The Political Economy of Policy Reform: Insights from Southeast Asia

Economists broadly agree on many key economic policy issues, but economics as a discipline has provided much less guidance on why and how economic policy reform occurs and how to develop institutional mechanisms that enable governments to adopt “good” economic policy. Political scientists are adept at identifying coalitions, constituencies, institutions, and interest groups, but they less commonly examine the implications for economic policy. Thus, work at the intersection between economics and politics—of why and how policy reform takes place—remains relatively unexplored territory. This is especially so in developing countries where political processes are more personalistic, institutions often less well established, outcomes more fluid, and the detailed case study literature on economic policy making still in its infancy. This paper provides an analytical survey of economic policy reform in Southeast Asia. It ranges across the major policy U-turns and the incremental reforms, with special reference to macroeconomic management and trade policy. On the basis of several case studies and set against the broader international literature, we advance nine conclusions on the political economy of reform.


I. Introduction
Although it may not seem obvious to non-economists, economists broadly agree on many key economic policy issues. But economics as a discipline has provided much less guidance on why and how economic policy reform occurs and how to develop institutional mechanisms that enable governments to adopt "good" economic policy. Political scientists are adept at identifying coalitions, constituencies, institutions, and interest groups, but they less commonly examine the implications for economic policy. Thus, work at the intersection between economics and politics, of why and how policy reform takes place, remains relatively unexplored territory.
There is no generally accepted template, much less a "rule book" for how to engineer successful policy reform. This is especially so in developing countries where political processes are more personalistic, institutions often less well established, outcomes more fluid, and the detailed case study literature on economic policy making still in its infancy. We therefore need more case study evidence, of both success and failure, to understand why and how successful reform occurs.
Southeast Asia offers a fascinating opportunity for social science researchers interested in these issues. The economic performance of most of the major economies for most of the period since 1970 has been significantly better than the developing world average. But there is much diversity in the record of growth and reform. The reform experience ranges across the major policy U-turns and incremental reforms, successes and failures, and the macroeconomy and the sectors, with both international and domestic factors playing a role. This diversity of the region, reflected also in levels of development and in political and institutional structures, both cautions against generalization but also adds to the richness of the subject matter.
This paper offers an analytical survey of the evidence on economic policy reform in Southeast Asia. To narrow down the topic to manageable, paper-length proportions, we focus primarily on macroeconomic and trade policy reform in three of the lower-middle income economies-Indonesia, the Philippines, and Viet Nam. Section II introduces the issues and provides some country and institutional context. Section III dissects a series of reform episodes in these countries. Section IV then draws some broader lessons and implications.

A.
Some Definitions I define "reform" for these purposes as a durable and significant policy change that improves aggregate socioeconomic welfare, consistent also with an objective function that recognizes distributional and environmental considerations. The underlying rationale is concern for general welfare, the public interest, rather than particular vested interests. Economists have typically defined reform as measures that increase productivity and growth rates, but these goals could obviously be redefined to encompass a broader set of non-economic objectives. In addition to particular measures, reforms can also be about policy-making processes, for example greater transparency in policy making such as when firms claiming special assistance have to submit to a process of public scrutiny and justification.
Obviously, not all policy changes would meet this definition of reform. For example, a redistribution program would not unless it could be shown that this program resulted in increased productivity or met more widely accepted social objectives (e.g., social stability). Similarly, programs that are essentially windowdressing exercises such as anti-corruption campaigns introduced by a deeply corrupt regime or "one-stop-service" investment programs without significant bureaucratic B.

Drivers of Reform
The literature typically identifies the key drivers of reform as a mixture of factors, including necessity, the triumph of ideas, and the conjunction of reformoriented political leadership aided by technocratic advisers. Typically, several of these factors are present in the case of successful reforms. An understanding of the drivers of reform in turn requires identification of the key policy actors. We summarize here some salient points in the literature.
First, the "crisis hypothesis" as a reform driver was a key conclusion of the comparative study of Lal and Myint (1996, 288), who concluded that "Turning points (in economic policy) are invariably associated with macroeconomic crises." The trigger could also be some other major event such as a military defeat (or threat), the cessation of external support, or a natural disaster. The underlying hypothesis is that reform is a difficult process, and societies have a natural ("Olsonian") tendency to become sclerotic. A crisis may be helpful in persuading the community that the current order is unacceptable and requires change. Political leadership may be emboldened and willing to tackle difficult issues. For example, Bardhan (1998) draws attention to India's 1991 balance of payments crisis as a trigger for reform, enabling the government to push aside the formidable vested interests that had built up around the post-independence dirigiste regimes, both financial and ideological.
Obviously, however, not all crises trigger major reforms. Instead they may result in failed states or at least an inability to seize the opportunity to reform. That is, the hypothesis only works in certain circumstances. For example, the collapse of the communist regimes in the former Soviet Union and Eastern Europe led to a deep economic contraction, and several countries did not regain pre-crisis living standards for over a decade or two (Pomfret 2002). Moreover, deep crises may result in a political and institutional vacuum and incapacity to undertake effective reform. In the transition to a new, perhaps democratic environment, power is generally diffused, structures are established to deliberately weaken the state, and policy lacks credibility. Indonesia in 1998 and the Philippines in 1986 are relatively mild examples of these twin crises, although in Indonesia, economic recovery occurred surprisingly swiftly given the depth of the crisis (Aswicahyono, Bird, and Hill 2009).
Second, effective reform requires a coherent intellectual agenda, an analysis of what needs to be done, how, and in what sequence. Thus, ideas are a central prerequisite (Krueger 2007). These ideas spring from a variety of sources but are most commonly associated with think tanks and economics faculties of leading universities. A common characterization is that of a key group of technocrats receiving training from leading universities abroad-Chile's "Chicago Boys" and Indonesia's "Berkeley Mafia" are oft-cited examples-but the channels of influence are in reality much broader. 2 In some cases, the reform agenda is formulated and driven by observing success abroad, most commonly in the neighborhood. The bureaucracy may also become a driver where there is a realization that the system being administered has become increasingly dysfunctional.
Technocrats are generally politically powerless. To translate their ideas into policy, they need to convince political leaders of the case for reform and to work with them in implementing reform. That leader does not need to be a technical economist-in fact such cases are rare-but at least a person open to persuasion and able to grasp intuitively, if not technically, the case for reform. Obviously, there has to be a close working relationship and a sense of trust between these leaders and their technocrats. In centralized authoritarian regimes, the key is the technocrats having access to the president or ruling party. Here, the technocrats may engage in what Soesastro (1989) termed "low politics" in the Indonesian case. That is, the technocrats seized the opportunity created by the sudden decline in international oil prices in the early 1980s to persuade then President Soeharto to implement farreaching reforms but without having to engage in large-scale, high-profile public or parliamentary persuasion.
However, approaches to reform change, sometimes radically, in democratic regimes. Politicians have to persuade an electorate, while technocrats have to engage in the public discourse, persuade political and opinion leaders, and ameliorate or buy off potential opposition (and losers). There is higher potential for policy change in regimes with low dispersal of power, in the words of MacIntyre (2003). That is, policy reform is likely to be more difficult where power is more diffused and there are more veto players present. It might be argued that while reform is slower under a democratic regime, it is likely to be more durable since the reform process will be consensus-driven, with greater attention paid to potential losers. Nevertheless, in presidential regimes where both the executive and senior echelons of the bureaucracy are replaced, elections may result in significant policy change. Two examples since the mid-1990s where reformists were followed by democratically elected reform skeptics are Colombia and the Philippines. In both cases, the growth momentum decelerated. 3 Third, convincing politicians of the case for reform is a key challenge. Political leaders by definition have short-term horizons and a political predisposition to favor a particular constituency. Arguing the case for reform when the benefits may be uncertain and long term in nature, and the short-term costs potentially high, is perhaps the most important challenge in the policy reform agenda. The reforms need to produce a dividend as quickly as possible. Where painful decisions are needed, external support may occasionally be useful. Technocrats themselves may attempt to pick a political winner and lend their credibility to a particular candidate. There may also be scope for institutionally embedding reform momentum. Bates (1994, p. 30) for example argues for ". . . creating institutions that possess the power to commit (politicians) to collectively rational strategies." Examples include measures that impose fiscal discipline on a government or create "agencies of restraint" staffed by professionals and with public reporting responsibilities (e.g., a productivity or competition commission).
Fourth, as noted, the bureaucracy is a key actor. It may range from being a passive bystander to an active player in a negative or positive sense. The literature generally makes two presumptions concerning their role. The first is a division between the key economic policy agencies such as finance ministries and central banks, which are more likely to be staffed by economists and to favor "orthodox" policies, and line and sector ministries which are more likely to favor and be captured by sector interests. The second is the notion that much of the bureaucracy, the latter groups especially, will be reluctant reformers since most but not all reforms will reduce their discretionary authority and hence the scope for rent seeking. Reform outcomes will therefore depend on the relative strength of these contending groups. More generally, outcomes will be shaped by the relative power of the executive, the legislature, and the bureaucracy; and the scope of the reformers to variously persuade, co-opt, or bypass bureaucratic resistance.
Fifth, there are various conjectures concerning the impact of external actors and factors. Two have already been identified: crises, some of which are exogenous (in the form of negative external shocks) and which have unpredictable effects; and ideas, many of which originate from abroad. Other foreign influences may also shape the process. There is the demonstration effect of successful reforming economies, especially if they are located nearby. This is the "competitive liberalization" thesis referred to by Indonesia's Minister for Tourism and Creative Economy Mari Pangestu (2012) and others. The intellectual ascendancy of openness as an engine of economic progress is highlighted in several country studies of trade liberalization (Rajapatirana 2001). Foreign investors have become more interested in global economic integration, and therefore their earlier interest in establishing "tariff factories" behind high protective barriers in developing countries has waned (Bhagwati 2002).
The evidence on donor (particularly, the IMF and the World Bank) conditionality is mixed. Jeffrey Sachs (1994, 504) has opined that "Countries cannot be transformed without the generous and farsighted involvement of the international community." A large literature of course argues the contrary case (see, for example, Easterly 2006), that aid encourages the recipient countries to postpone difficult policy reforms. International agencies can play an effective role if there is a domestic interest in-and will for-reform (Krueger and Rajapatirana 1999). However, in the absence of these factors, externally-mandated reform attracts domestic opprobrium, implementation is likely to be spasmodic, and the reforms will therefore generally not be durable. These arguments are also consistent with the cross-country econometric evidence that finds aid contributes to growth only when "good policies" are present (Burnside and Dollar 2000).
Sixth, the more successful reforms are invariably comprehensive. Political constraints may in reality result in piecemeal reform. But the danger is that significant gaps in the reform agenda may undermine the entire process. The literature on the interaction between macroeconomic and trade policy illustrates this issue and also provides an intellectual rationale for the sequencing of reforms. As Rajapatirana (2001) and other analysts of successful trade liberalization point out, a willingness to allow a large depreciation boosts the competitiveness of tradable goods industries and facilitates a lowering of protection. Krueger (1978, p. 231) goes further, arguing that the ". . .failure to devalue by a sufficient margin will prevent sustained liberalization." Moreover, a ". . . realistic real exchange rate (is) an essential condition for sustained liberalization." In a similar vein, Pinera (1994, p. 228) warns against partial reforms on the basis of the Chilean experience: "It is no use freeing trade and opening up the capital markets if one is going to leave the labor markets untouched."

III. Southeast Asian Case Studies
In this section, we summarize four major Southeast Asian policy reform episodes. It needs to be acknowledged immediately that there are two forms of sample selection bias in these case studies. First, their selection is inevitably arbitrary, based on documented research and my own research interests. Moreover, the samples selected focus mainly on success stories, working on the principle that economic policy can fail for any number of reasons, but success is more elusive and therefore needs to be investigated. Second, to the extent that the case studies focus primarily on changes in policy direction, the three more advanced economies in the region-Malaysia, Singapore, and Thailand-are underrepresented since they have had much more consistent policy regimes since the 1970s, and therefore there has been less need to undertake far-reaching changes in the policy settings. These three, for example, belong to the tiny handful of countries that have remained "always open" in the Sachs-Warner sense and avoided serious inflation episodes (though not a major growth slowdown during the Asian economic crisis and the global economic recession).

A. Viet Nam's Doi Moi, and Beyond 4
Viet Nam's major reforms from the mid-1980s are of particular interest since they have been highly successful, yet they were undertaken in very difficult circumstances. The country was verging on being a pariah state: frozen out of relations with the US; at loggerheads with its neighbors, the PRC to the north and the Association of Southeast Asian Nations (ASEAN) to the south and west; about to lose the support of its principal international benefactor, the Soviet Union; and having minimal contact with international financial institutions (IFIs). There was a weak technocracy with very limited knowledge of how to manage the transition process and run a market economy. Riedel and Comer (1997) argue that policy makers learnt mainly from their bitter experience with a decade of central planning, including disastrous attempts at agricultural collectivization and nationalizations. In their words: The leadership of Viet Nam did not decide to "go market" because of any kind of ideological conversion from Marxism-Leninism to capitalism; instead it discovered the hard way that the alternative to a market economy does not work.
The Chinese experience, although in its infancy, was closely observed, as this was the country against which Viet Nam benchmarked its performance. Riedel and Comer (1997) stress that the term chosen at the Sixth Party Congress in December 1986, doi moi (renovation), connotes gradualism. However, the hyperinflation of 1986-1988 threatened to undo the early reforms, and this resulted in a successful stabilization program in 1989 that was ". . . pure IMF orthodoxy, albeit without the IMF behind it." The main elements were raising interest rates, devaluing and unifying the exchange rate, legalizing gold holdings, and reducing public sector deficits. Deficits were lowered by reducing the ratio of government expenditure to GDP by six percentage points. Subsidies to state-owned enterprises (SOEs) were largely eliminated, half a million soldiers were demobilized, and major state investment programs were cut. In early 1989, the decision was taken to liberalize prices and eliminate the system of state procurement.
Further reforms followed. SOEs saw a hardening of their budget constraints. They were weaned off central bank credits and increasingly forced to buy inputs from the market. Some were closed, and others brought under the control of the Ministry of Finance. Several laws were introduced clarifying the rights of enterprises. Liberalization of the regime for foreign direct investment began in 1988. Trade reform involved the freedom to engage in international trade and the establishment of export processing zones. Reform of the import regime proceeded more slowly, as did financial sector reform.
The pace of reform slowed in the mid-1990s, a factor compounded by the Asian economic crisis. However, a "second doi moi" got underway in the late 1990s, involving further reform of enterprise laws, more liberal trade and investment regulations, and additional SOE reforms. Viet Nam then enjoyed strong economic growth for a decade, until 2008 when a combination of domestic policy missteps and the global economic recession again slowed growth. This success has bequeathed further problems, including in macroeconomic policy, industry policy, and state enterprise reform (Leung 2010, Pham andRiedel 2012), but there seems little doubt that moderately high growth is now entrenched. This appears to be a case of reform initially triggered by necessity, anticipating a large reduction in its external revenues and disappointment with its central planning experience. It was undertaken by an authoritarian regime intent on national economic development and anxious to learn from and keep up with its neighbors. Apart from necessarily abrupt macroeconomic stabilization, the reforms were mostly gradual and effective. There was a strong export response to the decision to unify the exchange rate, adjust prices to international levels, and free up the trade regime. The country did not experience the Eastern European economic collapse owing to its effective reforms and the absence of a large and inefficient heavy industry sector. There was also some good fortune in the discovery of large oil deposits, which effectively substituted for Soviet aid. Two unusual features were the absence of a group of well-trained technocrats and the country's international isolation. It is difficult to think of a more compelling case of successful reform against formidable odds in recent times.

B. Trade Liberalization: Indonesia in the 1980s
Trade liberalization has been central to policy reform. This is based on the premise that once macroeconomic stabilization has been achieved and a workable political system established, openness is the key policy lever to ensure a competitive economy, by disciplining rent seekers and preventing policy backtracking. That is, the political economy dynamic, not guaranteed but more likely, is that the efficient, internationally-oriented sectors of the economy producing tradable goods and services will exert pressure on the unreformed sectors of the economy and will demand better quality governance and institutions. The struggle for trade liberalization also illustrates up close how and why reform succeeds. In the words of Bhagwati (2002), the literature on the political economy of trade liberalization emphasizes the interplay of "ideas, interests, and institutions." We focus here on two major, though quite different, trade liberalizations in Indonesia and the Philippines.
Indonesia achieved comprehensive reform in the mid-1980s, and this elevated growth rates and almost certainly averted a serious debt crisis. This followed an earlier, more significant, and highly successful change in policy direction in the period 1966-1968. While we focus here mainly on trade reform, the broader context is also relevant. By way of background, the Indonesian economy grew strongly over the period 1967-1982, driven by the return to sensible and credible economic management and large oil and aid revenues. In the early 1980s, however, the global economy began to slow down and oil prices fell sharply, from about $30 per barrel to less than $10 per barrel. With oil, gas, and related commodities generating about three-quarters of merchandise exports and two-thirds of government revenue, the Indonesian economy looked precarious. Growth slowed considerably in the early 1980s, but by the end of the decade the economy was growing as fast as it did during the oil boom period.
The details of the reforms are explained elsewhere (see e.g., Hill 2000). Fiscal policy remained prudent, with immediate adjustment on the expenditure side (mainly the shelving of an ambitious heavy industry program) and a series of effective tax reform measures that lifted revenues. Donors also responded quickly and generously. In addition, there were two large nominal exchange rate depreciations, in 1983 and 1986. Combined with low inflation, these provided a major boost to competitiveness. Once macroeconomic stabilization was secured, the government turned to microeconomic measures and implemented a comprehensive reform package. In trade policy, most nontariff barriers were gradually removed, while tariffs were lowered and unified. Exporters were placed on a free-trade footing through an effective duty exemption and rebate system. A sweeping reform of customs sidelined deeply corrupt and obstructive import/export procedures. Foreign investment restrictions were relaxed. The financial sector was deregulated and the stock market reactivated. Many regulatory barriers to entry were removed, particularly in sectors formerly dominated by SOEs such as the strategically important interisland shipping industry.
What explains the success of these reforms? 5 As most analysts of this episode note, strong opposition to the changes was to be expected. The dominant ideological predisposition of the influential policy community was suspicious of liberalism. As soon as the macroeconomic stabilization and liberalization of the late 1960s began to bear fruit, the pendulum swung back towards dirigisme and control, reinforced by the huge commodity windfall gains. As an indication of the sensitivities, whenever liberal reforms were introduced, they were always referred to by the neutral term "deregulation." Moreover, vested interests had built up around the complex system of controls and intervention in the business sector, the SOEs, and the bureaucracy. There was by contrast a weak export sector and a tiny, marginalized intellectual community calling for reforms.
The key to the success of the reforms was an able, coherent, and powerful group of reformers known as the "technocrats." This group, the so-called Berkeley Mafia, had occupied all the major economic policy portfolios since the beginning of the Soeharto era. Although lacking any significant political party support, they had strong technical credentials. Most importantly, they had developed close relations with Soeharto before he came to power, and they had overseen the remarkably successful stabilization and recovery of the economy in the second half of the 1960s. This was moreover a political system characterized by Mackie and MacIntyre (1993) as one in which "Soeharto (was) in supreme control." In addition, from the margins, external actors were helpful. Relations with Japan were exceptionally close. Japan had become the country's major donor and investor, and it viewed Indonesia as a strategically crucial partner. Throughout this period, it extended its credit lines on highly concessional terms and rolled over most of Indonesia's debt. This was still the cold war era, and Indonesia's relations with the US were also very close. The IFIs provided useful policy and analytical advice on a range of issues. However, it should be noted that the reforms were not part of any formal IMF and World Bank conditionality, a factor that made them easier to sell domestically.
There were, in addition, three facilitating factors that enabled the reforms to be introduced with little opposition and that boosted their effectiveness. One was broad-based development presided over by an authoritarian regime that had nevertheless delivered rapid growth. Second was the absence of serious domestic resistance to the reforms that could mobilize popular opposition. The "economic nationalists" and those in the large SOE sector were either neutralized or pushed aside. The personal, egregious vested interests centered on the Soeharto family were then not significant, unlike a decade later at the time of the Asian financial crisis. Third, the regional (East Asian) climate was conducive-other countries were liberalizing; the Plaza Accord was opening up trade and investment opportunities with Northeast Asia; and the PRC was not yet then a really serious export competitor.
In his comprehensive assessment of the reforms, Soesastro (1989) argues that the process was driven by necessity much more than theory and ideology. The reformers deliberately maintained a strategy of "low politics," avoiding grand ideological debates that would have been polarizing and may have derailed the reforms. Although there was opposition from within the bureaucracy and vested interests in the protected sectors, the packages were implemented effectively and there was a steady flow of new initiatives. Basri and Hill (2004) explained these trade policy dynamics with reference to changes in the relative influence of several key policy actors over this period. That is, they identified the key policy actors, their general trade policy preferences, and how influential they were during each major episode of the Soeharto period.
The drift towards increased protection in the 1970s occurred because both the technocrats and foreign influences on policy were on the wane, at least in the realm of microeconomic policy. They were less needed during these "good times," and there was less imperative to follow their policy orthodoxy. Moreover, neither group was completely united on core trade policy issues. Economic nationalists were becoming increasingly powerful in this decade, and they were able to build opportunistic alliances with various rent seekers. By contrast, in the mid-1980s, the opposition to trade liberalization began to wane. The technocrats were united and stronger in their resolve to reform, and at that time of looming crisis, they had the ear of Soeharto. Foreign influences were clearly pro-reform, and they had more weight. Indonesia needed funding from the IFIs to help it adjust to lower oil prices. Foreign investors were becoming more interested in the country either as a low-cost export platform or as part of internationally integrated manufacturing operations rather than as a relatively small and protected domestic market. Neighboring countries, most especially the PRC, were liberalizing and growing rapidly, in the process constituting a powerful demonstration effect. Finally, the idea of coordinated, open, regionwide liberalization in the form of both the Asia-Pacific Economic Cooperation (APEC) and the ASEAN Free Trade Area (AFTA) was beginning to take root in elite government and business circles.
Summing up, this was a very successful reform in which the core elements were a group of able and credible policy advisors with access to the key source of power in the country and not seriously compromised by vested interests. The trigger, which enabled the technocrats to persuade Soeharto of the case for reform, was a developing external crisis. At the margins, various external actors and factors were helpful (technical advice, funds, and other countries reforming). The high quality of both design and implementation produced results, and won over a larger constituency. This is one of the best examples of successful reform in an authoritarian, growth-oriented state.

C. Trade Liberalization II: The Philippines
Philippine trade liberalizations were eventually just as effective and apparently as durable as those of Indonesia. But by contrast, they were much slower, spanning about 15 years that included a deep crisis, a transition from authoritarian rule to democracy, and three administrations. The case for reform was comprehensively argued in major academic publications from the late 1960s and by the country's leading university economics department, whose graduates have traditionally dominated the main economic policy institutions of government. The major international agencies were also heavily involved, both in advocacy and conditionsbased lending programs. The slow pace of reform, spread over more than 20 years, therefore attests to the strength of the opposition and especially the role of several key veto players.
Philippine trade and industry policies have been extensively documented and analyzed, probably more than in any other developing Asian country. 6 The introduction of "temporary" import controls in the late 1940s in response to a balance of payments emergency combined with an ideological predisposition to support "national firms" resulted in one of the most comprehensive and prolonged periods of import substitution in the developing world. Reform since then has been halting and piecemeal. The peso depreciation of 1970 and the introduction of export incentives provided some relief for export-oriented activities but had little overall effect on the incentives regime owing to the widespread use of quantitative restrictions. By the late 1970s, the intellectual battle for liberalization was largely won, and the World Bank provided a major program of structural adjustment assistance. Average tariff rates and their dispersion around the mean began to fall from 1980, and import licensing was relaxed. A major political and foreign exchange crisis from 1983 to 1986 temporarily set back the reforms, as comprehensive controls on foreign exchange and imports were introduced. However, the crisis-driven exchange rate depreciation boosted competitiveness for the tradables sectors, and there was renewed reform momentum starting in 1987. By the end of the decade, the original trade liberalization program was back on track, albeit delayed. The reforms continued through the 1990s, during both the Aquino and Ramos administrations, and with only a brief and temporary halt in the wake of the 1997-1998 Asian economic crisis. Bautista and Tecson (2003) emphasize the key role of the professional economics community, which staffed major economic agencies by the 1980s. Economists at the University of the Philippines were the key actors here, combined with a quasi-independent government agency, the Philippine Institute for Development Studies, which employed many of its graduates. Supporting this intellectual foundation were three additional sets of factors. First, World Bank programs in the late 1970s and early 1980s provided additional financial and human capital resources particularly during the adjustment phase. Second, there was a realization by the late 1970s that the Philippines was both growing and liberalizing more slowly than its East Asian neighbors, and thus competitive liberalization became a factor of some influence. Third, the reformist Ramos administration (1992)(1993)(1994)(1995)(1996)(1997)(1998) inherited the trade liberalization agenda and implemented it vigorously, not only by completing the schedule of tariff cuts and decontrol but also by a range of other major policy advances, including macroeconomic stabilization, the floating of the currency, and the removal of many regulatory barriers to competition. Perhaps most importantly, the faster economic growth over this period was the most significant reform dividend for a country where "growth pessimism" had become widespread owing to decades of poor performance.
This was a case of slow but apparently durable reform in a number of respects. It commenced under the Marcos regime at a time when the reformers were being increasingly pushed out by the inner circle of "crony capitalists" (Sicat 1985). There was a temporary setback during the crisis of 1983-1987, but the reforms were reinstituted by the Aquino administration, which in other respects was regarded as a rather indecisive and weak regime, attempting to manage economic recovery from a deep crisis and a sudden transition to an unpredictable democracy and against a backdrop of frequent coup attempts. The reform process was then largely completed under the more effective reforming Ramos administration. Many of the policy implementers remained in the bureaucracy over this period, and academic economists continued to occupy the high ground in the debate.
Consistent with the analysis above, Bernardo and Tang (2008) and De Dios and Hutchcroft (2003) identify several sets of drivers. First, the crisis and its aftermath had, with a lag, a galvanizing effect in strengthening the reformers. Second, there was a growing awareness that the Philippines was falling behind in the global trend towards openness, combined with a range of looming regional and multilateral obligations, including the World Trade Organization (WTO), APEC, and AFTA. Third, the removal of the US bases in 1992 ". . . had left the country feeling more exposed" (De Dios and Hutchcroft 2003, 54), and aware of the need to engage more with its neighborhood. De Dios and Hutchcroft (2003, 55) also emphasize the importance of leadership, in particular ". . . the deft and savvy leadership of the president and his key advisors, especially Almonte (who, like Ramos, was a former military officer)." One striking feature, common in such episodes, had been the reformers' "often expressed marked distrust of the Philippine business elite," many of whom were regarded as beneficiaries of the status quo and therefore as obstacles to reform.
One interesting political economy issue is that while protection for manufactures in the Philippines has declined significantly, that for agriculture has risen and now on average exceeds manufacturing. A similar trend is also observable in Indonesia over the past decade. David (2003) offers three explanations, all of which are applicable to post-crisis Indonesia. First, the sustained intellectual reform effort and the subsequent policy response were concentrated where the problem was, that is, high and variable levels of manufacturing protection. Second, agricultural interests were able to exploit loopholes in various international trade agreements thus permitting the imposition of various protectionist measures in the guise of other objectives such as health and quarantine. The slow pace of agricultural trade liberalization in the OECD north was also seized upon by local vested interests. Third, democratization empowered influential rural constituencies, who were able to dress up their demands for protection by playing on sentimental notions of food self-sufficiency (and rural development more generally).

D. Legislated Central Bank Independence and Fiscal Rules
Several Southeast Asian economies have adopted explicit policies designed to ensure central bank independence, tighten the supervision of the financial sector, and limit deficit financing by imposing fiscal policy rules. We examine here reforms of the central banks of Indonesia and the Philippines and Indonesia's fiscal policy law of 2003.
The crises in both countries, Indonesia in 1997-1998and the Philippines in 1985-1986, triggered a reappraisal of macroeconomic management. Both had experienced bouts of high inflation, especially Indonesia, which had hyperinflation in the mid-1960s and again briefly in 1998. Also, in both countries, the central banks were effectively an arm of government, with little operational autonomy and extensive interference. The crises had a devastating impact on public debt through the socialization of financial and corporate debts. As a result, there was a determination to improve macroeconomic management, which attracted broad political support and was consistent with the IMF programs that operated in the wake of the crises. Of the two countries, Indonesia had had the more prudent fiscal policy after it adopted the so-called "balanced budget" rule in 1970, which meant that the government could spend no more than the sum of its domestic revenue and official development assistance (ODA).
While similar in important respects, the modalities of reform differed. The Philippines embarked on a major overhaul of its central bank in 1993, when a new institution, the Bangko Sentral ng Pilipinas (BSP), was established (Gochoco- Bautista, Socorro, and Canlas 2003). The former practice of the board being dominated by cabinet secretaries, who had an interest in the central bank accommodating fiscal deficits and a bias towards a strong peso, was disbanded, as was the objective of exchange rate targeting. The BSP gradually moving to what is considered monetary policy best practice of inflation targeting, formally adopted in 2001, and a floating exchange rate regime.
This was one of the most important and successful reforms in Philippine economic history. The BSP has a highly credible record of monetary policy management, operating as an island of excellence in a public administration system not otherwise known for its high institutional quality. Inflation has remained low throughout the post-reform period, which has been characterized by great volatility, including political turbulence and large exogenous economic shocks. The exchange rate operated as the necessary "shock absorber" in response to the sorts of events that in the past would have resulted in a significant economic slowdown in the country and possibly a balance of payments crisis. Moreover, the financial sector has remained intact without any serious bank runs or failures since 1993.
Fiscal policy settings were also notably improved during the Ramos administration, with three successive years of budget surpluses in the mid 1990s, a highly unusual event in the country's economic history (Sicat and Abdula 2003). However, fiscal policy rules were not institutionally embedded, and for much of the Estrada and Arroyo administrations (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010), the government ran substantial deficits. For several years, the Congress blocked appropriation bills resulting in "re-enactment" provisions; that is, the government simply reverted to the previous year's budgetary provisions, resulting in a substantial reduction in real government expenditures, including civil service salaries.
The Indonesian story differs in two respects. First, although the government formally adopted the principle of central bank independence, the path to reform has been rocky and the inflation record less impressive. Second, however, fiscal policy has been more prudent such that public debt fell remarkably fast, from about 100% of GDP in 2000 to 24% in 2011. The independence of Bank Indonesia (BI) became law in 1999 during the early, chaotic post-Soeharto period. Although BI could no longer purchase government bonds to finance the fiscal deficit and operational autonomy has been more or less preserved, the bank has been the subject of continuous controversy, with three successive governors ending their terms either in jail or house arrest. With regard to fiscal policy, law number 17/2003 required the budget deficit to be no greater than 3% of GDP and public debt to be less than 60% of GDP.
This measure was essentially modeled on the Maastricht Principle but, unlike the EU, Indonesia has kept well within these limits even during the global economic slowdown of 2008-2009. Why and how were these major reforms introduced? In both cases, there was a constellation of forces at work. First, they were introduced after very deep crises. There was a broad recognition of the costs of bad policy and a predisposition to reform. Second, the reforms did not confront any immediate and powerful vested interests. They were not controversial, and there was no grand ideological debate over them. In fact, especially in the Indonesian case, they were introduced without much fanfare, almost "reform by stealth." Third, they had strong backing inside government, from key technocrats in the central bank and ministry of finance. Fourth, they occurred under the presidency of leaders who were both predisposed to reform (especially Ramos) and inclined to listen to their technocratic advisers. 7 Fifth, the role of the IFIs was mixed. All three measures occurred while the countries were under IMF programs, which is presumptive evidence that the fund was a significant player. But in both countries, the fund's role was controversial (and still is in Indonesia), and so it is unlikely that the reforms could have been achieved if any of the four factors mentioned above were strongly negative. Moreover, Indonesia's fiscal law was introduced precisely because the government wanted to exit the IMF program a year ahead of schedule owing to its unpopularity, and this strengthened the hand of President Megawati's advisers, who urged that some institutional restraints on fiscal policy needed to be in place prior to the exit.

IV. Summing Up: Nine (Cautious) Conclusions
The political economy of reform is a complex, multi-dimensional issue in which the analytical literature provides at best a suggestive template. There is, therefore, a tension in the literature, between the academic desire for analytical parsimony and the case study literature that (rightly) emphasizes the complex interplay of history, institutions, ideas, leadership, diverse actors, and external influences. This paper has attempted to steer a middle path, drawing on Southeast Asian and other case study material to highlight factors that appear to be consistently, or least substantially, present during significant and durable reform episodes. The caveat of course is that it is difficult to generalize across a highly diverse set of institutional circumstances, development stages, and policy issues. What worked in the disciplined, authoritarian Soeharto era may not do so in freewheeling and unpredictable Philippine politics. But several recurring themes stand out-these are typically interactive so that their aggregate impact is greater than the sum of their parts.
First, ideas are needed to drive an intellectual agenda, sometimes well formulated in advance, but on other occasions developed in response to specific circumstances. From this "ideas factory," there also needs to be a group of individuals willing to assume public office, interact closely with political leaders, and work together as a united team. However, the link between ideas and policy is an indirect and tenuous one. As the experience of countries as diverse as India and the Philippines demonstrates, there may be long lags between the articulation of ideas and their adoption. These two countries also illustrate that having a strong domestic economics profession is no guarantee that good policies will be adopted, at least quickly.
Second, political leadership is essential, generally featuring a key individual or group of leaders who understand the case for reform and are prepared to actively promote it. Reforms are obviously more likely to be durable the more institutionally embedded they are and the less they depend on a particular individual. One of the challenges of Philippine economic policy making, for example, is that new administrations may well change key policy settings substantially.
Third, major negative exogenous shocks, economic crises, the imminent cessation of external support, and a dawning realization that "the system is broken" have all played a role. The first (a sharp terms-of-trade decline) was the trigger for Indonesia's major reforms in the 1980s. The third and fourth were the key factors in Viet Nam's doi moi, and they were of some relevance in the Philippine reforms of the 1990s. The second resulted in substantial macro and financial sector reforms in the economies affected by the 1997-1998 Asian financial crisis (AFC). But crises are at best only a possible precipitating influence, and there is no guarantee of positive impacts. The AFC reportedly slowed reform in Viet Nam for several years. The mid-1980s Philippine crisis in effect incapacitated government for some time, and it took that country 20 years for its per capita GDP to recover to early 1980s levels. The current global economic recession has so far had little positive reform impact and may have spurred anti-globalization sentiments in some quarters.
Other external factors evidently have mixed effects. The Southeast Asian experience lends support to the international literature suggesting that, putting aside the special but important case of humanitarian assistance, aid works only if accompanied by good domestic policies. It is not clear that donors can influence the domestic reform agenda. Donors have worked effectively with growth-oriented regimes in East Asia, but there is no decisive evidence that donors underpinned the establishment of the regimes.
Conditions-based programs have a very mixed record and often invite a nationalist backlash. A stronger view (associated with William Easterly and others) asserts that aid has a malign influence since it enables recipient governments to postpone hard policy decisions. With the possible exception of the Philippines (and perhaps Cambodia), the latter view receives little support in Southeast Asia, in contrast arguably to the South Pacific and parts of Africa. 8 As the region as a whole progressively moves into the middle-income group, ODA as a share of GDP is anyway declining, to about 0.3%-0.4% of GDP for the larger lower middle-income countries. Where donors can perhaps be most effective is in supporting domestic "agents of change" through building up local analytical capacity, training a future generation of policy makers, and (discreetly) supporting reform-oriented think tanks.
An increasingly powerful external factor is the demonstration effect of a successful reforming economy, resulting in a process of "competitive liberalizations." This factor seems to be much more important in Asia than either Latin America or Africa, with Singapore and the PRC (and more recently India) as the standouts.
Fourth, reforms are durable only if they deliver and thereby win over a constituency of support. This requires that they be reasonably comprehensive so that they are not sabotaged by "unreformed" sectors of the economy. This also implies that implementation is critical. However, the evidence on explicit compensation measures to facilitate reform is mixed. Macroeconomic stabilization is invariably the bedrock upon which reforms are built. For example, trade liberalization can be jeopardized by misaligned exchange rates resulting from macroeconomic imbalances. The mounting protectionist pressures in contemporary Indonesia, for example, appear to be the result in part of a relatively strong currency owing to the strong terms of trade and capital inflows.
Fifth, reform is not a linear progression, and thus long time horizons are needed. For example, the lag between the articulation of the case for trade policy reform and its implementation took over 30 years in India and over 20 years in the Philippines. Advocates of reform have to be prepared for setbacks. Donors rarely have the patience or time horizons to stay the course. The sometimes slow pace of reform emphasizes again the importance of having strong ideas embedded in key domestic institutions (including universities, think tanks, and sections of the bureaucracy) on hand to quickly take advantage of (sometimes unexpected) reform opportunities.
Governments may also experience reform fatigue. An example of this occurred in Indonesia after the appointment of the 1993 cabinet. The role of the technocrats was then downgraded. As a result-and this is at best an exploratory counterfactual-they did not have the capacity to follow through on the financial liberalization they had introduced a few years earlier.
Moreover, the key reformers may be increasingly bypassed, for instance during the Marcos regime in the late 1970s (see Sicat 1985) and the increasingly populist Thai economic policy in the Thaksin era and beyond (Ammar 2011). A key reform strategy is therefore to "lock in" and institutionalize reforms, insulate key technocratic institutions, and render backtracking by a future regime more difficult. Examples, all adopted by some Southeast Asian governments, include independent central banks with a clear inflation objective, legislated restrictions on the extent of fiscal deficits, agencies that require recipients of public subsidies to be subjected to some form of public scrutiny, and broad regional and multilateral trade agreements. Of course, there can never be guarantees against the emergence of a really venal regime other than through a system of democratic checks and balances.
Sixth, the rules of the game change, sometimes dramatically, in the transition from authoritarian to democratic systems where voice, accountability, and public persuasion become important arbiters of reform success. This is most clearly illustrated in the two Southeast Asian countries that have swung from authoritarian to democratic rule in recent times, Indonesia and the Philippines. The two major changes concern the speed and modality of reform. Since there are fewer policy actors in authoritarian regimes, once the inner circle is convinced of the need for change, decisions can be taken very quickly. Reformers do not have to first win their case in the courts of public opinion and parliaments, and fewer concessions need to be made to potential losers. Conversely, it might be argued that while reform in democratic systems is slower, it is more likely to be durable as consultative processes have garnered more widespread community support. Moreover, as Nye (2011) emphasizes in the Philippine context, with effective leadership, democratic space may provide scope to mobilize the support of those groups disadvantaged by politically-inspired favors (e.g., regulations on restrictive practices and barriers to entry) to achieve reform.
Seventh, institutions in some broad sense are critical, but it is not necessary to have "high-quality" institutions to reform. The PRC, Indonesia, and Viet Nam began to institute effective reform programs with very weak bureaucracies and at extremely low levels of per capita income. What mattered had been a clear reform agenda; political commitment; and a sequence for reforms, tackling the major challenges first such as macroeconomic stabilization, the unfettered operation of markets, openness to trade and investment, and major supply-side investments. These experiences therefore cast some doubt on the "institutions rule" hypothesis commonly associated with Rodrik (2003).
But the expression of institutions, that is bureaucracies, clearly do matter, especially where implementation (as distinct from "stroke-of-the-pen" reforms) is central such as tax reform, decentralization, and judicial development. The general presumption is that the bureaucracy is a reluctant reformer to the extent that reform entails a loss of privileges. But this glosses over the heterogeneous nature of most bureaucracies, which typically range from reform-minded segments with analytical strength, such as ministries of finance and central banks, to patronagebased sector and infrastructure departments. The relative strengths of the executive and the bureaucracy and the institutional independence of the latter also matter. If, as in the Philippines, senior echelons of the bureaucracy turn over with each administration, the executive is generally able to operate with little bureaucratic resistance. 9 Eighth, the Southeast Asian experience suggests that it is easier to implement relatively prudent macroeconomic policies and broadly open commercial policy than it is to undertake microeconomic reform. Two political economy factors are at work here. One is that the political consensus in most countries now generally recognizes the costs of macroeconomic instability, and therefore key policy actors are willing to accept that central banks and large fiscal deficits are broadly "off-limits" to political interference. Trade policy is also increasingly governed in substantial measure by the ASEAN and other regional commitments. The second factor is that these policy settings, especially macroeconomic policy, are easier to sustain because there are fewer "veto" players, in contrast to industry policy, state enterprises, government procurements, and so on, where political considerations intrude to a far greater extent. In both these policy areas, the role of the three more advanced Southeast Asian countries (Malaysia, Singapore, and Thailand) as traditionally open, low inflation economies has also been important in setting regional benchmarks.
It is important not to overstate the macro/micro distinction, however. Trade policy remains contested and politicized in most of the countries, and the foundations of macroeconomic policy are shaky in several of them. Examples of the latter include Viet Nam's recent macroeconomic instability, the Philippines' budget travails for much of the past decade, large and highly distorted subsidies in Indonesia and Malaysia, the recent bout of fiscal populism in Thailand, and much else.
Ninth, there does not seem to be any clear association between the propensity to reform and the level of corruption. Corrupt regimes that are also growth-oriented frequently display a capacity for partial reform on the presumption that growth offers greater opportunities for both political longevity and rent seeking. Hun Sen's Cambodia and Soeharto's Indonesia are perhaps the outstanding Southeast Asian examples. In such regimes the nature of the corrupt activities switches, primarily from tradables (where rent seeking is more likely to be disciplined by trade openness) to non-tradables. Of course, there is a corruption threshold beyond which regimes begin to lose political legitimacy and the will to reform, and institutions are undermined. Soeharto around the mid-1990s and Marcos in the early 1980s are the clearest Southeast Asian examples of this phenomenon.