The origins of regional autonomy in Indonesia: experts and the marketing of political interests.

Smith, Benjamin

This article argues that, in contrast with prevalent choice-theoretic accounts of institutional origins in new democracies, the passage of Indonesia's regional autonomy laws in 1999 took place despite the interests of powerful political actors rather than because of them. Lacking the past experience to calculate retrospectively the likely electoral payoff from supporting an effort to devolve political power to Indonesia's city and regency governments, New Order-era political elites in Jakarta gambled on the advice of a team of experts. The experts assured them that supporting the effort would give them strong and salient reformist credentials on the eve of free elections. The conclusion of the article suggests that the political origins of regional autonomy in Indonesia have broad implications for the understanding of institutional genesis in new democracies, and that the potential impact of expert advisers is a fruitful focus of future research.

KEYWORDS: Indonesia, decentralization, federalism, institutionalism, transition

**********

On May 21, 1998, Bacharuddin Jusuf Habibie, President Suharto's final vice-president, assumed the presidency of Indonesia in the aftermath of a devastating economic crisis that spurred massive anti-regime protests and eventually Suharto's resignation after thirty-two years in office. Although according to the Indonesian constitution Habibie had a legal mandate to fill Suharto's term in office--which would have ended in 2003--he inherited a de facto obligation to prepare the country for a political transition much sooner than that. The buzzword for all of the demanded and actual changes to Indonesia's political system--Reformasi--became a hotly contested term. Little certainty or consensus existed, however, over the shape of reform or the transition, leaving Habibie the latitude to try and define them himself even as he tried to parry demands from activists, separatists, regime loyalists, and the international community.

Hoping to ride his success at managing the transition--and ideally to win an anticipated presidential election on his own--Habibie relied on a "change team" of experts, who came to be known as Team 7, to help him define Refonnasi. (1) Originally charged with reforming Indonesia's electoral system in preparation for the elections ultimately held in June 1999, (2) Team 7, after completing that task in late 1998, turned to writing a comprehensive reform of the country's unitary system that would devolve power to its regions. Only a few months later, the team presented the first draft regional autonomy bill to President Habibie, who agreed to support the bill, and then to the parliament.

In May 1999, the Indonesian People's Consultative Council (Dewan Perwakilan Rakyat, or DPR) passed Law 22 on regional autonomy and followed it with Law 25 on intergovernmental fiscal relations. (3) It has since been described as the world's largest political decentralization project. Nearly 2 million civil servants were transferred from the central government to the regency (kabupaten) and municipality (kota) levels along with authority over more than 40 percent of government expenditures and more than 60 percent of the national development budget.

More notable still is that these laws were pushed by a president and passed by a parliament made up entirely of holdovers from the Suharto era: the three official parties during this period and the army faction. (4) All three parties were highly centralized and vested in and committed to both the unitary state that held their interests and to the continuation of the centralized nature of the party system. (5) Moreover, although they all stood to gain from supporting democratic reforms in the eyes of the citizens whose votes they hoped to win, it remained highly uncertain which "reforms" would hold most salience. In short, it was unclear whether decentralization would prove to bring much in the way of electoral payoffs.

President Habibie, whose short-term political benefit from actively supporting decentralization was uncertain, had in the long term little to gain politically, either with his own party or in terms of diluting executive power, by agreeing to devolve so much authority to Indonesia's regions. In Indonesia, the party leaders who supported decentralization were all New Order-era appointees eager to save their positions in a new political system. Given an instrumental assessment of the certainty of giving up power and resources versus the uncertain benefits, politicians who should have opposed decentralization went along nearly unanimously. Why? More generally, why would politicians who ultimately stand to lose from decentralization support it anyway?

I argue here that regional autonomy became politically viable because the members of Team 7 took advantage of the conflicting short-and long-term interests between Habibie and the ruling party (Golkar) in (1) presenting democratic bona tides to voters and (2) retaining the advantages that political centralization provided them, respectively. Under conditions of immense political uncertainty, in which an outgoing authoritarian president and parliament wished also to be incoming democratically elected politicians, Team 7 experts played a crucial role in defining for them what the key issues were. In short, the members of Team 7 convinced both the president and the ruling party elite that the electoral benefits they could accrue by supporting regional autonomy would outweigh the longer-term political costs of doing so. This turned out decidedly not to be so, but by June 1999 the bills had become law. (6)

Much has been made of decentralization in the last decade. As a savior of both markets and democracy, decentralization has become a ubiquitous policy choice in new democracies, an economic policy choice considered among many others to be part of broad structural reforms, and an almost uniformly "good thing" for policy analysts. (7) Decentralization deserves the attention it has received. However, the political motivations behind its adoption have received less attention in this literature and are often assumed by virtue of the positive functions that decentralization can serve. But most political science research on this topic, focused on the question of why rational politicians in centralized systems would willingly give away elements of their own power, approaches the question from the assumption that political actors fully understand decentralization's effects. (8)

In this article, I use the experience of Indonesia to illustrate some scope conditions to the choice-theoretic models of decentralization that currently provide our foundational understanding of decentralization's political origins. I suggest that, where decentralization is enacted before transition elections have fleshed out the political landscape, the devolution of political, fiscal, or administrative authority is just as likely to be carried out in the absence of understanding of its political dynamics and effects as it is because of them. Incumbent politicians, facing significant uncertainty in upcoming elections, are open to suggestions about strategies that might allow them to retain office following a transition and gamble that decentralization might help them do this. In these conditions, intellectuals and bureaucrats in charge of designing the new system can exercise disproportionate influence over the decisions that politicians make; the former have a decided informational advantage over the latter.

I develop this argument using a narrative of the passage of regional autonomy laws under Habibie in post-Suharto Indonesia. It was the promise, rather than the reality, of decentralization that proved decisive in Indonesia. Indeed, legislators sent the first decentralization bill, which transferred nearly 2 million civil servants and responsibility for 60 percent of government spending from the central to city and municipal levels of government, to passage with few changes. (9) Given that all of them belonged to parties or other groups with powerful vested interests in maintaining unitary government and centralized politics, this decision is puzzling. I argue that it is the result of politicians' severe shortage of information with which to strategize, and their subsequent willingness to act as agenda-takers from a group of idealistic intellectuals eager to seize a window of opportunity.

The article proceeds as follows. First, I outline a number of ways in which scholars have conceptualized the role of political interests in explaining why political leaders choose to decentralize. Second, I suggest an alternative conceptualization of politicians' interests in transition periods and argue that decentralization may be enacted despite those interests rather than because of them. Third, I present an overview of the process of enacting Indonesia's decentralization, specify the shape of that decentralization, and detail how it succeeded even though it ultimately garnered little political benefit for the parties whose leaders pushed strongly for decentralization and cost them power and resources. Finally, I conclude by discussing the implications of Indonesia's decentralization for other new democracies.

Why Decentralization? Interests and State Reformation

I outline here the major conceptualizations of why decentralization was adopted and extrapolate some possible arguments for its adoption. Each of these arguments follows a pithy logic summarized nicely by Barbara Geddes: "Institutions reflect the interests of those who devise them." (10) This microeconomics-inspired take on institutional genesis is a compelling one, and it is not my intent here to dismantle it; what I argue here is that there are some important scope conditions to its applicability. One such condition is reliable information about what is likely to be gained politically from enacting such political reforms. To foreshadow my critique of these works as a whole, how do politicians facing an entirely new political game know how to define, let alone advance, their interests? From where do they acquire the information necessary to think about signaling their democratic credibility to voters? These are both crucial questions if we are to know how they decide which kinds of institutions to devise. Because these theoretical models assume perfect information, they tend to lead us to overlook the central role that uncertainty can play in highly fluid political environments.

The decision to enact wide-ranging decentralization of a mix of political, administrative, and fiscal authority is by nature a decision by politicians at the center to give away control of power and resources. Given the reasonable assumption that politicians would rather not dilute their own power, this is a puzzling choice in need of explanation. Catherine Boone argues that variations in the power and goals of rural elites shape the choices available to central rulers. (11) Her argument casts a valuable light on such choices by placing in it the context of demands from the countryside. It does rest, however, on a presumed demand from the periphery, a presumption that she notes is partly a function of the evolution of democratic politics. The further a country moves temporally forward from a democratic transition, the more likely both central and local elites are to be cognizant of both their own interests and the balance of power between them.

Kathleen O'Neill proposes an explanation for the choice to decentralize based on the subnational performance of political parties in past elections. (12) Parties whose subnational support appears more secure than their prospects in national elections are likely to support decentralization as a way of shaping institutions to increase their future electoral success. This argument, too, provides a powerful and elegant means for understanding the choice to decentralize in polities that are at least a few elections down the road from democratization, or in countries with a recent enough history of democracy for current party elites to know something about their subnational support. Again, though, in countries facing their first posttransition elections, parties in general, but especially new ones, are unlikely to possess such information.

Pauline Jones Luong moves away from a center-focused, choice-theoretic approach and suggests that, in several former Soviet republics, decentralization is less a choice for central actors and more a formal response to informal realities aimed at avoiding an appearance of weakness. Local elites with leverage effectively wrestle authority from the center in informal interactions, she suggests, and only later are those interactions codified formally. In essence, central politicians in Jones Luong's model are agenda-takers, salvaging what they can from a power imbalance in which "the state-building process is driven as much from below as it is from above." (13) In this framework, the interests of central elites are not taken as translating smoothly into institutional outcomes, a helpful step toward giving proper emphasis to factors other than elite interests. As compelling an argument as this is, it leaves us unable to account for situations in which local actors play little or no role in the decision to decentralize. In cases like Indonesia, where local actors hold little real power in the debate, we must look elsewhere.

Christopher Garman, Stephan Haggard, and Eliza Willis argue that the scope and pace of decentralization are shaped heavily by preexisting institutional, electoral, and party structures. (14) Rulers (and in all of their cases it is presidents who initiate decentralization) weigh the potential political gains to be had by providing greater efficiency of public goods against the costs of releasing power over both resources and agents. In transitional settings, the dynamics are likely to be somewhat different, but we can think productively about democratic efficiency gains as well as fiscal ones. Presidents in situations like Habibie faced in Indonesia, seeking public support for democratic reform measures, need some way to gauge whether voters will in fact respond favorably to various possible reforms. Absent past electoral cycle results to assess retrospectively--or public opinion polls to gauge prospective support for--such measures as decentralization, I would suggest that politicians face a distinct challenge. In transitional settings that approximate political tabulae rasae, there are few reliable options for obtaining a reliable sense of which reform issues are likely to play well with voters.

Another possibility is that autocratic rulers might decentralize selectively, allocating power and resources to regions where they know they have strong support, as a way of locking in some political gain even if decentralization is a second- or third-ranked strategy preference. David Samuels outlines this political logic in Brazil's military regime, showing how directed decentralization helped strengthen the regime's civilian supporters in smaller, more friendly states while limiting the power of opposition-friendly states. (15) While the type of regime differs in Samuels's analysis from the democracies that usually enact decentralization, his approach shares the presumption of minimal, or manageable, uncertainty that characterizes other studies.

Given the enthusiastic support for decentralization from international lending agencies such as the World Bank and International Monetary Fund (IMF), one might also expect to see it adopted by political elites eager to latch on to foreign aid as a way of protecting their political positions. (16) Critics of the Washington Consensus often point to foreign pressure to decentralize when explaining its overly hasty enactment and subsequent inability to do more than redistribute patronage vertically. In Indonesia, the IMF undoubtedly played a major role in the country's crisis-era political economy but exerted no pressure whatsoever on the question of regional autonomy. The team of academics Habibie appointed to rewrite the electoral law brought in external consultants to aid in that project but wrote the draft regional autonomy law in-house, free of pressure from the IMF and other foreign lenders. So, it is possible to set aside this explanation.

Finally, we might ask whether there was such a powerful public voice demanding autonomy for Indonesia's regions that both the president's office and the outgoing parliament realized the political weight of the issue independently and chose to support it because it was an obviously popular issue. This was not, and is not now, the case. Despite its arguably momentous implications for Indonesia's political future, it continues to garner relatively little public awareness. As already mentioned, elections crowded it almost completely out of public discourse by the time the bill came before the president, and it has never regained a central place in the political arena. That this would be so, however, was extremely uncertain in late 1998 and early 1999.

This group of theories to explain the political origins of decentralization is persuasive, and I do not mean to discount these arguments by questioning their applicability to the case of Indonesia and other "brand new" democracies. What I do hope to show is that there are some scope conditions for theories based on instrumentalist assumptions. Where politicians have at least some access to reliable information about how institutional choices will affect them, Geddes's elegant summary is often the best place to start in explaining those choices. This assumption sets the bar pretty high, however, in new democracies in which few or no actors have experience in democratic politics. The crux of the problem is that politicians often have little idea of what their interests vis-a-vis voters actually are beyond the rudimentary desire to stay in office, so that starting with interests misses the most important part of political processes whose outcome we want to explain.

Let us turn to Indonesia's transition to illustrate some of these problems. A draft law on decentralization arrived at the DPR from the president's office in late 1998. The president had no solid way of knowing how his party might fare in the regions, especially in the local elections that O'Neill illustrates were so important to Andean decentralization; it is therefore unlikely that his decision to support decentralization came out of a calculation that Golkar would benefit from it. Given the party's entrenched patronage networks, especially off Java on the outer islands, Habibie and the Golkar faction members in the DPR might have known they could count on a sizable vote share no matter what; that is to say, the party and Habibie would do well whether or not decentralization was enacted. But, given the immense public hostility showered on Golkar, the armed forces, Habibie himself, and most other organizations or officials associated with the Suharto era throughout 1998, the prospects for a strong Golkar electoral showing appeared very uncertain despite the prospects of playing the patronage card. As a result, Habibie and his party needed clear signs of reformism to present to voters.

At the time, the issue of regional autonomy attracted little attention from the public as well. With wrangling over the election laws recently concluded and the elections themselves looming (in addition to increasingly violent clashes between students and the military in late 1998), regional autonomy was an important issue almost exclusively for a handful of provinces such as Irian Jaya and East Timor, whose demands failed to generate broad support. Moreover, the bill arose not in response to stiff demands from local elites (e.g., governors, regional parliaments, mayors) such as took place, according to Jones Luong, in Kazakhstan, but in the near absence of real pressures from such elites. Indeed, what is remarkable about this early debate over regional autonomy in Indonesia is the almost complete lack of involvement of local actors. This was a centrally driven affair, one that moved through a political arena occupied by actors who, in the end, stood to gain little and to lose much if it became law. Finally, politicians had little to guide them in navigating their new roles in an emerging political system other than constant and fairly vague demands for reform.

This collection of events culminated in the passage of a law that seems to belie much of what we know about the origins of decentralization in other contexts. I want to suggest not that these explanations are incorrect, but simply that we need to recognize some important scope conditions for fruitful application of them. I return to the issue of scope conditions in the conclusion of this essay, but in the next section I move to my own explanation for the origins of decentralization in Indonesia and preelection new democracies in general. Uncertainty, rather than strategic calculation based on perfect or substantial information, takes center stage.

Selling Decentralization as a Signal of Reformist Credentials

The first several years following a democratic transition, especially in countries with no recent history of democracy, are ones of tremendous flux and political uncertainty. (17) As a result, authoritarian-era actors presenting themselves as good democrats often populate new democratic polities, creating two important dynamics that are important to understanding decentralization and other democratic reforms. First, political actors who thrived in authoritarian systems are unlikely to have a very solid grasp of the strategies required to succeed in democratic politics. Marek Kaminski, for instance, notes the amateurish fashion in which ex-Communist Party elites in Poland ran their election campaigns in the first free elections. (18) Anna Gryzmala-Busse illustrates the wide variation in the abilities of authoritarian-era communist parties to respond to democratic incentives as a function, in part, of their experience in negotiating with opposition groups before the transition. (19) Where we might intuitively have expected such seasoned politicians to adapt readily to a new environment, there is much comparative evidence to suggest that old habits die hard.

Second, the impetus to decentralization also parallels the choice of electoral systems in new democracies more than it actually does the choice to decentralize in "older" democracies. Where multiple elections have put flesh on actors' visions of the political landscape and their places in it, we can much more safely presume an ability to make choices based on reliable information. Absent those conditions, we ought to look to the origins of politicians' perceptions of the effects of their decisions about institutional choice and, in particular, to external sources of "expert" information. Given the unpredictability of institutional effects and of the voter salience of particular issues, such information may take on disproportionate importance.

Third, this lack of expertise in the democracy game manifests itself in important ways. The most important is an early strategy of endorsing major political changes--foremost among these is political decentralization--despite much understanding of their concrete implications and effects. Far from having access to full or even half-complete information about the consequences for themselves or their parties of such changes as decentralization or electoral system engineering, holdover politicians from the old regime often have very little information. Because of this, an approach based on either rationality or functionality assumptions is likely to read too much intent into the political endorsement of such changes from their effects. Politicians are likely to endorse political changes simply because they are associated with "reform" or "democracy." The real holders of information--intellectuals and other perceived experts--are more often the agenda-setters in new democracies, although international actors may also be important. These individuals can endow decentralization with the status of democratic reform, rendering it attractive to politicians eager to signal to voters their democratic bona fides.

Finally, political actors face an upcoming electoral battle in which they have to overcome a legacy effect as holdovers from the authoritarian era to demonstrate their democratic bona tides to the electorate. For these reasons, holdover politicians face a need to show a commitment to reform, and to do so early and decisively. The urgency of signaling reformist credentials at such moments in the transition process is likely subsequently to trump medium- and long-term instrumental strategies. In short, holdover politicians in such settings may face a forced move of accepting the advice of experts.

To state the argument briefly, agenda-setters with information can exert leverage over politicians who lack information simply by casting decentralization as a democratic reform. In the next section, I use the case of decentralization in Indonesia to illustrate how experts can exert inordinate influence over holdover politicians by pushing a specific reform issue on them. In this particular case, the members of Team 7 made a strong pitch for supporting decentralization to Indonesia's final authoritarian-era president and parliament, arguing that it would be a powerful issue on which to build reformist credentials. This turned out not to be the case, but absent more solid ways to assess voter preferences, it was convincing.

Adopting Decentralization in Indonesia

The push for decentralization in Indonesia began in 1998, during the months following President Suharto's resignation after thirty-two years in power. B. J. Habibie, the vice-president chosen by Suharto in January 1998, assumed the presidency after Suharto's resignation and began to prepare the country for free parliamentary elections, ultimately scheduled for June 1999. A presidential election, which Habibie hoped to win, was to take place in October 1999, in which the incoming parliament would elect the president as previous ones had done under Suharto in accordance with the constitution.

In addition to preparing the country for elections, Habibie also sought to lay a foundation for his own political survival in a post-authoritarian political system. His primary strategy was to couch himself as a reformer--to signal to the Indonesian electorate that he was more than Suharto's lackey. What was less clear, however, was how he might do that outside of shepherding Indonesia through elections. For the answers to that puzzle, Habibie turned to Team 7, the group of political scientists he had asked to advise the government on electoral reform. Team 7 used the opportunity afforded by this proximity to power to construct a reform agenda and to sell it to the president.

The team rewrote Indonesia's election laws in a handful of months under the leadership of Ryaas Rasyid, a US-educated political scientist who was serving as rector of the Indonesian Institute of Governmental Science. Working around their respective full-time academic positions, team members often clashed over ideas: "We put everything on the table and criticized each other. The only thing we didn't do is up-end the table." (20) Ultimately, Ryaas and the team settled on a variant of an open-list proportional representation system, reasoning that it struck a balance between encouraging maximal governability and diverse representation. Once that task was complete, several members of Team 7 contemplated a fruitful "next step." (21) Ryaas, with Andi Mallarangeng, proposed the decentralization of Indonesia's political system, since the mid-1960s one of the most highly centralized in the developing world. For Ryaas in particular, who was, like Mallarangeng, a native of South Sulawesi (one of the larger non-Java provinces in Indonesia), the issue of regional autonomy was a personal one and one on which he had written previously. (22)

Since Indonesia's independence, and particularly since the ascent of the New Order in 1966, Javanese domination of the outer islands has been a sore point in Indonesian politics. A rebellion on Sumatra in the late 1950s; lingering separatist movements in Aceh, in Irian Jaya (now renamed Papua), and in East Timor (now an independent state); and a more symbolic resentment of the early days of independence when Indonesia was formally a federalist system of the United States combined to produce both enthusiasm for, and wariness of, handing power from the central government to local levels of the state. Nonetheless, public sentiment in the months after Suharto's May 1998 resignation and a sudden surge in ethnic and religious conflict around Indonesia gave the issue of center-periphery relations some increased urgency and at least a balance between pro- and antidevolution voices.

Beginning in mid-1998, the issue of regional autonomy acquired temporary momentum and began to fill the void left by the resolution of the "when elections?" question. Even in West Java, the heavily populated province surrounding the special capital district of Jakarta, calls came for extensive power to be handed to the regions: "It has to happen, not just for East Timor, but all regions must be given autonomy before they demand it." (23) This warning, which began to echo in academic and government seminars as well as headlines, became more and more pointed as time went by: "Indonesia's sociopolitical 'flare-ups'--each and every one of them--have been caused by the regions never having been given full autonomy by the central government." (24) For a brief time, the issue of regional autonomy had a public voice.

Team 7 wrote a draft of the decentralization law in fewer than four months, incorporating elements of a similar bill under development at the Ministry of Home Affairs. The bill had first been proposed in 1992 by then Home Affairs Minister Rudini and, while this early version was similar to the sweeping Team 7 bill in its focus on subprovincial units, the version passed that year provided only for city and regency units to have some control over levying local taxes. (25) To assuage fears of creeping federalism, (26) and worries that devolving to provincial governments would make governors too powerful, the Team 7 bill proposed to transfer nearly 60 percent of government revenues, and 2 million civil servants directly to city- and municipal-level governments, bypassing the provinces completely and making walikotas (mayors) and their rural counterparts (bupatis) the new recipients of dramatically increased authority and funding. This single step, the authors hoped, would take the bite out of the counterargument that regional autonomy would make it impossible for the central government to control secession-minded governors. (27) It would be much harder, Team 7 reasoned, for 300 mayors and regents to act collectively than for 28 governors to do so. This objection to province-level decentralization was in sizable part a function of the political legacy of regional separatism in the 1950s, which ultimately led to the abolishment of Indonesia's federal constitution and its replacement with a unitary one. (28)

In late 1998, Ryaas presented a draft of what became Law 22 to President Habibie. (29) At first, Habibie was lukewarm, saying he would "prefer not to give any autonomy to the provinces." (30) Ultimately, Ryaas couched the proposal as a good reformist policy program that could win Habibie, and Golkar, seats in the upcoming elections. As Mallarangeng put it, "Ryaas was able to use this opportunity, where Habibie was very vulnerable. [Habibie] wanted to be seen as progressive, reformist." (31) Ryaas responded by appealing to the president's desire to portray himself as Indonesia's first reformist president, rather than its last authoritarian one, and to the rhetorical appeal of decentralization as another opposite of the Suharto-era pattern of governance. Ryaas banked on a kind of "negative learning" effect in which the president might be convinced that supporting the opposite of any Suharto-era trend in governance was a safe strategy for winning voter support. (32) In the end, Habibie was swayed by the promise of using his support for regional autonomy to put a reformist face on his interim presidency in hopes of winning the presidential election in October 1999. He endorsed it with almost no changes and promised to throw his weight behind it in parliament.

It is worth returning at this point to two possible explanations for the president's decision to support decentralization. O'Neill and Jones Luong suggest, respectively, that it may emerge from the calculation of future electoral advantage or from a desire to "save face" by codifying a de facto power grab by local elites. In Habibie's case, he had no past elections from which to gauge the electoral boost that decentralization might give him, and little apparent sense of how his party, Golkar, would fare in regional elections. On the one hand, Golkar was the one party in the country with strong patronage ties throughout the country, especially on the outer islands, and might be able to count on that support base. (33) On the other hand, the party remained tied to the New Order legacy of coercion and political exclusion; the tension between these two poles of political interest formed the basis for any decision Habibie--and Golkar parliamentarians as well--had to make.

Moreover, regional elites had little power, could make little in the way of forceful demands on the center, and in general played at most a secondary role in his decision. In short, there was little concrete information on which Habibie could base the decision. Ultimately, it rested on a perception of how regional autonomy might play with the electorate, and, in the absence of any polling on the issue, he had to turn to those he trusted to interpret a sporadic series of demands from various provinces. By emphasizing the reform aspect of regional autonomy, Ryaas and his team of experts were able to wrap the issue and the legislation in a political cachet that overcame Habibie's, and certainly the rest of his cabinet's, reservations.

Thus, with only minimal executive input, a bill designed to overhaul completely the structure of Indonesian political and administrative authority arrived at the DPR for consideration. At this point, decentralization ceased to be a purely administrative or intellectual exercise; it entered Indonesia's premiere political arena, the DPR, still populated by a holdover parliament that included appointed representatives from the army, navy, air force, and police in addition to the three sanctioned New Order parties. An observer might have expected decentralization to run into its stiffest opposition here: a parliament full of authoritarian-era legislators who stood to lose much by agreeing to hand over power that was effectively theirs (in a political system weighted heavily toward the legislative branch). (34)

The Debate Over Decentralization

The process by which the decentralization bill moved through the DPR is telling. In Ryaas's words, "The discussion in the DPR was peaceful, smooth, and productive." In part, the members of Team 7 who worked with members of parliament could rely on trust and good will generated during negotiations over the laws on elections and political parties. During these earlier negotiations, the team had compromised on several key issues, most important of which was the agreement to rely on a proportional system of representation rather than single-member district elections, which fell to party leaders' preferences. (35) In addition, Team 7 garnered much public exposure for their central role in redesigning Indonesia's political institutions: "In those days, two or three times a day I was on TV. For a time I was the center of focus; thank God I never betrayed them [members of the DPR]." (36)

Nonetheless, various factions of the DPR and representatives of other parts of the government weighed in while the final shape of both Laws 22 and 25 underwent consideration. This discussion took a public form in Indonesia's already widespread print media, making it possible to some degree to chart the interests of, and the intended signals to be sent by, these parties and political groups.

The three parties of the New Order--the ruling Golkar; the Indonesian Democratic (Partai Demokrasi Indonesia, or PDI, the official secular party); and United Development (Partai Persatuan Pembangunan, or PPP, the Islamist opposition party)--remained highly centralized, with nearly all important decisions taking place in Jakarta. Also important in this last New Order parliament was the military faction, which should have been opposed to decentralization at least as much as the parties were. Finally, bureaucrats in the Ministries of Finance and Home Affairs also made numerous public pronouncements about the ongoing revisions to the bills, especially Law 25 on fiscal issues. Their public statements capture most of the central concerns that motivated debate over regional autonomy. In this section, I present the central themes from these statements to flesh out where the parties and other important political actors stood on both the issue and the structure of decentralization.

Habibie's office, and other agencies and media he could count on, continued after presenting the bill to the DPR to refer to its reform-enhancing effects and deemphasized the power-shifting aspects. Republika, a dally newspaper tied closely to the All-indonesian Association of Muslim Intellectuals (Ikatan Cendakiawan Muslim se-Indonesia, or ICMI), of which Habibie was the first leader, presented a supportive public forum for his office's message, calling decentralization representative of "a more democratic spirit as demanded by the reformasi movement." (37) His minister of finance, Bambang Subianto, and other ministry officials made similar proclamations about both bills, stressing the "transparency" and "public participation" gains to be had from the intergovernmental fiscal balance bill in particular. (38)

Once Habibie had endorsed the decentralization bills himself, Golkar faction members eventually fell into line as well. Nonetheless, as the ruling party, and one ostensibly committed to a unitary state, its leaders expressed concern about weakening the coherence of the country: "Granting autonomy must not cause the disintegration of the nation." Even in doing so, these objections were more mechanical than absolute, calling for a focus on decentralization rather than deconcentration so as to enhance "regional-level democracy" while not weakening the state. (39)

Indonesia's Muslim elites had a different set of concerns than the unitary-state ones central to Golkar party leaders. PPP's leaders, because their party was the official Islamic party of the New Order polity, warned of the possibility of local religious majorities' legislating or decreeing official discrimination against local minorities, describing the potential for "Hindu Shari' a." (40) A similar message issued from the Majelis Ulama Indonesia, who warned that the central government must maintain regulatory authority over religious issues. (41) This warning led to the exclusion of religious authority from the final menu of additional responsibilities handed to local governments. In addition to Bali, a number of provinces in eastern Indonesia are either evenly balanced between Muslims and Christians or are Christian dominant, which could have raised the prospect of non-Muslim local religious control in a country composed overwhelmingly of Muslims. Nonetheless, once that issue was resolved, PPP faction members began to use decentralization as another shorthand signal of reformism: "With autonomy we develop and strengthen the pillars of democracy to strengthen the unity and integrity of the nation." (42)

In addition, two groups with concrete experience in government, the army and navy, weighed in on the bill. (43) The army's main concern lay in a long-standing territorial role in which it held primary responsibility for domestic order; the navy's lay in jurisdictional questions over protecting Indonesia's shoreline and ocean territories, which are substantial. Team 7 agreed to leave marine authority, like religious affairs, within the purview of the central government. Army faction spokesmen in the DPR actually pushed a new line of defense for regional autonomy: that it would liberate regional business potential from heavy-handed central regulation and from "dependence on handouts" from the central government. (44) Ultimately, ABRI (the collective name for Indonesia's four armed services), like all three party factions, threw their support behind regional autonomy after some modest modifications to the fiscal balance bill, and both bills passed into law in May 1999.

In terms of revision from the original versions drafted by Team 7, only the fiscal balance underwent any substantive changes. This was in large part because, not being economists, the members of the team left the early draft somewhat vague on actual procedures for assessing the amounts to be transferred from the central government to regencies and municipalities. (45) For them, the primary task was to establish political decentralization, and then writing a law from the initial draft of the fiscal balance bill was a more collective endeavor. Nonetheless, after consultation with various nongovernmental economic organizations (like the Econit Advisory Group) and through extensive involvement of the Ministries of Finance and Home Affairs, issues such as subsidy transfers to poorer regions were resolved on April 23, 1999. (46)

Again, it is worth revisiting a choice-theoretic approach to explain the quick passage of the regional autonomy laws. Given the speed with which they moved through consideration to passage, we might infer that a majority of some kind--whether parties or military factions--saw decentralization as being in their concrete political interests. This appeared not to be the case after the 1999 elections, in which decentralization took a back seat to most other parts of the Reformasi agenda. On the contrary, as I discuss below, the government's preparation for enacting the bill in 1999 and 2000 catalyzed opposition by parties, the military, and many ministries, who suddenly realized how much regional autonomy would directly threaten their interests. That is, once these parties realized how their interests would be affected, they responded accordingly. By then, however, the decentralization law had moved through parliament, as it had through the president's office, despite rather than because of legislators' concrete interests.

Seven academics wrote Indonesia's regional autonomy law and sold it to an outgoing (and would-be incoming) government as a major democratic reform, almost sneaking it into law between the reform of Indonesia's election laws and preparations for the election itself. B. J. Habibie, Suharto's hand-picked presidential successor, along with various military factions and the largest parties in Indonesia's last New Order parliament, (47) supported decentralization strongly and pushed it through with almost no opposition, in large part because they saw it as part of the Reformasi package that would mark them as reformists in advance of the 1999 elections. Just as importantly, at the time, few had much sense of the likely implications of giving so much administrative and fiscal authority to lower levels of government.

This changed with the creation of the Ministry of Regional Autonomy by President Abdurrahman Wahid in 2000. Wahid asked Ryaas, who had headed the directorate of regional autonomy within the Ministry of Home Affairs, to serve as state minister for regional autonomy. He accepted and appointed Mallarangeng his deputy. It was under this new ministry that the real work of preparing Indonesia for devolution began, and it was with this preparation that political actors across the government began to grasp the law's implications. As Ryaas and Mallarangeng started to carry out the administrative changes necessary to hand over power, personnel, and revenues to municipalities and cities across the country, their opponents in Jakarta became more numerous and more vocal.

The Ministry of Home Affairs, which had seen its institutional and jurisdictional control of regional autonomy vanish by presidential decree, led the opposition to the new ministry. It was soon joined by the forestry and mining ministries, whose upper echelons realized that their rent-granting monopolies were about to disappear. Soon, the military too realized that its longtime ability to dominate regional government would come to an end with the direct election of local officials. Finally, the two largest party factions in the new, democratically elected parliament that took office in October 1999--Golkar and the Indonesian Democratic Struggle Party, or PDIP--realized that decentralization posed a threat to their highly centralized control of candidate nomination and patronage networks. Despite their active mobilization to stop the law's enactment, or at least dilute it, Indonesia's "Big Bang" decentralization was carried out beginning in January 2001 and has substantially changed the political and institutional landscape of the world's fourth largest country as a result. (48)

The fact that all the major actors who supported decentralization in principle opposed it in practice, to my mind, has much to do with their late realization of its impact on their interests as well as the realization that it carried limited electoral benefit. Moreover, it demonstrates clearly how these actors came collectively to understand their own interests, and how various institutional changes affected those interests, even after only one electoral cycle. The same was true of Indonesia's electoral system, written in 1998 by Team 7 just before the group turned to drafting a decentralization law. By 2002, the major parties in the DPR had a keen sense of how to modify the electoral system in such a way as to maximize their own control of the process. The result was an officially open-list proportional system that in fact was far from open. Inserted deep in the 2003 Law on Elections (see Article 107) was a clause that in practical terms closed the list again and restored party leaders' capacity to fill seats despite the "openness" of the list. (49)

Conclusion: Indonesian Decentralization

Indonesia's experience adopting decentralization does not fit easily into a choice-theoretic explanation. I suggested here that the reason hinges on the magnitude of political uncertainty. Theories that rely on assumptions about instrumental behavior rely on a further presumption that these actors know what their concrete interests are and can therefore assess long-term versus short-term political trade-offs. Where information is highly uncertain, I argue, experts gain influence over politicians. These experts come to government service with their own commitments and, under the right conditions, can use the opportunity afforded by their status to suggest strategies for politicians. As a result, the search for the origins of decentralization and, importantly, other components of democratic reform is best aimed at these advisory groups.

To the extent that I believe Indonesia can shed light on the decision to decentralize, or indeed to adopt other reforms such as electoral system change, it is by virtue of outlining systematically what Indonesian decentralization is a "case of." I argue that it is a case of codifying dramatic institutional changes in the interim between authoritarian breakdown and the election of an incoming democratic government. It is, in addition, a polity with such a distant democratic past (1949-1957 in Indonesia) that no currently active political elites had experience in navigating politics outside of relying on networks emanating from an authoritarian regime. During such periods of immense fluidity and uncertainty, we can reasonably expect actors to have at best an incomplete sense of how democratic institutions affect their interests, and how major changes like decentralization might impact their positions. In short, Indonesia at the time that decentralization was adopted, reflects conditions of almost maximal political uncertainty for incumbent politicians and, more importantly, of extremely limited concrete information with which those incumbents can act to preserve their political careers.

Where these conditions hold, it is plausible to expect appointed advisory groups like Team 7 to be the source of such important changes, and it is in these groups that we are likely to find the most compelling explanations for institutional innovation in preelection transitional democracies. The primary comparative proposition I draw from Indonesia's experience in passing the regional autonomy laws is the following: where decentralization comes for adoption before free elections have established some baseline for party power in regions, and in polities with no recent history of democratic party politics to guide current elites, expert advisory groups are likely to exercise preeminent authority. It is possible that these groups might consist of external advisers rather than domestic ones, depending on the perceived salience of decentralization for external lenders. In the end, however, the crucial point to take away is that "experts" can trump political interests given conditions of high uncertainty. Where those conditions obtain, our analysis of key moments in democratic transitions ought to focus away from politicians and toward those from whom they "learn" what their interests are. (50)

Notes

Thanks to Jason Brownlee, Ketut Putra Erawan, Bill Liddle, Rizal Panggabean, Dan Slater, and David Waldner for comments on earlier versions of this article. Special thanks go to Andi Mallarangeng, former deputy minister, and to Ryaas Rasyid, minister of regional autonomy, who gave generously of their time during a busy election season, and to Kevin O'Rourke, who as always was generous with his time in sharing his encyclopedic knowledge of Indonesia's political elite.

(1.) They were Hamid Awaluddin, Djohennansyah Djohan, Afan Gaffar, Andi Mallarangeng, Ryaas Rasyid, Ramlan Surbakti, and Anas Urbaningrum. Team 7 actually predated Suharto's fall: Hartono, Suharto's minister of home affairs, brought the members together under Ryaas Rasyid's leadership to design and coordinate political reforms. Kevin O'Rourke, "Interview with Ryaas Rasyid," VanZorge Report, November 1999. Thanks to Kevin O'Rourke and Bill Liddle for extensive conversations about this period.

(2.) Author's interview with Andi Mallarangeng, October 29, 2003; see also Nadirsyah Hosen, "Indonesian Political Laws in Habibie Era: Between Political Struggle and Law Reform," Nordic Journal of International Law 72, no. 4 (2003): 490-491.

(3.) Republic of Indonesia, Undang-Undang Nomor 22 Tahun 1999 Tentang Pemerintahan Daerah (Law no. 22, 1999, on regional government), and Undang-Undang 25 Tahun 1999 Tentang Perimbangan Keuangan Antara Pemerintah Pusat Dan Daerah (Law no. 25, 1999, on fiscal relations between central and regional government).

(4.) Under Suharto's New Order, electoral interference was accompanied by a process by which members of parliament were appointed rather than directly elected. It was a sort of closed-list proportional system in which the ruling Golkar allowed only two other parties to contest elections and then had to approve the candidates they could name to parliament.

(5.) Indeed, their pressures were successful in another arena of institutional reform: electoral laws. Party leaders were successful at pushing for a variant on proportional representation that looked to provide candidates with a meaningful incentive to build local support but that in actuality left most seat-allocating power in the hands of Jakarta-based party leaders.

(6.) My research in late 2003 and early 2004, three years after implementation and just before Indonesia's second posttransition parliamentary elections, revealed little concern for regional autonomy as a national issue, even among local party leaders. This despite the apparent determination of President Megawati Sukarnoputri to roll back significant portions of the regional autonomy law and to reclaim much authority for the executive branch. Indeed, the two parties that made regional autonomy a central issue for their platforms in the 2004 elections--both led by nationally known political personalities, one of whom was Ryaas Rasyid--captured less than 2 percent of the total vote and only 4 of 550 seats in the incoming DPR. Author's interviews in Jakarta, Yogyakarta, and Denpasar, October-November 2003.

(7.) John Cohen and Stephen Peterson, "Administrative Decentralization: A New Framework for Improved Governance, Accountability, and Performance," Working Paper, Harvard Institute for International Development, 1997.

(8.) See Catherine Boone, "Decentralization as Political Strategy in West Africa," Comparative Political Studies 36, no. 4 (May 2003): 355-380; Kathleen O'Neill, "Decentralization as An Electoral Strategy," Comparative Political Studies 36, no. 9 (November 2003): 1068-1091; Kent Eaton, "Designing Subnational Institutions: Regional and Municipal Reforms in Postauthoritarian Chile," Comparative Political Studies 37, no. 2 (March 2004): 218-244; and Christopher Garman, Stephan Haggard, and Eliza Willis, "Fiscal Decentralization: A Political Theory with Latin American Cases," Worm Politics 53, no. 2 (January 2001): 205-236. One notable exception is Pauline Jones Luong, "Economic 'Decentralization' in Kazakhstan: Causes and Consequences," in Pauline Jones Luong, ed., The Transformation of Central Asia: States and Societies from Soviet Rule to Independence (Ithaca: Cornell University Press, 2003).

(9.) Law No. 25, the partner bill that specified fiscal balances between center and region, underwent more substantial modification, in part because none of its original authors in Team 7 were trained economists. For discussions of the debates over this second law and clear statements from most of the major parties to the negotiations, see Republika, February 12, 1999, March 16, 1999, and April 7, 1999; Suara Karya, February 20, 1999, and February 23, 1999; Kompas, February 22, 1999, March 31, 1999, April 22, 1999, and July 30, 1999; Bisnis Indonesia, February 22, 1999; Suara Pembaruan, April 8, 1999; and Media Indonesia, April 20, 1999, and April 24, 1999.

(10.) Barbara Geddes, "A Comparative Perspective on the Leninist Legacy in Eastern Europe," Comparative Political Studies 28, no. 2 (July 1995): 239.

(11.) Boone, "Decentralization," footnote 3.

(12.) O'Neill, "Decentralization," footnote 3.

(13.) Pauline Jones Luong, "Economic 'Decentralization' in Kazakhstan," 182-212.

(14.) Garman, Haggard, and Willis, "Fiscal Decentralization." See especially p. 209 for a discussion of the potential costs and payoffs for presidents who initiate decentralization.

(15.) David Samuels, "Federalism and Democratic Transitions: The "New" Politics of the Governors in Brazil," Publius 30, no. 2 (Spring 2000): 52-54.

(16.) Vedi Hadiz, "Reorganizing Political Power in Indonesia: A Reconsideration of So-called 'Democratic Transitions'," Pacific Review 16, no. 4 (2003).

(17.) By this I mean countries in which democratic government, if it existed previously, was in the sufficiently distant past that generational change has removed from active politics actors and organizations with direct experience in democracy. This is one scope condition of the argument I develop here.

(18.) Marek M. Kaminski, "How Communism Could Have Been Saved: Formal Analysis of Electoral Bargaining in Poland in 1989," Public Choice 98, no. 1-2 (January 1999): 83-109.

(19.) Anna Gryzmala-Busse, Redeeming the Communist Past (Cambridge: Cambridge University Press, 2002).

(20.) Andi Mallarangeng, quoted in Far Eastern Economic Review, September 17, 1998.

(21.) Author's interview with Andi Mallarangeng, April 8, 2004.

(22.) M. Ryaas Rasyid, "Regional Responses to Central Government Authority: A Comparative Study of South Sulawesi and Aceh" (masters thesis, Northern Illinois University, 1988); M. Ryaas Rasyid, "State Formation, Party System, and the Prospect for Democracy in Indonesia: The Case of Golongan Karya, 1967-1993" (PhD diss., University of Hawai'i-Manoa, 1994).

(23.) Quoted in Suara Karya, August, 25, 1998, emphasis added.

(24.) Merdeka, September 30, 1998, emphasis added.

(25.) See Republic of Indonesia, Peraturan Pemerintah Republik Indonesia Nomor 45 Tahun 1992 Tentang Penyelenggaraan Otonomi Daerah Dengan Titik Berat Pada Daerah Tingkat II (Government regulation no. 45, 1992, on the enactment of regional autonomy with centrality for level II regions).

(26.) For example, see the overview of this debate in Kompas, October 19, 1998.

(27.) This rapid drafting process also garnered criticism from NGO and civil society groups in Indonesia, who complained that only governors and bupatis had been consulted at the subnational level. See Down to Earth (August 2000), http:dte.gn.apc.org/46Suh.htm (accessed August 9, 2004). For their part, Team 7 members replied that NGO leaders, when asked to provide consultation, either failed to appear at scheduled meetings or had not read background materials pertinent to consultation meetings.

(28.) There is a broader point to be observed from this pragmatism on the part of the members of Team 7. I do not mean to suggest that they wrote a wholly idealistically driven law and sold it willy-nilly to Habibie and the DPR; they did not. They worked pragmatically to produce a regional autonomy law that (1) would be more palatable to nationalists than federalism, and (2) tracked administratively with a much more limited bill under construction at the Ministry of Home Affairs, whose approval the bill would need.

(29.) Team 7 later wrote a draft of the law on intergovernmental fiscal relations that underwent much more substantial revision after consultation with the Ministries of Finance and Home Affairs and the party factions in the DPR. I owe thanks to an anonymous reviewer for pushing me to clarify the process by which Law 25 was produced and revised.

(30.) Author's interview with Ryaas Rasyid, April 12, 2004. It was unclear whether Habibie meant that he would rather pass on any form of regional autonomy or simply one that devolved power to provincial-level governments.

(31.) Author's interview with Andi Mallarangeng, April 8, 2004.

(32.) Thanks to Rizal Panggabean for describing it in these terms.

(33.) Indeed, there is some evidence to support this. Amid dramatic changes in the electoral performance of PDI-P, Partai Demokrat, and Partai Keadilan Sejahtera, among others, Golkar captured almost exactly the same share of votes nationally in 1999 and 2004--roughly 22 percent. Habibie could not know this in 1998, of course.

(34.) The constitution gave the parliament the authority to elect the president. While under Suharto this election became more ritual than anything else, it gave the party blocs the dominant role in the post-Suharto political system.

(35.) This preference actually ran counter to the interests of Golkar, at least, but remained the central position of all three official parties in the last Suharto-appointed parliament. Party leaders apparently failed to realize that their extensive regional ties would provide them with a first-comer's advantage in fielding good candidates for single-member districts. Thanks to Bill Liddle for bringing this to my attention.

(36.) Author's interview with Ryaas Rasyid, April 12, 2004.

(37.) Republika, February 12, 1999.

(38.) Quoted in Bisnis Indonesia, February 20, 1999. Subianto was one of the appointees of Habibie himself in the aftermath of the resignations that accompanied Suharto's. A long-time ministry official without a political clique of his own to protect, Subianto largely served simply as Habibie's voice.

(39.) Functional Development (Golkar) faction spokesman J. M. Nailiu, quoted in Suara Karya, February 20, 1999.

(40.) Shari'a is the Arabic word for Islamic law. "Hindu Shari'a" refers to the Hindu majority on the island of Bali, whose population is only roughly 8 percent Muslim, as opposed to the 90 percent majority nationwide. The fear on Java was that Hindus on Bali might use regional autonomy to enact religious laws discriminating against the small Muslim minority on the island.

(41.) Kompas, March 31, 1999.

(42.) PPP fraction spokesman Chozin Chumaidy, quoted in Suara Karya, February 20, 1999. Suara Karya was closely aligned to Golkar at this point, so it is somewhat surprising to see such contradictory language in an article that before this quote lays out the Golkar fraction's concerns about national disintegration.

(43.) Ibid.

(44.) ABRI fraction spokesman Namoeri Anoem, quoted in Suara Karya, February 23, 1999.

(45.) See Jakarta Post, February 23, 1999, and Bisnis Indonesia, March 2, 1999.

(46.) For discussions of the final negotiations on transfers and subsidies, see Suara Karya, March 16, 1999, and Suara Pembaruan, April 8, 1999. On the final agreement to vote on the law, see Kompas, April 24, 1999, and Media Indonesia, April 24, 1999.

(47.) The New Order parliament reserved large blocks of seats for representatives from all branches of the Indonesian armed forces and for the police.

(48.) The effort to water down regional autonomy continued throughout the Megawati presidency. One of only four bills under consideration by the outgoing parliament in late 2003 (which in the next parliament would reflect a considerably lower number of total seats for the Golkar-PDIP axis) was one that would return many areas of authority to the central government and make mayors and regents formal representatives of the president. I was accidentally invited to a "discussion session" on October 29, 2003, in which a draft of this "recentralization" law was presented by the Ministry of Home Affairs to a handful of regional autonomy specialists, including Ryaas Rasyid and Andi Mallarangeng, well before its existence became public. The debate over the bill at this meeting was, by Indonesian standards, extraordinarily raucous, and the cross-table yelling subsided only after sunset when, because the meeting took place during Ramadan, participants could break their fast and eat dinner. The bill was later withdrawn by the Megawati administration.

(49.) Article 107 established a quota for automatic election in any given district: the number of valid votes divided by the number of seats in that district. Given that there were often eight or ten times as many candidates as seats available, it is unsurprising that only about 5 percent of the incoming DPR was elected by surpassing the quota; the rest were selected by party leaders. Hadar Gumay of the Center for Electoral Reform confirmed that Article 107 was crafted and forced into the bill by the leaderships of Golkar and PDIP. Author's interview with Hadar Gumay, October 31, 2003.

(50.) Later rounds of bargaining and conflict over institutional changes may, and probably will, come to reflect dynamics that lean closer to full-information models as actors acquire greater certainty about how issues affect their political power and how they play with voters.

Benjamin Smith is assistant professor of political science and Asian studies at the University of Florida.