaysia's economy never regained the high-speed rates of growth that had prevailed prior to the Asian financial crisis of the late 1990s, recovery continued to be driven by export competitiveness right up to the time of the 2008 election.
Even so, though the NEP remained in force, there were reasons for the Malays to doubt distributive fairness. And if the country's economy continued to recover, doubts crept in too over developmental performance. During 2007, global petroleum prices surged, prompting the government to cut local fuel subsidies for ordinary consumers. But at the same time, it renewed subsidies for “highly lucrative” independent power producers (IPPs), many of them owned by Malaysia's “politically connected and wealthiest business families” (Netto 2008b). This policy decision, backgrounded by longstanding perceptions of “cronyism” and “money politics,” corresponds with weakening evaluations of legitimacy at this juncture. In polling conducted by the Merdeka Center in late February 2008, shortly before the election, more than 60 percent of Malay respondents agreed that in the tendering of state contracts, UMNO politicians benefited most. At the same time, fewer than 30 percent agreed that UMNO was “fighting for Malay rights” and that to vote against the party would weaken Malay unity (see table 2).
Percentages of Respondents
Source: Author's own compilation.
These results are especially striking because as Ibrahim Suffian (2009: 93), director of Merdeka Center (2008), Malaysia's leading political polling research company, observes, the country's citizens are typically inhibited by their unfamiliarity with survey activities and by a “conservative and restricted political culture.” Accordingly, Ibrahim suggests that respondents tend reflexively to favor incumbent politicians, “particularly those from the ruling party.”
Even so, claims derived from public opinion data at a single point in time must be cautiously advanced. While the evaluations made by the Malays over the government's distributive fairness may have been critical in 2008, the unavailability of similar data from the 1980s-90s raises questions over whether they might have been equally scathing during the decades before. However, in combination with other evidence, it can reasonably be interpreted that the opinion data do indeed indicate that popular resentments had reached a new nadir. Much of this additional evidence is found in the quickening pace and new character of protest activities that occurred during 2007-2008. These upsurges will be explored more thoroughly in later sections, but it is worth mentioning one incident here. In January 2008, Malay demonstrators, organized largely by PAS officials, gathered in front of Kuala Lumpur's “iconic” Petronas twin towers, christened for their owner, the national petroleum company, Petroliam Nasional. And until roughly dispersed by riot police, they protested vigorously against the government's reductions in fuel subsidies (Fauwaz 2008).
The Malays also looked more skeptically on developmental performance. Despite economic recovery, growth rates fell far short of the cracking pace that was popularly remembered as hallmarking the Mahathir years. Projects like the Northern Corridor, then, while appropriately tinged with Malayness, sparked little investor interest (Hew 2008: 217-20). Inflation, as well as unemployment, also began to rise (see figure above). Accordingly, the Merdeka Center's polling revealed that for a plurality of Malays (31 percent), as well as for Chinese (20 percent) and Indians (26 percent), living costs had become their topmost concern.
In these circumstances, the evaluations made by the Malays of the legitimacy of the government's policy outputs began to slip. We find evidence for this not just in opinion data and protests, but also in the new stridency with which alarmed UMNO politicians, in issuing piercing communal appeals, tried at their party's annual assemblies to reenergize their constituencies. During his tenure, Mahathir had discouraged avowals of Malay supremacy in these meetings’ sundry speeches and “debates,” for they deterred the non-Malay entrepreneurism upon which his industrializing visions had come finally to depend. But Abdullah now gave freer rein. Thus, at the assemblies held in 2006 and 2007, the leader of UMNO's youth wing, doubling as education minister, drew a ceremonial Malay dagger during his address, the mystical keris, in a striking revival of communal ascendancy (Zahiid 2008). Other speakers then rushed to express their commitments to what had come to be hailed as the “Malay Agenda” (Lee 2008). And the vehemence of these displays, as well as the enthusiasm with which they were met by delegates in attendance, were duly transmitted nationally by party-aligned television.
The extent to which these appeals may have strengthened perceptions of legitimacy among the Malays is unclear. But there can be little doubt that the assessments made by the non-Malays, never having risen above base support, grew more critical. Thomas Pepinsky (2009: 109-110), in a sophisticated analysis, attributes the electoral setback that the UMNO-led Barisan would soon suffer to Abdullah's personal moderation, dissuading him from intimidating the non-Malay communities in the ways that Mahathir had early in his tenure. But if Abdullah himself recoiled from communalist threats, he acquiesced in their use by others, remarking at the end of the 2007 UMNO assembly that “the keris is a weapon, but is also a weapon to protect yourself and your friends” (Beh 2007b).
Conformity to Procedure
As the Malays grew steadily more ambivalent and the non-Malays more deeply alarmed, many citizens began to extend their scrutiny from policy outputs to procedures, focusing intently on the ways in which UMNO politicians exercised and renewed their state power. From the start of his prime ministership, Abdullah was aware that grievances over corruption had been simmering. He thus took steps to counter it. In brief, Abdullah introduced a National Integrity Plan, then gave it form with a Malaysian Institute of Public Ethics (Lyall 2004). He ordered spot checks on government agencies most associated with corrupt practices (Lopez 2003-2004). An independent body to investigate complaints against the police, perhaps the most distrusted element in the state apparatus, was also set up. Charges over conflicts of interest were brought against a former deputy cabinet minister, some top civil servants, and a prominent Chinese tycoon (Lopez 2004), actions that were unprecedented in their challenging an ethos of elite-level impunity. And several dubious megaprojects were suspended (Derichs 2007).
As mentioned above, the UMNO-led Barisan was rewarded with a handsome electoral victory in 2004. Abdullah was thus encouraged to demonstrate still more of his government's conformity to procedures. He required now, for example, that government MPs declare their personal assets (Theophilus 2004). But much more strikingly, the judiciary recovered some of its earlier independence. In late 2004, the courts overturned the conviction of former deputy prime minister, Anwar Ibrahim, jailed for sexual misconduct and corruption after having challenged Mahathir's leadership in 1998 (Jayasankaran 2004). Anwar was thus immediately released, enabling him swiftly to return to public life.
However, after so raising expectations among citizens over procedural reforms, Abdullah was confronted by the immutable requirements of singleparty dominance. As James Chin and Wong Chin Huat (2009: 83) make clear,
Patronage politics is hard-wired into the UMNO and [Barisan] party machinery; no party leader who tries to rip this infrastructure out is likely to survive politically.
Thus, at UMNO's party election in 2004, held barely eight months after Barisan‘s triumph in the general election, Abdullah was persuaded to lift a ban on “campaigning” that he had earlier imposed, seeking to halt the payments made routinely by aspirants for delegate votes. Though these exchanges besmirch UMNO's image, they tighten the party's apparatus upon which elite-level cohesion depends. As one hopeful at the UMNO meeting queried, “What's the fuss about? Whatever money is given helped us recoup some cost. I do not think that the leadership should be too worried” (Pereira 2004a). And Abdullah finally concurred, lamenting, “What can I do?” (Pereira 2004b).
Later, the government officials and lone business tycoon who had been charged with corruption, an action that had so animated citizens, were acquitted. Barisan MPs slackened in reporting their assets. Indeed, one UMNO legislator, his company having been found by customs to have smuggled illegally cut timber into Malaysia, was reported to have asked agents “to close one eye” (Case 2006). And in defending himself afterward before parliament and the media, he made plain the futility of Abdullah's efforts to instill probity in the party, insisting, “I don't know whether my company was involved. Maybe yes, maybe no. If yes, so what? Why can't an MP take care of his own interest?” Finally, if some of the projects that benefited UMNO politicians had been suspended, new ones were started. Indeed, Abdullah's own family members were reported to have entered into deals involving government contracts and the privatization of assets, allegedly enriching his son and son-in-law (Gatsiounis 2007).
Resistance to accountability also cropped up within the police, a vital locus of power over which a dominant party must keep its grip if activated social forces are to be contained. On an internal website, top police officials made known their contempt for Abdullah's proposal to set up the complaints commission (Kuek 2006). Indeed, while the police had once so dutifully served Mahathir in suppressing the reformasi movement that had been activated by Anwar's arrest in 1998, they threatened now to switch their allegiance from UMNO to PAS. Dan Slater (2003) has shown how Mahathir had “personalized” the security forces during his confrontation with Anwar, yet maintained their functionality, hence shedding new light on institutional adaptation and authoritarian durability. Under Abdullah, though, we observe that efforts to depersonalize the police and hold them accountable risked new dysfunctionalities, driving them into the arms of the opposition. Abdullah thus retreated, heavily watering down the complaints body that would eventually be formed. Scandals during 2007 also engulfed the director of the Anti-Corruption Agency (ACA), itself a fount of corruption, as well as the chief justice of the Federal Court, with a prominent lawyer caught on a secretly made video “brokering” judicial appointments and promotions with him. Yet despite the public resentments that flared, these officials were left quietly to retreat from view, their contracts lapsing, but without charges being laid.
Thus, with citizens already doubting the rightness of policy outputs, they now extended their scrutiny to procedural abuses. Seeking to prevent this, Abdullah had promised thoroughgoing reforms. But after so raising expectations, he failed utterly to fulfill them, thereby breeding new levels of cynicism. Analysts, activists, and opposition politicians then spotlighted abuses through unprecedented usage of the internet. And with their blogsites circumventing the controls on mainstream media outlets that electoral authoritarianism imposes, they in Chin and Wong's (2009: 80) estimation “helped to create a major shift among the middle class and sent many concerned citizens into political action.” Indeed, Chin and Wong (2009: 78) regard the release of the video clip involving the chief justice as a “catalyst”, after which “the whole political climate […] changed.” And as elections drew nearer, we will see that citizens grew more vexed still over the deep manipulations that they anticipated.
Legitimacy Deficits and Public Protest
Preliminary evidence that the government's legitimacy had slipped is found in the results of opinion polling conducted among ordinary Malay citizens. It is also found in efforts of UMNO politicians to reenergize their constituencies with increasingly strident communalist appeals. But we locate still better evidence for weakening evaluations of legitimacy made by the Malays, as well as declining calculations over support made by the non-Malays, in the character of public protests that followed. Direct action, when mounted autonomously from below, is prohibited under any electoral authoritarian regime. After all, the logic of its hybridity lies in its squeezing the activism of often unruly civil society into manageable electoral arenas. Thus, much more than through survey responses or in camera speeches, we find evidence for legitimacy deficits in citizens taking their grievances to the streets (see Gilley 2008: 273). Indeed, the anti-system character of public protests under these conditions is declared by their very illegality.
Large-scale protests were mounted throughout 2007 and early 2008 in Kuala Lumpur's center, its suburban environs, and even in the rural Malay states. They were mostly led by officials in PAS, the DAP, and the National Justice Party (PKR), the latter vehicle having been formed at the time of Anwar's arrest and that now mostly recruited young reform-minded Malays. However, among the protests inspired by grievances over distributive fairness at this juncture, it was perhaps those mounted by ethnic Indians, organized through the Hindu Rights Action Force (HINDRAF), that grew most potent. Barred from most public sector positions under the NEP, but unable freely to enter private domestic markets controlled principally by the Malays and Chinese, many Indians viewed their community as having benefited least from Malaysia's rapid industrialization. And galvanized now by the demolition of Hindu temples as plantation land was cleared for new projects, Indians from “all class backgrounds” (Welsh 2007: 2) marched under HINDRAF banners in protest. But like those who had earlier demonstrated over fuel costs, they too were met by riot police. Indeed, HINDRAF leaders were arrested and held without trial under the country's Internal Security Act (ISA).
Protests over distributive fairness also took on new dynamics, spreading from their usual inter-ethnic tinderbox to new sites of intra-ethnic disparity, especially within the Malay community. As noted above, though economic recovery continued during this period, anxieties mounted among citizens over living costs. Increasingly, then, as doubts about distribution seeped into assessments of developmental performance, protests were joined now by “low-income Malays,” decrying the patronage that a “conspicuously consuming elite” so habitually extracted through the NEP (Baradan 2007).
In addition, as doubts about rightness spread across, but also within ethnic communities on the policy dimension, citizens began collectively to scrutinize procedures. An umbrella movement labeled Bersih – a Malay word for “clean,” but also an acronym for Coalition for Clean and Fair Elections – took shape. And though Bersih articulated sundry grievances, it focused most intently on electoral manipulations. In late 2007, Bersih was denied a permit by the Kuala Lumpur police to mount a large protest. Yet the movement's leaders, Anwar Ibrahim and top PAS officials, pressed ahead, their anti-system actions plainly displaying their low evaluations of the government's legitimacy. They at the same time, however, cunningly portrayed themselves as remaining within Malaysia's rightful political and socio-cultural order, implying the greater rightfulness that they possessed. In particular, after demanding electoral reforms, Bersih‘s leaders and its 60,000 followers moved upon the royal palace, there to petition the country's king (Lee 2008: 198), high symbol of constitutional restraint and arbiter of traditional Malay culture.
To be sure, illegal public protests have regularly punctuated Malaysia's political record, with upheavals taking place during the late 1980s, then again during the late 1990s. But their pace grew quicker during 2007-2008. As Chin and Wong (2009: 79) observe, “to have such big protests within a short span was unheard of.” The character of these protests was also qualitatively different. Most signally, unlike the recent demonstrations, earlier ones had sprung from ructions within UMNO itself. Following the trajectory sketched by Brownlee and Greene, economic shocks had depleted public sector resources and state patronage, placing such strains on UMNO's party apparatus that factional splits and leadership challenges ensued. In brief, during the late 1980s, UMNO split into two factions designated Team A and Team B, with the leader of the first, Mahathir, the patron of “Malay millionaires,” confronted by Tengku Razaleigh, championing small Malay contractors (Khoo 1992). As their rivalries deepened, reverberating through UMNO party elections, parliament, the courts, and connected conglomerates, team leaders appealed for mass-level support through raucous public rallies. Indeed, Team A resorted to communalist strategies, with the rallies they held bristling with banners and chanting that grew sharply anti-Chinese in tone. At the same time, Team B organized a splinter party, Semangat ‘46, that rose in opposition.
A decade later, Mahathir's leadership was again challenged at an UMNO assembly, this time by Anwar Ibrahim (Case 2003). Anwar was then purged from UMNO and the cabinet. And as he responded by appealing directly to citizens over Mahathir's abuses, large rallies again took place. Anwar was arrested under the ISA and, in a notorious episode, beaten by police. In these circumstances, PAS was able rapidly to recruit large numbers of aggrieved Malays, while Anwar's supporters organized the PKR. In addition, social activists launched the reformasi movement, filling the interstices between opposition parties while invigorating civil society organizations (Weiss 2006).
Accordingly, during the late 1980s and the late 1990s, protests involved citizens who, amid economic shocks, had been mobilized through top-down, sometimes communally-ordered appeals made by fractious UMNO politicians. And just as Brownlee and Greene would predict, as the economy recovered and public sector resources and state patronage were replenished, UMNO reasserted its dominance and elite-level cohesion. Thus, if the share of popular votes won by the UMNO-led Barisan fell in the elections held in 1990 and 1999, it was each time restored in the contests that followed. Indeed, as we have seen, the government won by its grandest margins ever in 2004.
The protests that took place during 2007-2008 were qualitatively different, their dynamic commencing from “outside-in.” There had been no prior erosion of single-party dominance, with UMNO politicians divided by team rivalries or purges. It was instead the ways in which elites maintained their party's dominance and their own cohesion, so weakening assessments of legitimacy and worthiness of support that citizens rose up autonomously. The evidence presented in this section turns on the illegal and hence, anti-system nature of the direct action to which citizens resorted. Indeed, they came to exercise their civil liberties with a fervor that electoral authoritarianism proscribes. But it is also a measure of their new autonomy that in their rejectionism, they began to breech the communal walls to which they had for so long been attuned. Cross-ethnic participation in protests organized by PAS, the PKR, and the DAP rose markedly. Of course, ethnic criteria for the evaluation of legitimacy did not simply dissolve. In 2009, PAS leaders would split over their party's collaborating with the DAP, with one faction even seeking engagement with UMNO (Ong 2009). But during 2007-2008, citizens had grown so alienated that deficits in legitimacy and support began to trump ethnic loyalties. And as the next elections approached, these deficits would grow even deeper.
The 2008 Election
Malaysia's most recent general election, held in March 2008, amounted to a liberalizing electoral outcome. Yet, despite this shift in the country's political trajectory, we have seen there had been no weakening beforehand in the dominant party's apparatus or any cascading elite-level defections. Nor had the party lost control over public sector resources and state patronage. Rather, the methods by which UMNO perpetuated its dominance – marked by a fusion of party and state apparatuses, a monopolization of resources, and a pulsing collusion between elites – came to diminish its legitimacy and worthiness of support in the eyes of many citizens.
Further, as they grew disillusioned with policy outputs, citizens reviewed the medley of procedures by which they had earlier been lulled, reassessing corrupt practices more critically, but also the electoral manipulations that they had mostly overlooked. Hence, as the next election approached, these manipulations, while having helped earlier to disperse societal discontents, now inflamed them. And in accreting with other grievances, they raised fresh doubts about legitimacy and support, observable in the protests that citizens mounted. But in addition, citizens would now make concerted use of the electoral competitiveness left to them, taking their protests from the street into the polling station.
What is more, we will see that in a reversal of the causal ordering specified by institutional and resource theories, it was only after the election had been held that cracks in UMNO's party apparatus appeared, with losses in patronage leaving it vulnerable to elite-level defections. Opposition parties, by contrast, buoyed by the protest votes that citizens had cast, upgraded their collaborative front into a somewhat hardier arrangement, Pakatan Rakyat (People's Alliance). And while acquiring resources through the five state-level governments that they now formed, Pakatan angled to replace the Barisan at the national level by recruiting defectors. In this way, one of the world's most enduring cases of single-party dominance and electoral authoritarianism was brought to the brink of democratic change.
A Liberalizing Electoral Outcome
As doubts about the rightness of a government's policies spill over onto the procedural plane, Bruce Gilley (2005: 61) contends that “low legitimacy will tend to create pressures for changes to the state itself.” Jason Brownlee (2007) counters, however, that in their sturdiness, dominant parties may instead continue effectively to mediate patronage among elites, rendering these parties, the state apparatuses they infest, and the authoritarian regime that they operate highly resistant to exogenous pressures. In Brownlee's interpretation, then, if democratic change is to take place, it must occur from inside-out, with top politicians, aroused by shortfalls in patronage rather than any deficits in legitimacy, weakening the dominant party by defecting. Kenneth Greene (2007), in his analysis of Mexico, similarly ascribes the resilience of dominant parties to their mastery of patronage, even if giving greater emphasis than does Brownlee to the inclusion of ordinary citizens, with patronage winning them over as voters, while blunting the appeals of opposition parties. And Magaloni (2006: 18), we recall, though paying yet greater attention to citizens and the government's need of their support, argues that dominant parties, in perpetuating “ample spoils,” avoid elite-level splits. Thus, it is only when patronage falls short, usually through contractions in public sector resources, that the dominant party falters, ceding scope in Greene's schematic for an opposition party to rise from its ideological “niche,” overshadow its rivals, wrest away voters and take power, therein advancing democratic change.
However, these accounts ignore the extent to which a dominant party, in conducting business as usual – mediating patronage in ways that perpetuate elite loyalties while wrong-footing opposition parties – can deeply alienate citizens. And it is by invoking the notion of legitimacy deficits, first on a policy dimension, then on a procedural dimension, that we discover the varying intensities with which grievances may set in, finally driving citizens to pose the transformative pressures that Gilley describes. In addition, under electoral authoritarianism, citizens are encouraged to continue their protest at the ballot box. Indeed, their alienation is now deepened by the manipulations that they anticipate will distort their preferences, causing the electoral flywheel to slow, even to reverse.
Malaysia's twelfth general election has become one of the most studied contests ever to have been waged in the Southeast Asian setting. Given the number of good analyses, then, that have recently become available (e.g., Brown 2008; Kee 2008; Maznah 2008; Ong 2008; Ooi et al. 2008; Ufen 2008; Welsh 2008; Chin and Wong 2009; Moten 2009; Pepinsky 2009; Weiss forthcoming), only a short account is necessary here. Though the UMNO-led Barisan won this election, it was dealt a severe setback. The government's share of the vote fell from the approximately 55-60 percent of the total that it has usually commanded to a bare majority nationally, even less on the peninsula. Notwithstanding, then, the knock-on effects of the plurality-based single-member district system, it won only 140 of parliament's 222 seats, leaving it shy for the first time ever of a two-thirds majority. 4 Thus, the government lost its capacity to alter institutional functioning unilaterally, in Magaloni's (2006: 39) estimation the main rationale for operating a singleparty dominant system. Even more strikingly, it lost control for the first time also of four state governments (including Malaysia's two most industrialized, Selangor and Penang), while failing to retake a fifth.
For full results, see nstonline 2008.
With the Malays reevaluating the government's legitimacy, they produced a significant swing of five percent against the UMNO-led Barisan. And more than fleeing to PAS, they greatly boosted the PKR, then even reached out in some constituencies to the DAP (Brown 2008; Ong 2008). Further, as our framework would predict, the non-Malays shifted even more profoundly in their support. Because balloting is secret, of course, the precise pattern of ethnic voting cannot be known. But using ecological inference methods, Ong Kian Ming (2008) calculates that in Peninsular Malaysia, Barisan won only 58 percent of the Malay votes, 35 percent among the Chinese, and 48 percent among the once highly supportive Indian community.
In casting their ballots, citizens sought mostly to continue their protest. And they did this so concertedly that they imposed far greater casual force through the election than Brownlee allows. But in hindsight, it is also plain that some citizens were drawn more positively to visions of cross-ethnic unity. To this end, Anwar had campaigned tirelessly across the country, appealing to multiethnic audiences, while portraying the opposition parties as amounting to a viable alternative. He thus avoided the winnowing, go-it-alone strategies that Greene (ch. 5) identifies in the Mexican case wherein a pragmatic new leader of the opposition National Action Party (PAN) pushed past “early joiner” ideologues within his own organization, made an “end-run” around rival opposition parties, then adopted a “catchall” disposition through which to win over enough floating voters that that the ruling Institutional Revolutionary Party could be ousted. In Malaysia, with the PKR so new, there were no unyielding veterans whom Anwar had to overtake. Nor did the PKR try to outflank other elements in opposition. Rather, in appealing mostly to reform-minded Malays, the party served as the lynchpin between the Islamist PAS and the secular, largely Chinese DAP. In this way, Anwar took to mediating in order to forge a more cooperative front.
But if Anwar avoided the strategies that might have produced the streamlined opposition that Greene describes, so too did he fall short of the deep coalescence that Howard and Roessler (2006) specify as necessary for producing a liberalizing electoral outcome and that Bunce and Wolchik (2009: 103) identify as the “first requirement” for democratization by election. As Pepinsky (2009: 116) observes, Malaysia's opposition parties had “by any measure” been more united when contesting general elections in 1990 and 1999 than they were in 2008. Thus, if in Malaysia's most recent election Anwar's leadership mattered, citizens were more motivated to protest against the government than to embrace the opposition, signaling clearly their weakening assessments over legitimacy and worthiness of support.
In this way, in 2008, the PKR made striking advances, the DAP meaningful gains, and PAS at least held steady, enabling them collectively to deny the government its extraordinary majority in parliament. When the government had been dealt a similar blow in 1969, it regained its footing by absorbing most opposition parties into its new Barisan coalition, then firmly subordinating them (Mauzy 1983). But in 2008, the opposition was so energized by the results that it sought now to take power at the national level in its own right. Thus, even as leaders of the opposition parties bargained over the formation of the state governments that they now controlled, Anwar continued to hold rallies, striving to heighten pressures for change among citizens. He sought also to force by-elections in constituenices where the government had won. And he tried to entice some of the government's parliamentarians to defect, thought principally to be members of parties based in East Malaysia who were disgruntled over their peripheral standing.
Meanwhile, the UMNO-led Barisan was quite stunned, with top politicians in all the component parties engaging in uncharacteristic introspection, then canvassing reforms. As Staffan Lindberg (2009: 90) observes in such instances, startled incumbents may seek to “‘get ahead of the curve’ by legitimating themselves as forward-thinking reformers.” Gilley (2008: 273) writes similarly of a rushed adoption of “new value orientations.” Accordingly, in a front-page editorial, the New Straits Times (2008: 1), widely regarded as the government's English-vernacular mouthpiece, intoned that “the people have long been disgusted with the kind of boorish and loutish behavior that UMNO leaders have exemplified because of their grip on power since independence in 1957.” And the former UMNO chief minister of Selangor, Mohd Khir Toyo, succeeded now by a PKR assemblyman, contended that the “election results [were] a reflection of UMNO having lost touch with reality. This leaves the party with no option but to tread the path of reform” (Zahiid and Omar 2008).
Hence, in recognizing the government's legitimacy deficits, top UMNO politicians sought hurriedly to realign procedural functioning with the sentiments of citizens, canvassing a striking new round of reforms. Proposals included separating the ACA from the Prime Minister's Department, thus placing new checks on the executive, while setting up a judicial appointments commission, giving still more independence to the courts (Baradan 2008; Netto 2008a). Proposals were also made to ease requirements for the annual licensing of print media, the restrictions on student participation in politics, and even the conditions under which dissidents were detained under the ISA, therein strengthening civil liberties. At this juncture, then, with the opposition forging ahead, while the government pondered reformist concessions, speculation mounted that however slow moving, Malaysia's twelfth general election had precipitated a peaceful transition from singleparty dominance and electoral authoritarianism to two-party alternation and democratic politics (see Baradan 2008).
An “Outside-in” Trajectory
If a democratic transition has indeed begun in Malaysia, its dynamics have departed significantly from the weighting and sequencing of factors that have typically been delineated under conditions of single-party dominance. In particular, we have seen that leading theories negate the autonomous evaluations made by citizens and the separate causality of elections, contending instead that contests are won or lost by governments much earlier. As Brownlee (2008: 112) succinctly puts it, “elections do not destabilize regimes; [dominant parties] destabilize their own elections.” Though framing his argument differently, Greene (2007: 14) concurs, writing that “dominant parties’ pre-electoral advantages and in particular their virtual monopoly over patronage resources mean that they usually win elections before election day.” Thus, in crafting their institutional and resource theories respectively, they begin with the government's loss of state patronage, with Greene focusing closely on shrinkage of the public sector through economic shocks or privatization. And it is in these straitened circumstances that Brownlee turns next to elite-level defections from the dominant party to the opposition, with top politicians having lost confidence in their party's serving up what they adjudge as their rightful deserts.
But we noted that prior to the election in Malaysia, there had been no economic shock, contraction in public sector resources, or shortages in state patronage. Upon coming to power, Abdullah did suspend a few megaprojects. Yet this ran counter to a larger trend that had begun with the Asian financial crisis, one wherein assets that had earlier been privatized were reacquired by the state from now floundering Malay millionaires. Taking a leaf from Singapore's model, public enterprises were to be revitalized as “GLCs” (government linked corporations), then expanded with renationalized assets. Moreover, the government's investment vehicle, Khazanah Holdings, was restructured and more vigorously deployed as a sovereign wealth fund (SWF Institute n.y.). And even during the decade prior to the crisis, when privatization had been most fervently carried out, it was widely viewed as yet another form of patronage (Gomez and Jomo 1999: 91-98), nourishing the sinews between UMNO politicians and their corporate allies.
Thus, as the 2008 election approached, no defections took place of the kind upon which Brownlee's explanation for democratic change depends. Far more striking at this juncture than the ceaseless factional scheming that reverberates through any political party was the habituated cohesion with which politicians in UMNO prepared for the election. Indeed, though some incumbents were dropped as candidates (Welsh 2008), most politicians in UMNO continued to look upon the party in just the way that Brownlee portrays during good times, as a clearinghouse through which so long as they remained loyal, they could reliably extract largesse. If anything, then, it was the opposition that suffered defections, with the PKR's youth chief, its treasurer, and several of its state-level and division leaders so alienated by Anwar's “behaving like a ‘dictator’” that in 2007, they left for other parties, including UMNO (Beh 2007a; see also Muda 2007).
It was only after the election, then, that the UMNO-led Barisan‘s resources and patronage began to slip, with control over five state assemblies and bureaucracies now held by the opposition. To be sure, UMNO itself had fared reasonably well in the election, its candidates, whether incumbents or newly selected, winning two-thirds of the seats that they contested. But with the NEP having remained basically intact, a five percent swing in Malay voters against the government is significant. Further, in Malaysia's divided society, party dominance must not be strictly equated with UMNO. It must instead be understood in terms of the cross-ethnic coalition, Barisan Nasional, upon which UMNO, however pivotal, relies so heavily for non-Malay support. By itself, UMNO might win simple electoral majorities. But it could never gain the extraordinary parliamentary majorities that Magaloni demonstrates are integral to single-party dominance. And on this score, we note that Abdullah's acquiescing in the revival of communalist imagery, made most apparent at the party's annual assemblies, sacrificed the appeal of its non-Malay partners in Barisan, yet without fully reproducing Malay loyalties.
And it was thus now too, with the UMNO-led Barisan gravely weakened, that elite-level fractiousness set in, increasing the risk of cascading defections. In brief, Khir Toyo initially refused Abdullah Badawi's call to serve as opposition leader in Selangor's state assembly. At the same time, in two rural states where the National Front had been returned, UMNO politicians unfavored by Abdullah defied him by flattering their respective sultans and seizing the chief ministership posts. Next, Mahathir and his son, Mukhriz, a top position holder in UMNO's youth wing, demanded that Abdullah resign, a call that resonated among division leaders throughout the party's apparatus. Abdullah wavered, offering first to transfer power to his deputy prime minister, Najib Razak, but then vowing to defend his position at the party's next general assembly (Vasudevan et al. 2008). Mahathir derided Najib as a “coward” for failing to confront Abdullah (Choi 2008), prompting Najib to round on Mahathir. Mahathir abruptly resigned from the UMNO, while urging other members to do the same, then return to the fold after Abdullah had been forced out. Abdullah's law minister also resigned, protesting over the security minister's having detained under the ISA opposition figures, journalists, and bloggers who had been emboldened by Pakatan‘s rise. And in mid-2009, the one-time minister joined the PKR.
This fractiousness among top politicians in UMNO also spread throughout Barisan. A key official in Gerakan (People's Movement Party), still a component in the ruling coalition, but voted out now from its longtime governing role in the state of Penang, agreed to serve in a new administration led by the DAP. In addition, some 20 Gerakan division and branch leaders, finding their prospects in Barisan bleak, also defected to the PKR (Waheed and Abas 2008). And the Sabah Progressive Party (SAPP), an East Malaysian vehicle, tried to act even more decisively, seemingly in response to Anwar's entreaties. More than defecting, SAPP attempted to topple the government of which it was a member by spearheading a parliamentary vote of no-confidence (Aziz and Ghazali 2008). But though this vote did not take place, fractiousness deepened within UMNO, prompting Abdullah to agree finally to step down at the party's next assembly, indicating in yet another way how decisive the sentiments of citizens had become. Howard and Roessler (2006: 372) claim that amid a cluster of precipitating variables, a change in incumbent leadership can help in paving the way to a liberalizing electoral outcome. But in Malaysia, it was instead the election, by activating citizens and rattling elite relations, that triggered leadership succession.
Further, if the UMNO-led Barisan‘s electoral setback cannot be attributed to any prior elite-level defections of the kind outlined by Brownlee, neither can it be matched with the patterns of oppositional dynamics elaborated by Greene. As summarized above, in Greene's trajectory, the shrinkage of the dominant party's public sector resources and state patronage cede new opportunities for mobilization by opposition parties. Thus, in Mexico, amid widespread expectations of economic crisis, a new leader of the opposition PAN pushed past ideologues within his own organization, overtook rival opposition parties, then won over enough voters that the PRI could be ousted and democratization advanced. Little of this applies to Malaysia. Instead, the PKR, in appealing mostly to reform-minded Malay constituents, strove mightily to bridge the gap between PAS and the DAP. The PKR thus avoided forging the streamlined vehicle that Greene depicts. But equally, it fell short of building the broad and deep-seated coalition that Howard and Roessler (2006) and Bunce and Wolchik (2009) call for in their models. Notwithstanding Anwar's leadership abilities, then, the PKR fell limply between the poles, cobbling together an extremely diverse, internally tense, and indeed improbable front.
Hence, what stands out in Malaysia's 2008 election is that citizens were never activated by competing appeals from rival politicians in the UMNO-led Barisan. There had been no such open cacophony. Nor were they principally motivated by the overtures of collaborating politicians in what would later become the PKR-centered Pakatan Rakyat. The dissonance between its members remained too great. Instead, citizens were driven by their own critical reevaluations, made manifest in deficits in legitimacy and support. And it was this that most raised prospects for a liberalizing electoral outcome.
A last question emerges over timing. The alienation of citizens, evinced by opinion surveys, the rhetoric to which UMNO politicians resorted, and the public protests that followed, had grown intense. But why were these sentiments so much more heartfelt in 2007-2008 than they had been during earlier decades? What had changed, so extending popular scrutiny from policy outputs to procedural abuses that mass-level grievances, long simmering, were brought finally to the boil? This is intriguing because, if anything, political controls under Abdullah had grown laxer.
Many factors have been cited by observers, but two bear underscoring. First, with Abdullah's policy outputs failing to assuage the Malays, even while greatly alienating the non-Malays, his government's conformity to procedures grew critical. Thus, after promising more intently than any previous prime minister to roll back corruption, his utter failure to deliver bred deep disillusion. Second, the feeble leadership and continuing abuses that marked Abdullah's tenure were trumpeted ceaselessly over the internet, eluding the controls on communications imposed by electoral authoritarianism. And as citizens gained levels of insight that they had never before possessed, their grievances were redoubled. Abdullah had led the Barisan in winning its largest parliamentary majority ever in 2004. But just four years later, he led it also in its worst performance, starkly registering the shift in assessments made by citizens of legitimacy and support. As Abdullah himself observed after the election, though his government kept a tight rein on mainstream media, it had “lost the internet war” (quoted in Ramirez 2008).
Conclusions
In Malaysia, a transition from single-party dominance and electoral authoritarianism appears to have started. But its context differs from comparable cases in two important ways. First, in Malaysia, democratic change is occurring in what Pepinsky identifies as “non-crisis” conditions. Second, it is unfolding within a starkly divided society. As such, this transition has not been instigated by elites, with an economic shock so diminishing their party's resources and patronage that they have split and defected. Rather, it has been initiated by citizens, alienated over the procedures by which the party has perpetuated its dominance and elite-level cohesion. However, in doing this, citizens have not been uniformly motivated. Rather, across the country's Malay/non-Malay societal dyad, the intensity of sentiments has varied.
How can we best understand a transition that unfolds in this way, initiated by citizens, but complicated by their uneven motivations? Leading institutional and resource theories, in tracking the elite-level defections that take place amid material shortfalls in patronage, offer little help. The framework developed in this paper, then, begins with the mass-level alienation that appears amid deficits in legitimacy. It also addresses variations in sentiments across different communities, making a crucial distinction between the loyalties rooted in viscous legitimacy and the more fleeting sentiments that arise from conditional support.
In analyzing Malaysia's general election in 2008, most observers have fixed on the extent to which the non-Malays abandoned the UMNO-led Barisan. But by interposing the lenses of legitimacy, we are able to see that even more worrisome for the government than the loss of support among floating non-Malay voters is the slippage in emotive attachment among bedrock Malay constituencies. Support can be recovered by the government's offering the non-Malays better terms. But this can only be undertaken so long as the Malays remain forbearing. There is much evidence, however, that their patience has worn thin, with opinion data and protest activities revealing that many Malays perceive UMNO politicians, in sharing out the patronage upon which elite-level cohesion depends, to have taken too much. The entitlement of the wider community, grounded in indigenousness, has been violated, then, weakening perceptions of rightness. Accordingly, the sticky loyalties of the Malays have begun to grow as brittle as the support that is so conditionally given by the non-Malays. And in this situation, the government is even harder pressed to manage the demands of rival communities in ways that enable it to perpetuate its single-party dominance. Hence the government's trepidation over even a five percent swing in Malay voters against it.
In sum, the electoral setback dealt the UMNO-led Barisan in 2008 cannot be attributed to any crisis-induced shrinkage of resources and patronage. Evidence shows that it originated instead in UMNO's skewed used of constant resources, so favoring its politicians that citizens grew alienated. Of course, the corrupt practices through which patronage is conveyed have long been embedded in Malaysia's political economy. And their inadequate cloaking by the NEP has not recently grown any more threadbare. What is different, then, is that against a backdrop of deepening skepticism over policy outputs, Abdullah Badawi, in having so boldly pledged conformity to procedures, but then dashing the hopes that he had raised, drove even the Malays to reevaluate his government's legitimacy on the procedural dimension. And then, in seeking to revive their flagging loyalties, he allowed communalist rhetoric to resurface, therein shattering already tenuous support among the non-Malays. Grievances in both communities, moreover, were inflamed through an unprecedented intensity of internet usage, evading the controls on communications that electoral authoritarianism typically imposes.
Citizens grew more agitated still over procedures as the next election drew near, for they anticipated that just as corrupt practices persisted, so would electoral manipulations recur. To be sure, electoral competitiveness was probably greater under Abdullah than under his predecessor, Mahathir. But as Magaloni (2006: 196-98) explains, citizens make evaluations based on lengthy records of government performance. And thus, in Mexico, though there had been no economic crisis prior to the 2000 election, citizens were still prompted by the PRI's “dismal record” of economic management to expect one, encouraging them to vote against the dominant party. In Malaysia, though electoral manipulations may have grown no worse, citizens were similarly influenced by their recollections of the UMNO-led Barisan‘s checkered past. And despite the government's possessing a full hand of what Greene has subtly termed “pre-electoral advantages,” voters cast their ballots so concertedly in protest against it that they produced what Smith and Roessler have labeled a liberalizing electoral outcome.
The distinct causal impact that elections can have is inaccessible to institutional and resource theories. It can readily be discovered, however, through a framework based on legitimacy. Thus, in recognizing that legitimacy deficits can cause flywheel reversals, we see the pivotal effects of Malaysia's election in 2008. It was not before, but after this election had been held that the government's resources and patronage began to slip, with opposition parties now controlling legislative assemblies and bureaucratic apparatuses in five states. And thus, it was only after this contest too that the government finally grew vulnerable to cascading defections, while the opposition parties, astonished by their electoral windfall, drew closer together. Though no rationalized catchall party of the kind identified by Greene emerged, cooperation grew firmer under Pakatan Rakyat.
Against this startling electoral setback, it is unknowable whether Pepinksy (2009: 110) is right that the UMNO-led Barisan would have fared better in this election had it clamped down, resorting more heavily to the manipulations and coercion that it had used in the past: “jailing opponents, deploying resources to turn out supporters and suppress the opposition vote, and using the Malaysian media” to harp ceaselessly on the rise of communalism and radical Islamism. But with many citizens having become so alienated by the government's behaviors on policy and procedural dimensions, one can as reasonably speculate that any return to coercive tactics would sooner have brought them to the boil. We recall the quite counterproductive effects of the UMNO Youth leader's having raised the keris to menace the non-Malays on national television.
But whatever Malaysia's suitability for a framework that centers legitimacy, this paper offers only a single case study. It but narrowly illuminates a particular pathway along which a dominant party and electoral authoritarianism might yield to two-party alternation and democratic politics. Thus, it may apply less well to some of the other contemporary and historical cases placed commonly in this category, namely, Mexico, Taiwan, Singapore, Senegal, Zimbabwe, and increasingly Guinea-Bissau. Even so, while Malaysia's transition might be distinguished by the non-crisis conditions and stark social divisions within which it its unfolding, comparativists interested in the fates of single-party dominant systems and electoral authoritarian regimes must come to grips with it. And even if before completion, this progress provokes a sharp backlash, this paper succeeds still in demonstrating how an exemplary case of single-party dominance may now be unraveling. Indeed, its demise is past due, having lasted well beyond the 30-year life span that this regime type is said to average (Smith 2005: 423).