WPS5792 Policy Research Working Paper 5792 Ripe for a Big Bang? Assessing the Political Feasibility of Legislative Reforms in the Philippines’ Local Government Code Yasuhiko Matsuda The World Bank East Asia and Pacific Region Poverty Reduction and Economic Management Unit September 2011 Policy Research Working Paper 5792 Abstract In the Philippines’ highly decentralized political been made. This paper assesses the prospects of legislative system, smooth functioning of inter-governmental reforms on the revenue side of the decentralization relations is key to effective service delivery and good framework. Using a variety of approaches ranging governance overall. Although considered a milestone, from a historical analysis to institutional analysis of the 1991 Local Government Code, the Philippines’ the legislative dynamics in the Philippine congress, it basic legislation governing inter-governmental relations, assesses the prospects of a major overhaul of the Local contains provisions that thwart vertical and horizontal Government Code and concludes that a significant resource equalization among local government units, reform is highly unlikely under the conditions prevailing and contributes to mismatch between expenditure in the late 2010s. By implication, any effort to improve assignments and the fiscal capacities of the local the Philippines’ inter-governmental framework will have government units. Numerous technical reports have to settle for sub-optimal incremental measures within the called for adjustments to the existing revenue and inefficient revenue assignment arrangement. expenditure assignments, yet no tangible progress has This paper is a product of the Poverty Reduction and Economic Management Unit, East Asia and Pacific Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at ymatsuda@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Ripe for a Big Bang? Assessing the Political Feasibility of Legislative Reforms in the Philippines’ Local Government Code Yasuhiko Matsuda1 JEL Codes: D72, H77 1 Sr. Public Sector Specialist, East Asia & Pacific Region, World Bank Acknowledgement: This paper has benefited from comments from Phil Keefer and Motoky Hayakawa as well as research assistance by Allan Layug, Justine Diokno-Sicat and Jay Carizo. Introduction Making decentralization “work� is a key challenge for better governance in the Philippines given its highly dispersed geography and diverse cultural and ethnic groupings throughout its territory. The World Bank has recently prepared a set of analytical reports aimed at identifying critical reform priorities to improve the workings of the Philippines’ inter-governmental relations with particular emphasis on strengthening the ability of Local Government Units (LGU) to provide devolved public services. These reports/notes focused on improvements deemed desirable for efficient functioning of the Philippines’ inter-governmental relations from technical perspectives. One of the findings, which confirms those of other previous studies of the Philippines’ decentralization, highlights the need to amend or possibly overhaul the Local Government Code (LGC) which defines the key parameters of inter-governmental relations.2 As a companion to these reports, the present paper explores the prospects for legislative reforms of the LGC, based on a mix of methodologies, namely (i) analysis of the historical episodes of the decentralization reform in the Philippines and the motivations of the key drivers of the reform; (ii) international comparison of typical conditions/circumstances that lead to a major reform of a country’s decentralization framework; (iii) newly compiled data on bills related to decentralization filed in the Philippine Congress since the passage of the LGC; and (iv) analysis of the incentives of national legislators (as potential opponents of further fiscal decentralization) and those of local chief executives (as potential proponents/beneficiaries of further fiscal decentralization). The paper concludes with a summary of the findings and the implications for reform options. What Is Wrong with the Philippines’ Decentralization Framework? There is a broad technical-level consensus that decentralization is not “working� as well as it could in terms of improving local service delivery and allowing LGUs, at least those with potential, to become effective facilitators of dynamic local economic development.3 The Philippines’ decentralization framework is defined mostly by the 1991 LGC and many experts agree that certain provisions of this framework legislation are problematic, especially those that relate to the assignments of revenues, and to some extent the expenditures. On the revenue side, the oft-mentioned distortions include: 2 For a more comprehensive summary of the technical diagnosis of the weaknesses in the current decentralization framework and its workings, see “Approach Paper on a World Bank Strategy for Supporting Decentralization Reforms�, the Discussion Note (23) “Decentralization: Improving Local Governance for Better Service Delivery,� and the forthcoming Bank ESW report on Local Service Delivery. 3 There are notable exceptions to this generalization, a countable number of LGUs which have established themselves with innovative citizen-oriented programs and good governance in general, and the presence of these LGUs suggests that the current framework allows those well-intentioned and innovative local chief executives to accomplish a lot even within the existing institutional and resource constraints (See “LGU Binding Constraints Case Studies� (tentative title) for examples of such LGUs). However, evidence from other analytical work strongly indicates that the current inter-governmental arrangement in the Philippines leaves much room for improvement. 2  The IRA distribution formula and tax assignments that lead to vertical and horizontal imbalances in LGUs’ resource bases – absence of equalization, mismatch with expenditure assignments;  The tax assignments that favor one type of LGUs – cities – while depriving other levels (especially provinces) of adequate tax bases. The combination of the current IRA distribution formula and the tax assignments exacerbates the vertical imbalance in revenue assignments across levels of LGUs as well as mismatch between devolved service delivery responsibilities and LGUs’ fiscal capacities. On the expenditure side, the incomplete implementation of the legally mandated devolution, thanks to “loopholes� in the LGC and some national government agencies’ reluctance to fully devolve certain functions, has created confused situations of expenditure assignments and murky accountability. The unclear de facto assignments of expenditure responsibilities and the distorted revenue assignments weaken the LGUs’ incentives to cooperate with each other, and this tendency is accentuated by the hyper fragmentation of the LGU jurisdictions. A recently completed Bank report on local service delivery in the Philippines provides rare empirical data on the skewed resource allocations among different levels of LGUs and clear mismatch between expenditure assignments and resource bases even after taking the IRA into account. An in-depth review of spending data from several case study LGUs revealed the troubling patterns of expenditure allocation such as one province spending more than 2.5 times as much on health on a per capita basis than another province in the sample, or one city receiving 1.5 times as much in total revenues (including the IRA transfer) than another city in the sample, resulting in the former being able to spend 1.5 times more on roads and 1.7 times more on health, again on a per capita basis, than the latter (World Bank 2011). Relative levels of spending are only proxies of the quality and adequacy of service delivery, and it is conceivable that some LGUs may make use of limited resources more effectively than others. Nonetheless the wide disparities in revenue assignments across LGUs are likely to be correlated with the adequacy of service delivery and these clearly reflect weaknesses in both the tax assignments and the transfer distribution formula, as currently specified in the Local Government Code. The national government has no explicit policy to compensate for this disparity in LGU revenue assignments. What Sorts of Reforms Are Desirable from Technical Points of View? An ideal solution to the problems enumerated above would be an overhaul of the LGC, probably through omnibus amendments or even its repeal and replacement with whole new legislation. Key content of the reforms would include, inter alia:  A revised IRA distribution formula which approximates the principle of “finance follows functions� in the vertical assignment of transfers and includes a stronger element of horizontal equalization – the direction of the change should be to increase the provinces’ share of the IRA, reduce the cities’ and possibly eliminate the barangays’ as the latter could be absorbed into the administrative structure of the cities/municipalities; 3  New tax assignments, again in the direction of strengthening the provinces’ tax bases (e.g., permitting provincial sales tax with a rate cap, taxes on exploitation of natural resources within the province, etc.) while retaining the robust tax bases especially for cities. In contrast to revenue reforms, refinements in the expenditure assignments are possible largely through administrative measures within each devolved sector. Although the LGC contains a legal loophole that has allowed national government agencies (NGAs) to continue their direct involvement in devolved service delivery, the national government could, if it so wishes, voluntarily shift its roles from direct service delivery to more policy-making and supervisory roles envisioned in the spirit of the LGC. An exception is if the Filipinos want to devolve education as part of the new decentralization framework since education is not included among the sectors formally devolved to LGUs in the LGC. But even without formal devolution, the Department of Education has been able to pursue, albeit slowly, transfers of greater resources to schools as a key element of its sector decentralization strategy. The LGC also provided the LGUs with the ability to levy a special tax earmarked to education and as a result, many LGUs actually spend more on education than on some of the devolved sector responsibilities already. Therefore, an immediate issue in the education sector with respect to center-local relations is better coordination in investments and service delivery between the national government and the LGUs rather than formal devolution of the sector responsibilities. There is much that can be accomplished within the existing legal framework. This suggests that progress on the revenue side will require legislative measures whereas reasonable progress on the expenditure side, especially if pursued on a sector-by-sector approach, will not. The latter then becomes largely a question of the incentives among the LGUs and the relevant national government agencies to coordinate with each other. In some sectors (e.g., agriculture), there may also be a need to revisit the whole policy framework that governs the government’s approach to service delivery before tweaking the distribution of roles and resources across levels of government and in some cases this may require a sectoral legal reform. The prospect of expenditure-side reforms through administrative measures would also depend on whether or not higher-tier entities, be they NGAs or provinces, have the ability to induce voluntary cooperation of the LGUs to achieve more rational distribution of expenditure responsibilities vertically across levels of government and horizontally among LGUs at the same level. This paper mainly focuses on assessing the political feasibility of a legislative reform on the revenue side of the decentralization framework, with particular focus on the reform of the inter-governmental fiscal transfers, Internal Revenue Allotments (IRA). LGUs’ incentives to provide public services are analyzed in a separate paper based on an original survey of LGUs and households in a single province (Khemani 2011). Political Economy of Fiscal Decentralization Reforms: Three Analytical Perspectives The fiscal reforms needed to address the imbalances in the current system of decentralization involve redistributing financial resources from the central government to LGUs and from some LGUs to others. We would never expect the first set of reforms to be approved by central government officials unless 4 there was a significant change in the political calculation of national-level politicians, such as a change in influence of LGUs (e.g., governors) on the political futures of central government officials. We would never expect the redistribution of resources among LGUs unless the losers began to wield significantly less political power than the winners (Garman, Haggard and Willis 2001). This note shows that these conditions are rarely met and that these conditions are not currently present in the Philippines. The main thesis of this paper is that a major decentralization reform, especially the kind that requires a legislative reform, is highly unlikely under the prevailing political economy conditions of the Philippines. We support this thesis with the following sets of analytical considerations. First, we review the political history of the decentralization reform in the Philippines and try to understand the context and the prevailing incentives that led to both the decision to decentralize and the specific content of the reform. We ask if the specific historical conditions that led to the passage of the Local Government Code are present today. Second, we seek lessons from other countries about conditions under which governments have tended to launch decentralization reforms and ask if similar conditions exist in the Philippines today. Third, we review legislative activities since the passage of the LGC to ascertain the extent to which attempts have already been made to reform aspects of the LGC, as an indicator of latent demand for reforms, and ask what changes have already been adopted and what other reform proposals with some prospect of success are still pending. Finally, we posit two hypotheses to explain the limited extent of legislative reforms observed so far and ask if there is any reason to believe the validity of these hypotheses will weaken in a foreseeable future. Political History of Decentralization in the Philippines Political context: In trying to understand political incentives for decentralization, a useful approach is to study the context in which the original reform (the passage of the LGC) unfolded and try to understand motivations of the key actors involved in that process. A relevant question is whether those same incentives and impetus that led to the launch of the reform generally considered a “big bang� still prevail today or if a new coalition eager for a change has emerged.4 The move toward decentralization in the early 1990s cannot be discussed separately from the broader political development which involved the sudden downfall of the Marcos dictatorship and the electoral victory of President Aquino, who acted with limited political ambitions of her own. During the Marcos rule, centralization of power created discontent among local officials because of his centralizing and discretionary handling of local affairs.5 For example, Marcos used to channel the IRA selectively to his allies and supporters, instead of releasing it automatically to all local governments.6 Therefore, the 4 The Bank report East Asia Decentralizes (2005) characterizes decentralization in the Philippines, along with Indonesia’s, as a case of a “big bang� reform in the regional context. 5 Some local officials had to travel to the Malacanang Presidential Palace in Manila for trivial requests such as purchase of certain capital equipment. For some of them far from Manila, the transactions cost surpassed the cost of the equipment to be bought resulting in a huge and unnecessary waste of government resources. 6 Interview with Atty Eleuterio Dumogho, Director of Local Government and Political Affairs of Sen. Aquilino Pimentel, Jr and former Chairman of the Technical Working Group (TWG) during the bicameral conference of SB155 and HB31046, conducted by Justine Diokno-Sicat for a background paper “Political History of Fiscal Decentralization.� 5 Aquino administration’s push for decentralization was partly its answer to the pent-up discontent among local officials as well as its intent to mark a break from the centralizing authoritarian past of the Marcos regime.7 Upon assuming power, President Aquino promised a broad public sector reform package including the devolution of political and administrative authority to local governments (Diokno 2003; Hutchcroft 2004). As part of the transition towards a fully democratic regime, local elections for governors (provinces) and mayors (cities and municipalities) were held in most jurisdictions in January 1988.8 At these elections, pro-administrator candidates captured 12 of the 17 mayoral posts in Metro Manila and the majority of the 62 provincial governorships contested, which must have strengthened the president’s incentive to push decentralization.9 In May 1988, President Aquino created the Cabinet Action Committee on Decentralization “to study how certain additional powers and responsibilities could be devolved to local government units.� Shortly after its creation, the Committee launched the Pilot Decentralization Project (PDP) which transferred administrative and budgetary responsibility over the delivery of certain services to four provinces. Through this pilot, provinces were allowed to implement budgets without prior approval of the Department of Budget and Management (DBM), and were each given PhP 120 million for development projects. The pilot was later expanded to 15 other provinces. The President instructed her secretaries and heads of line agencies to deconcentrate more duties in their regional directors so national agencies could be more responsive to local problems and needs. The Presidency called these efforts, de facto decentralization and “people´s power institutionalized, with or without Congress’ imprimatur.�10 In Congress, legislative action was first taken by Senator Aquilino Pimentel Jr. in 1987 when he introduced a first draft. In November 1989, Senator Pimentel consolidated several Senate bills and created a single working draft in the upper house (Senate Bill Nº155).11 Meanwhile in the House, the Aquino administration had introduced HB Nº31046 in August 1989 with the sponsorship of congressmen Celestino Martines and Ciriaco Alfelor. The President certified this bill as priority legislation and urged Speaker of the House, Ramón Mitra, to file the bill through to the Committee on Local Government. Yet the Bill was not read in Committee until June 1990. Referring to the delaying tactics in the House, a report by the Office of the Presidency states: “But perhaps, many congressmen saw that it would give local governments more access to resources. Ergo, power. The bill gathered dust in Congress for five years. 7 Other probable reasons behind the passage of the 1991 LGC may include: (1) the worldwide trend of decentralization and privatization in the 1980s to 1990s; (2) the belief of President Aquino that democracy could be restored further by institutionally establishing it at the local level; (3) the agenda of legislators to garner the support of local executives by pushing for decentralization; and (4) the influence of international donors that helped shape the law and support the implementation (Hutchcroft 2004). 8 Barangay elections were differed until March 1989. 9 In other countries (e.g., Peru in the 2000s), triumph of the opposition in sub-national elections has often dampened the administration’s incentive to pursue decentralization (Eaton 2009). 10 Quote from the report “Power to the Regions� prepared by the Presidential Management Staff in June 1992. 11 Sen. Pimentel was passionate about having the IRA released automatically to local governments (Dumogho 2009). According to an interview with Ex-Governor Roberto Pagdanganan (2009) who was President of the League of Provinces of the Philippines at the time, the IRA was their main concern in the code. 6 But slowly, the political landscape began to change with the President´s relentless, indefatigable campaign of tapping the grass roots, of linking with the rural folk and NGOs. Furthermore, the President met regularly with the mayors, the governors, and their constituencies. She called on them in planning sessions.� Politics of LGC passage: The House’s delay in initiating a discussion on the bill certified by the President as a priority and the subsequent changes in some of the key content suggest the House was not acting merely as a conduit of the administration. In the end, the final Local Government Code in many ways was a product of compromise, with the final provision taking the “middle ground� between the House and the Senate Bills. However, there were also some provisions that departed from either bill’s position. Key differences found among the two bills and the LGC include: (1) the distribution formula of the Internal Revenue Allotment (IRA); and, (2) the treatment of the education sector. One of the most contentious issues in the passing of the LGC was the design of the IRA. The main stakeholders behind the IRA provisions in the Bills were local officials. The original proposal by the Aquino government was that local governments receive, unconditionally, 20% of internal revenue collections and an additional 5% subject to local government tax effort. HB 31046 proposed that the unconditional IRA be 25% and there be an additional 5% based on revenue effort. In SB155, in contrast, the proposal was an unconditional 35% of net internal revenues given to local governments. In the end, the LGC ended up allocating 40% of the internal revenues as IRA, higher than either the House or the Senate proposals. The two bills differed in their proposed IRA shares among the levels of LGUs as well. The HB31046 proposed a larger amount, 25% to provinces compared to only 20% in the SB155. The Senate proposal gave a larger share of 30% to cities compared to the 25% share in the House bill. The Municipal Mayor’s League of the Philippines (MMLP) lobbied for municipalities to get 45% share in the IRA with the justification that municipalities had “the largest constituency� among the local governments. Curiously, the LGC ended up giving the barangays highest increase in allocation from 15% in both the House and the Senate Bills to 20%, but we were not able to get a credible account of why this increase was agreed on in the Bicameral Conference Committee that hammered out the differences between the two bills, when the other aspects of the IRA design all seem to have followed the logic of “splitting the difference.� The relative “loser� from this bargaining were the cities whose share in the IRA was reduced from 25% in the HB and 30% in the SB to 23% in the LGC. The proposal to devolve education was originally included in senate deliberations and amendments approved.12 However, various public school teachers and parent-teacher organizations opposed the devolution of education services citing as reasons: (1) fear that devolution would compromise the quality of education (Philippine Public School Teachers Association (PPSTA) 1990, Department of Education, Culture and Sports (DECS) Region IV Division of Quezon District of Plaridel 1990); and (2) concern that the devolution would politicize the management of the sector (MACO North District 12 Sen. Edgardo J. Angara and Sen. Pimentel were strong supporters. 7 Teachers Association Resolution 1-90 1990, DECS Region XI Office District of Davao Oriental).13 Another, equally or even more plausible, explanation is that national legislators feared that devolution of teachers, who also served as the board of election inspectors and were responsible for protecting ballots and votes, would be subject to political control and manipulation by local politicians, many of whom are national legislators’ political rivals. Table 1: Comparative Features/Highlights of HB 31046 & Senate Bill 155 (Proposed Local Government Code) on the Powers, Functions and Responsibilities of LGUs Local Government Code Issues HB 31046 SB 155 of 1991 Comments (RA 7160) IRA Gross internal revenue 35% of net IR collection 40% of internal revenue The Aquino collection; 25% of the based on the second collections of the third administration collection from the IR fiscal year preceding fiscal year preceding the initially taxes computed on the the current fiscal year current fiscal year (Sec. proposed 20% basis of the gross Sec. 247 (1) 284 (c)). unconditional collection of the second and 5% subject fiscal year preceding to revenue the current fiscal year. effort. IRA shares 25% to Prov 20% to Prov; 23% to Prov; The final figures 35% to Mun 35% to Mun; 34% to Mun; are higher for 25% to Cities 30% to Cities 23% to Cities; barangays. 15% to Brgys 15% to Brgys. 20% to Brgys (Sec. 285) IRA formula Population 60% Population 35%; Population 50%; Land area 20% Land area 35%; Land Area 25%; Equal sharing 20% Equal sharing 30% Equal Sharing 25% (Sec 223) (Sec. 285) IRA additional An additional 5% No additional incentives incentives incentive for collection in the final draft. efficiency among the LGUs. Elementary Sec. 27-B, Devolved the Elementary school Amended in school system elementary school system was deleted and Senate system only the school-building deliberations program was devolved. but dropped in the final law. This brief and selective sketch of the politics of the LGC passage implies that: (a) the decentralization reform should be understood as an extension of the democratization process following the fall of the centralizing Marcos regime; 13 There were allegations of lobbying from the Education Secretary to remove the education system as one of the government services to be devolved and against other provisions pertaining to the education system in the proposed local government code as supported by the PPSTA (1990). 8 (b) while there clearly was some discontent with the centralized political system under Marcos especially among local politicians, the available historical account reveals little evidence that civil society and voters at large were strongly demanding decentralization of the government structure; (c) instead, available accounts of the history highlight the importance of the personal interest by President Aquino and the role of a few policy entrepreneurs such as Senator Pimentel Jr. who emerged as strong champions of decentralization and interests of local politicians acting through their own leagues as well as through sympathetic national legislators; (d) yet the final shape of the LGC was very much a product of political compromise among the House and the Senate (as well as the administration), showing various telltale signs that national legislators were loath to give large resources and power to local politicians, especially to provincial governors. In sum, although the available evidence is limited and far from unambiguous, there is a sufficient ground to conclude that the decentralization reform in the Philippines was launched more as a top-down initiative. Once launched, however, the reform’s content was shaped through political compromises among national politicians some of whom represented local politicians’ interests while others were more antagonistic or ambivalent at best. The municipalities and the barangays, collectively, came out with relatively beneficial fiscal positions as far as the IRA shares were concerned, whereas provinces emerged as relative “losers� in terms of resource shares. Cities fared worse in the final version of the law than in either of the bills. However, they still ended up with relatively generous levels of funding given the extent of devolved responsibilities and the relatively small total number of cities among which the cities’ total share would be divided. A plausible hypothesis is that the national legislators chose to favor those local jurisdictions that were (a) less likely to challenge their own political status (unlike provincial governors who have tended to emerge as political rivals of members of the House in some localities) and (b) more likely to be in control of voters at the grass-root level (i.e., barangay-level leaders and some municipal mayors).14 The final decision not to devolve education, a peculiar choice in comparison to decentralization in many other countries, can also be explained in terms of the national legislators’ election-related concerns. Contextual Factors Favoring Changes in Inter-governmental Relations: Lessons from International Experiences International experience shows that decentralization tends to be a highly politically charged topic because it involves redistribution of power and resources across and among levels of government and is often a constitutional matter. As such, changes to the country’s intergovernmental relations seem to 14 Evidence that corroborates this hypothesis is the account of political posturing by two prominent national legislators during the process. Before the passage of the 1991 LGC, the head of the House of Representatives, Speaker Ramon V. Mitra and the Senate President, Jovito R. Salonga were both planning to run for president in the 1992 elections. As a result, there was political grandstanding between the two which resulted in a higher than proposed IRA (Diokno 2003), from the initial proposal of a maximum of 25% to a 40% share of internal revenues. The final result was a staggered annual increase in the IRA from 30% during the first year of implementation to 40% for the third year after. 9 occur only sporadically under somewhat extraordinary political circumstances. Here we consider three such scenarios: decentralization as part of a broader political change (e.g., democratization), decentralization as a top-down strategy for regime consolidation/survival, and adjustment in inter- governmental fiscal relations under fiscal crisis. Decentralization as part of a broader political/constitutional change: International experiences as well as the Philippines’ own trajectory suggest that a major reform to a country’s center-local relations often obtains as part of a broader re-configuration of its constitutional system. Such an event seems to be more likely when, for example, an autocratic regime (typically with a tendency to centralize power and resources) is replaced by a democratic government (e.g., Brazil in 1988, Peru in 2001, Indonesia in 2001). As discussed above, the Philippines itself falls in this category as its decentralization reform unfolded during the first post-Marcos administration of Cory Aquino. Based on patterns seen in these countries, a major overhaul of the decentralization framework, once put in place, rarely repeats itself, especially if the framework is enshrined in the constitution (e.g., Brazil). In Peru, the strong fervor for decentralization swept through the country in the immediate aftermath of the sudden end of the semi- authoritarian regime (Fujimori) in 2000, but politicians’ appetite for pushing decentralization further quickly waned. Analysts now consider decentralization in Peru as paralyzed (Eaton 2010). Although the Philippines’ political system is often perceived to be unstable, events that could lead to a fundamental change in the country’s constitutional framework have not taken place since the fall of Marcos.15 Decentralization as a government-led strategy to strengthen the center: In some cases, decentralization has been a part of the national government’s strategy to diffuse pressure for political change and ameliorate popular discontent with government performance. Mexico’s Institutional Revolutionary Party (PRI) accelerated fiscal decentralization over the 1990s when the growth of the opposition began to threaten its 70-year hold on power. The measure was taken in parallel to other controlled actions to liberalize the political system and constrain the national executive (e.g., judicial independence, freedom of information). Vietnam’s decision to transfer greater fiscal resources and administrative responsibilities to sub-national levels is also attributable to the Communist Party’s concern about popular discontent with government performance (Guidotti and Gironde 2010). These instances of “center-led� decentralization are predicated on a relatively cohesive national government capable of making and implementing major decisions based on its political calculus and strategy. In both examples cited above, the national government was led by a dominant single party and the Philippines lacks such a cohesive government or party with political foresights and incentives to approach decentralization reforms strategically.16 15 Although the Arroyo administration flirted with a proposal to change the current unitary state to federalism, the move was seen by many to be politically motivated, aimed at a disguised way to seek a constitutional amendment that would allow the president to stay in power longer. Even Senator Pimental, a genuine supporter of the federalism proposal, withdrew his support for the idea temporarily. 16 O’Neill (2003) argues that governments tend to launch decentralization when the governing party enjoys more secure electoral support at the sub-national level than their prospects in national elections. In the absence of a cohesive party, such a condition will never materialize in the Philippines barring a major reconfiguration in the party system. 10 Adjustments under fiscal crisis: Besides a political change, fiscal stress has sometimes triggered constitutional/legislative reforms to adjust the inter-governmental fiscal relations. For example, Colombia has gone through a series of constitutional reforms partly in response to the large fiscal imbalance decentralization has created (Dillinger and Webb 1999). Repeated fiscal crises due to undisciplined sub-national fiscal behavior also led to adjustments in inter-governmental fiscal relations in Brazil, which culminated in the passage of the Fiscal Responsibility Law (FRL) in 2000 (Schneider 2006). Colombia has begun to reign in errant sub-national governments with very tight central oversight of fiscal, and increasingly sectoral, performance. Similarly Brazil’s national treasury now keeps very tight control over sub-national borrowing and states’ and municipalities’ compliance with the FRL rules. Evidence began to emerge that fiscal performance came to be seen as a key element of good governance with electoral consequences at the municipal level.17 Although the Philippines’ fiscal circumstance is tenuous at the moment, this is not because of an “excessive� transfer system or uncontrolled growth of sub-national debts. Unlike the sub-national governments in Argentina, Brazil and Colombia which have been accused of triggering national fiscal crises, the LGUs in the Philippines do not display fiscal populism, although there is some evidence that expenditures increase near election times. Rather than resorting to expansionary spending policies, the LGUs in the Philippines seem more content with keeping the level of service delivery low, relying on funding from a variety of national sources including congressional pork barrel, and blaming (only rhetorically, perhaps) lack of resources for their inability to meet constituent demands. Given this low- level equilibrium, it seems unlikely that an LGU-induced fiscal crisis will erupt and eventually induce adjustments to the intergovernmental fiscal relations as in those Latin American countries. Summary: Experiences of middle-income countries which have been grappling with problems of incomplete or inefficient decentralization suggest (at least) three contextual patterns to overhauling or adjusting inter-governmental relations. In many cases, initial launch of decentralization has been associated with watershed changes in the political regimes (usually transitions from a centralizing authoritarian regime to democracy) or the constitutional framework, often in response to pent-up demand from society to introduce greater pluralism in the political system so as to restrain the national executive’s ability to dominate politics. In a more limited number of cases, a strong national government has initiated a process of greater delegation or devolution of fiscal resources and administrative responsibilities as a way to thwart a crisis of “performance legitimacy� and to preserve the essential feature of the political system and regime survival. Finally, crippling fiscal crises that have resulted from decentralization in Latin America have resulted in the national government re-asserting its control over sub-national fiscal behavior, sometimes leading to a constitutional reform or an equally significant change in the nature of the center-local relations. The Philippines today fits none of these scenarios which elsewhere have led to political decisions to alter intergovernmental relations. 17 Eduardo Leoni and Lucio Rennó (2006) “Reelection and Fiscal Responsibility Law: Reducing Populist Pressures in Brazil.� (http://svn.cluelessresearch.com/representation/reelection.pdf) 11 Legislative Activities and Their Institutional Underpinnings Even if some macro political economy conditions that have induced decentralization reforms in other countries are absent, there may be other signs within the Philippines that point to a prospect of legislative reforms in the near future. Some political systems are capable of introducing major policy change gradually over a period of time through a series of incremental reforms.18 Although the Philippines’ political system is not generally known for such a quality, it is possible that frustration with the LGC may have translated into build-up of latent demand for change. One such indication may be found in the record of legislative activities related to the LGC. Is the LGC sacrosanct? Between the 8th and the 14th Congress (1987-2010), members of both chambers have introduced a large number of bills related to decentralization. Yet very few have ever reached an advanced stage of deliberations (i.e., third reading) and fewer still eventually received approvals by both chambers (and one was still vetoed by the president). Table 2: Number of Bills Related to Decentralization, 1987-2010* Senate House Bills introduced 302 465 Completion of second reading pending 32 55 Completion of third reading pending 9 16 Passage by both chambers 5 3 Vetoed by President 1 1 *Excludes the Local Government Code (RA7160) itself Box 1: An Overview of the Congress and the Law-Making Process The Philippine Congress is a bicameral body consisting of the Senate (Upper House) and the House of Representatives (Lower House). The Senate is composed of 24 senators each serving a 6-year term, half of whom are elected every three years. The Senators are elected on nation-wide ballots. The House of Representatives is composed of 267 members who are either district representatives or sectoral (party-list) representatives elected every three years. Twenty percent of the members of the lower house are party-list representatives. Except for bills relating to “appropriation, revenue or tariff, bills authorizing increase of public debt, bills of local 19 application, and private bills� which shall originate exclusively in the House of Representatives, either chamber of the Congress may propose a bill to become a law. The procedures of passing a bill are generally the same in both chambers. First, proposed bills are filed with the Secretary General who then labels and numbers the bill. This label is distinguished by the mark “H.B.� if filed in the House of Representatives, or “S.B.� if filed in the Senate. The labeled and numbered bill is then reproduced and calendared for business. All bills undergo three readings. On the First Reading, the bill is then referred to the appropriate Committee or Committees as the case may be. The Committee thereafter conducts public hearings or discussions, prepares a 18 Arguably Brazil falls in this category of a country which has managed to introduce gradual improvements in its constitutional framework through 64 amendments (as May 2010) since its promulgation in 1988. In contrast, the Philippine constitution which went into force roughly around the same time as the Brazilian constitution has seen no amendment whatsoever. 19 Sec. 24, Article VI, 1987 Philippine Constitution 12 report and transmits it for the Second Reading. The sponsorship, debates and amendments happen at the Second Reading. After the amendments are acted upon, the House votes on the bill and thereafter submits it for the Third Reading. Final voting happens at the Third Reading. If the bill is approved, it is submitted to the other chamber for concurrence and if disapproved, it is transmitted to the Archives. The required vote to enable a bill to be approved is a simple majority of the members present. The bill that is approved and submitted to the other chamber undergoes the same process. In case there are conflicting provisions or differences, a Bicameral Conference Committee is called. This Conference Committee is composed of Members from each House to settle, reconcile or thresh out differences or disagreements on any provisions of the bill. The Conference Committee then prepares a report which is submitted for consideration of both Houses. If the report is approved, the bill is then transmitted to the President. The President, in turn, may sign the bill into law or veto it. If the bill is vetoed, it, together with the message citing the reason for the veto, is transmitted to the House where the bill originated. The Congress, if it so decides, may override the veto by a two- thirds vote of the members of each house in separate sessions. Of the more than 700 bills related to decentralization filed over the last two decades, only four were eventually enacted into laws. These are:  RA 8185 (1996): Amending the Sec. 324 to specify more precisely the use of the Calamity Fund (CF) with the effect of reducing LGUs’ discretion in the use of the CF.  RA 8285 (1996): Amending the Sec. 284 which defines the LGUs’ share of the national internal revenue taxes (40%) to appropriate a fixed amount (P14,445 million) in 1997 (applicable only that year).  RA 9009 (2001): Amending the Sec. 450 to increase the average annual income requirement for a municipality or a cluster of barangays to be converted into a component city from P20 million to P100 million.  RA 9640 (2009): Amending the Sec. 140 to reduce the rate of the so-called amusement tax from 30% to 10% of the gross receipts from the admissions fees of theaters, movies, boxing, etc. Besides these four, the Senate Bill 1173 (1993), which was intended to give LGUs up to 3 years to absorb devolved health facilities, was also passed by both houses, only to be vetoed by the president. With a possible exception of the RA 9009 which was intended to control the proliferation of cities through conversions from municipalities induced by the larger IRA share the conversion entailed, none of the other three amendments addresses fundamental structural distortions with which the LGC was born. RA 8185 and RA 9640 were clearly designed to weaken the LGUs’ fiscal autonomy, though this is not necessarily a negative development in all cases. In the case of RA8185, tighter regulation of the use of the CF was probably a sound practice given the risk of misuse of the fund but RA 9640 appears to have been a case where special interests (i.e., the entertainment industry) obtained legislative concessions through lobbying the national congress. In other words, the main issue at stake in this particular legislative change had little to do with the architecture of the intergovernmental fiscal relations. RA 8285 was an ad hoc one-time measure made necessary because of a decision by the executive not to include appropriations for the IRA in the General Appropriations Act in that particular year (1997) as has been done in every other year before and since. Even the SB1173 would have had 13 only a temporary effect of slowing down the mandatory devolution process rather than changing the mix of devolved functions or their funding. Thus in spite of the flurry of legislative activities, the key parameters of the decentralization framework such as the IRA distribution formula and tax and expenditure assignments have remained intact for close to two decades. This fact alone may be the most straightforward indication that the status quo is the preferred optimal point for key political decision-makers. Latent demand for adjusting the LGC? It is also possible, however, that some of the political actors such as local chief executives or individual members of the congress are not entirely content with aspects of the existing intergovernmental arrangement. The very large number of bills filed to amend the LGC suggests that some changes are being sought by legislators, who presumably represent some underlying societal interests. A break-down of the bills filed in each chamber by category is suggestive of the types of changes sought by those legislators who filed bills related to the LGC. Table 3 shows the legislators’ interests have indeed concentrated on core architecture of the intergovernmental system, such as expenditure and tax assignments as well as fiscal transfers. Table 3: Bills Related to LGC by Category (House/Senate), 1987-2010 Congress Expenditure Tax Local Local financial Omnibus Inter- Total (years) assignment assignment capital management amendment gov’tal finance transfer 8 (87-91) 6/1 4/5 0/0 1/2 1/2 11/8 23/18 9 (92-95) 17/12 14/6 1/0 4/2 0/0 9/9 45/29 10 (96-98) 23/14 19/8 1/0 4/10 0/1 27/19 74/52 11 (99-01) 22/5 27/8 7/1 3/1 4/0 33/10 96/25 12 (02-04) 14/14 23/13 4/3 1/8 0/4 29/11 71/53 13 (05-07) 14/16 34/12 2/5 0/6 0/3 28/14 78/56 14 (08-10) 14/16 31/20 2/4 1/15 0/5 30/9 78/69 Total 110/78 152/72 17/13 14/44 5/15 167/80 465/302 Further breakdown of these three main categories, expenditure assignment, tax assignment and intergovernmental transfers, shows that proposals to adjust devolution of functions (usually in the direction of reducing LGUs’ responsibilities) and to specify the use of a variety of local funds (e.g., calamity fund, local development fund) dominated the bills filed related to expenditure assignment (Table 4). Of the bills related to local funds, the issue area that has attracted the most legislative proposals relates to LGUs’ budgets for disaster response. Thirteen of the 43 House Bills and 10 of the 11 Senate Bills call for measures ranging from allowing LGUs to use a portion of the Local Calamity Fund (LCF) more flexibly (e.g., 25-30% of the LCF to be used for “preparation� for future calamities) to allowing them to use a portion of their Local Development Fund for disaster response and be reimbursed by the national government subsequently. Other bills propose creation of a variety of earmarked funds for specific purposes ranging from funding social programs (e.g., nutrition, scholarships) to creating an equivalent of 14 PDAF for provincial council members or barangays, to other sector specific funding requirements such as an Agricultural Productivity Enhancement Fund, a Counterpart Fund for PNP, etc. Fourteen of the 25 House Bills call for re-centralization of the devolved functions, especially in health, followed by roads (especially the responsibility for maintaining barangay roads) and other services (agriculture, social welfare and development). These were all filed relatively recently – most of them were filed between the 11th and the 14th Congress (i.e., during the 2000s). Only 4 House Bills in this category were intended to reinforce devolution either by assigning additional functions to LGUs or strengthening LGUs’ regulatory power over certain matters (e.g., operation of public utilities within the LGUs). In contrast only 7 of the 43 Senate Bills that dealt with devolution called for any form of re- centralization. Similar to the House measures, however, most (6 out of 7) of these Bills were filed fairly recently between the 13th and the 14th Congress, suggesting an undercurrent of desires among some politicians to push for re-centralization of devolved functions even after more than 15 years of the LGC enactment. It is probable that, along with those calling for more flexible use of LCF by LGUs, the bills calling for re-centralization of devolved functions reflect latent desires of many local politicians and their sympathizers. Table 4: Bills Related to Expenditure Assignment (House/Senate) Congress Personal (years) Devolution of Local service functions Expanded SEF expenditures Local funds limitation Total 8 (87-91) 0/1 0/0 3/0 3/0 0/0 6/1 9 (92-95) 5/10 1/0 3/0 6/1 2/1 17/12 10 (96-98) 4/9 1/0 3/0 6/4 2/1 23/14 11 (99-01) 5/0 4/2 1/1 9/0 0/2 22/5 12 (02-04) 4/6 2/5 1/0 7/1 0/2 14/14 13 (05-07) 4/6 5/5 2/1 3/3 0/1 14/16 14 (08-10) 3/11 6/1 4/1 1/2 0/1 14/16 Total 25/43 19/13 20/3 43/11 3/8 110/78 With respect to the bills related to tax assignment, most of the legislators’ interests were focused on the provisions in the LGC related to local taxation. A review of the bills’ content clearly shows that thrusts of most the bills were to reduce tax burdens on various constituencies ranging from the amusement/ entertainment industry to property owners to specific government/quasi-government bodies such as electricity cooperatives. At least 37 of the 101 House Bills on “Local Taxes� called for exemption or reduction in local taxes – 21 bills on reducing amusement tax and 14 on real property tax. Ten other bills proposed to limit the increase in the real property tax payments by making mandatory real property assessments less frequent (from every 3 years to either every 5 or 10 years) or by outright prohibiting a rate increase under specific circumstances. Fifteen of the 35 Senate Bills classified as related to “Local Taxes� proposed some form of exemptions or reductions in existing taxes. Of these, 14 related to exempting or 15 reducing taxes on amusement and entertainment and were later consolidated and passed as RA9640 (2009). Bills intended to increase taxation have been filed, though these are fewer in number than those intended to reduce taxation. For example, eleven House bills were filed between the 10th and the 14th Congress to increase the rate of the tax on idle lands from 5% to 10%. Twelve House bills were intended to strengthen barangays’ tax shares (e.g., automatic retention of 50% of the community tax by barangays). In contrast, bills meant to strengthen provinces’ taxing power were less frequent (one HB and 4 SBs). Table 5: Bills Related to Tax Assignment (House/Senate) Congress Requirement (years) for LGU Schedule of Tax Local creation/ fair market assignment/ revenues Local taxes conversion value exemption Total 8 (87-91) 3/2 1/3 0/0 0/0 0/0 4/5 9 (92-95) 2/2 9/4 2/0 0/0 1/0 14/6 10 (96-98) 4/4 12/4 1/0 1/0 1/0 19/8 11 (99-01) 6/0 19/6 1/1 1/0 0/1 27/8 12 (02-04) 4/9 15/2 2/2 1/0 1/0 23/13 13 (05-07) 4/5 22/5 0/2 3/0 5/0 34/12 14 (08-10) 2/2 23/11 2/3 1/3 3/1 31/20 Total 25/24 101/35 8/8 7/3 11/2 152/72 With respect to the bills filed on intergovernmental transfers, most are related to the IRA. Some bills call for increasing the LGUs’ share in the national taxes from the current 40%. Others propose a new distribution formula and yet others refer to procedures for IRA’s releases, retention and so on. Of the 117 House Bills related to IRA, 10 are meant to earmark a portion of IRA for specific purposes such as health, 5 to require automatic releases of IRA to LGUs, 20 to reallocate IRA among LGUs, mostly by deducting the cost of devolved functions before distributing the rest following the formula but also by changing shares among LGU types (e.g., less for high-income LGUs and more for low-income LGUs, 23% each for provinces and cities), 15 to revise the distribution formula - among these 15 are at least 10 different concrete proposals, indicating there is little consensus among the legislators about what a better formula is. Twenty-four bills aimed to increase the LGUs’ share of the national taxes from the current 40% to a higher share (e.g., 50-60%) have been among the most popular among legislators. 16 Table 6: Bills Related to Intergovernmental Transfers (House/Senate) Congress (years) IRA LGU share in the national Total wealth 8 (87-91) 4/7 7/1 11/8 9 (92-95) 7/7 2/2 9/9 10 (96-98) 20/19 7/0 27/19 11 (99-01) 22/10 11/0 33/19 12 (02-04) 22/11 7/0 29/11 13 (05-07) 19/9 9/5 28/14 14 (08-10) 23/5 7/4 29/9 Total 117/68 50/12 167/80W Although the vast majority of these legislative initiatives have failed to culminate in legislation as shown above, the distribution of the types of bills filed serve as an indication of the aspects of the LGC about which legislators hold a degree of discontent. Broadly speaking, most of the bills touch on those fundamental aspects of the LGC that we as well as numerous other analysts have identified as problematic. This can be interpreted as latent demand for change in the LGC, although it is not clear from the data whether there is anything resembling a degree of consensus about the direction and the extent of changes sought. Prospect of an IRA reform: In the 14th Congress (2007-10), a number of bills for reforming the IRA were filed in both chambers, some aimed at increasing the LGU shares and others proposing a new distribution formula.20 Of these, two Senate bills and five House bills aimed at increasing the LGUs’ IRA share from the current 40 percent of the internal revenues to either 50 or 60 percent. Two Senate bills (including one of the ones for increasing the LGUs’ IRA share) proposed automatic retention of IRA transfers by LGUs. One Senate bill and one House bill each called for deduction of costs of devolved services from the IRA calculation so that those LGUs which have received costly devolved functions (e.g., tertiary hospital) could offset these costs and receive an IRA transfer following the existing formula. Finally, several House bills were filed to change the IRA distribution formula in a variety of ways, including some proposals that appeared to be in line with technically ideal formula (e.g., greater considerations of poverty or LGUs’ revenue capacity) presumably to improve horizontal equity of the distribution formula. Those bills that aimed to change the IRA distribution either by deducting the cost of devolved functions or by explicitly changing the formula itself were likely to face resistance from those LGUs that stood to lose their IRA shares since any change in the distribution within the existing IRA envelope would imply a zero-sum situation among the LGUs. Since most members of the Lower House who were elected from specific geographic districts were likely to face conflicting interests among the LGUs in their jurisdictions, with some losing and others gaining from the redistribution of the constant IRA resources, the dominant strategy (in a game theoretic sense) among these legislators is likely to be one of inaction. In contrast the bills aimed to increase the LGUs’ share in the IRA envelope could garner stronger support since such 20 Bills related to other aspects of the LGC may also have been filed but we did not track these specifically. 17 bills do not pit one group of LGUs against another as potential winners and losers. In these cases, the conflict will be between the LGUs as a whole on the one hand and the national government on the other. Here, a key question is to what extent legislators represent interests of the LGUs in their districts. We will turn to this analysis later on. Remarkably, only five House bills out of the total of 16 bills filed in both chambers were sponsored by more than one legislator. Although it is possible that some of these bills will acquire additional sponsors as they go through the legislative process, the single sponsorship is likely to be a reflection of the atomistic and uncoordinated legislative activities in the Philippine Congress with low probability of eventual passage as law. Only one of these multi-sponsor bills aimed to change the IRA distribution (by considering poverty incidence in the formula). The other multi-sponsor bills all proposed changes to the IRA that were not zero-sum in nature among the LGUs. All Senate bills were sponsored by a single author.21 Of the 16 bills filed in the 14th Congress, the one with the greatest likelihood of being enacted into a law was the HB No. 3708 which gathered 62 sponsors in the Lower House and did not propose a zero-sum change among the LGUs. However, the 14th Congress closed before the bill could be approved by the Committee on Local Government to reach the second reading in the plenary. The predominance of single-author bills does not augur well for their legislative success. Furthermore, a close look at the bills reveals many of them were phrased with similar or identical wording either because they were simply expired bills from previous congressional sessions filed again or because the authors apparently based the content on materials provided by the same advocacy groups such as the Leagues of LGUs. These anecdotes lend further credence to the perception that legislative activities among individual legislators were poorly coordinated both because of the absence of well-organized political parties which would play such a role and because, perhaps, of the legislators’ lack of real interest in seeing their bills become a law. 21 Two senators, Aquilino Pimentel (“father� of the LGC) and Jinggoy Estrada, a former mayor, sponsored two bills each. 18 Table 7: Key Features and Classification of Bills on IRA, 14th Congress Key Features Senate Bills HOR Bills 1. Increase of NG-LGU IRA Sharing a. 40-60 Sharing SB No. 8 HB No. 2768 SB No. 119 HB No. 2937 b. 50-50 Sharing HB No. 3533 HB No. 3708 HB No. 4920 2. Automatic IRA Retention SB No. 8 HB No. 4258 SB No. 118 3. New/Additional Variables in Inter-LGU IRA Sharing a. Deduction of Costs of Devolution SB No. 520 HB No. 3845 b. Inclusion of Population and Poverty Incidence HB No. 0181 HB No. 2413 HB No. 4232 c. Inclusion of Marine Waters HB No. 3506 d. Consideration of LGU’s revenue capacity HB No. 4988 Note: The multi-sponsor bills are denoted with italics. In sum, the legislative activities on decentralization matters in the post-LGC period are characterized with a high level of legislative activism and strikingly low legislative “productivity.� First, national legislators in both houses routinely file bills intended to amend the LGC. Second, most of these bills touch on some fundamental aspects of the LGC such as expenditure and tax assignments and the design of the intergovernmental fiscal transfers. This can be interpreted as a sign that there is latent demand for altering some core architecture of the intergovernmental fiscal arrangement enshrined in the LGC. But there does not seem to be anything resembling a consensus on the directions of the change sought. Third, despite the large number of bills filed, precious few have ever passed through the entire legislative process to become a law and those that did tended to deal with issues tangential to the core of the decentralization questions. Fourth, in the most recent, 14th Congress, legislators from both chambers filed 16 bills related to IRA, just an aspect of the LGC, though clearly a very important aspect. One of them in particular appeared to have some prospect of garnering a level of support, as it managed to get more than a third of the entire Lower House as co-authors already in the first reading. But even this bill did not prosper, at least during the 14th Congress. Explaining Low Legislative “Productivity� on Decentralization: Two Hypotheses We posit two sets of hypotheses to explain this pattern as an aid to predict likely developments in the future. These hypotheses are not logically mutually exclusive. In fact, both may be at play to reinforce legislative inaction in the Philippine case. One refers to the generic feature of the Philippines’ legislative process in terms of low institutional capacity for collective action. The other refers to the specific constellation of interests among key actors with respect to the topic of intergovernmental fiscal relations. While full empirical tests of these hypotheses are beyond the scope of this paper, the available information is fully consistent with both of them, which leads us to the conclusion that a legislative overhaul of the LGC is highly unlikely, short of a major change in the basic architecture of the 19 country’s political system, or some extraordinary exogenous shock (e.g., major fiscal crisis), which we had earlier concluded was absent in today’s Philippines. Weak institutional capacity for collective legislative action The Philippine Congress is known for its particularly fragmented institutional structure and its members’ particularistic tendencies.22 The absence of disciplined political parties capable of mobilizing and organizing votes of individual members means that the transaction cost of orchestrating a major legislative initiative is significantly high. It follows that even when a substantial number of legislators, including those in senior leadership positions, are in favor of a particular bill, regardless of its content, the probability that it will pass as a law through the legislative process is low.23 This observation holds for any type of legislation whether it is related to decentralization or not. It is also a crude argument because it is obviously untrue that no legislation is ever passed and enacted in the Philippines. Nonetheless the available data does demonstrate that the probability that a bill becomes a law in the Philippines is unusually low. An examination of legislative records shows very low rates of bill enactment – on average, only around 3% of the bills introduced to the House are ever enacted into laws (Table 8). Even in comparison to these low enactment rates, the fate of the bills related to decentralization has been even worse. The enactment rate of the Senate bills on decentralization for the period 1987-2010 is 1.7% (5 out of 302) and that of the House bills is 0.6% (3 out of 465). Table 8: Bill Enactment Rates at the House of Representatives, 1987-2004 Congress/Administration National application Enactment Local application Enactment (Passed/Introduced) rate (Passed/Introduced) rate th 8 /Aquino (1987-92) 191/5,237 3.6% 809/30,183 2.7% th 9 /Ramos (1993-95) 156/3,184 4.9% 306/11,448 2.7% th 10 /Ramos (1996-98) 147/3,785 3.9% 393/6,766 5.8% th 11 /Estrada (1998-2001) 67/4,197 1.6% 348/8,738 4.9% th 12 /Arroyo (2001-04) 89/2,920 3.0% 84/3,764 2.2% Total 650/19,323 3.4% 1,940/60,899 3.2% Source: Kawanaka (2010) For comparison, legislative data for Mexico for 1991-97 show how a government run by a disciplined single party can produce results in stark contrast to the Philippines’ “low legislative productivity.� While the legislative “productivity� in Mexico declined in 1997 when the ruling party lost the majority in the 22 It is also asserted that a majority of legislators, especially the members of the Lower House, represent “traditional elite� interests and fail to champion progressive causes. But for the topic of decentralization reform, socioeconomic support bases of the legislators are a secondary concern since the stake is mainly over control of fiscal resources between the LGUs and the national government and among the LGUs. 23 It is often asserted that legislative passage is a product of political compromise between the president and a majority in the congress and that the most frequently used tool for building such compromise is the president’s discretionary uses of budget release authorities. Precisely because there is no political party capable of disciplining elected members who stray from the party line, the president could buy votes from individual members of congress to garner enough support for passing a particular piece of legislation. 20 Senate and in 2000 when the party in power changed (for the first time in 71 years) and yet the winning party did not capture a majority in the congress, the bill enactment rates still remained 10 times higher than in the Philippines. One of the reasons is because the well-organized parties (Mexico today is essentially a solid 3-party system) could coordinate legislators’ law-making activities to reduce transaction cost involved in having a large number of individual legislators file bills separately. Even the bills introduced by the opposition have much better rates of being enacted into laws than the bills in the Philippines. Table 9: Bill Enactment Rates in the Mexican House of Deputies, 1991-2006 Year Government bills Enactment rate Opposition bills Enactment rate (Passed/Introduced) (Passed/Introduced) 1991-94 133/154 86.4% 16/89 18.0% 1994-97 90/103 87.4% 18/148 12.2% 1997-2000 43/118 36.4% 94/488 19.3% 2000-03 115/326 35.3% 160/881 18.2% 2003-06 72/400 18.0% 207/1,573 13.2% Total 453/1,101 41.1% 495/3,179 15.6% Source: World Bank (2007) Another striking difference is the absolute numbers of bills filed in the two countries. In the Philippines, the legislators appear to file an inordinate number of bills despite the historically low enactment rates. Based on interviews of some legislators, it is evident that a primary reason why they file bills despite the near-zero probability of having them enacted into laws is political posturing before their constituents. Some legislators prepare “accomplishment reports� which, among other items, list all the bills they have sponsored. Voters have few reasons why they should care about most of these bills because even though most bills filed, especially in the Lower House, address local matters, few of them touch constituents’ lives directly. Following the long-established tradition of clientelism, it is faster for voters to approach local politicians for specific help when they are in need than waiting for some national legislation to be enacted and begin to have policy effects on their daily lives. Once filed, the bills require aggressive lobbying efforts by the sponsoring legislators to move up the deliberation processes of each chamber. Certain officers of both chambers, such as the chairman of the relevant legislative committee and the Speaker of the House/the President of the Senate yield considerable powers over the fate of individual bills. Individual legislators will have to exert significant personal efforts to follow up with these officers and others if they truly seek to have their bills passed into legislation, a high transaction cost as well as use of political capital vis-à-vis fellow legislators, which no one can absorb very frequently, especially given limited pressure from their constituents makes such efforts less electorally worthwhile. Not only are Filipino legislators relatively poor at having their bills passed, those bills they file appear to be extensions of local constituency work rather than attempts to craft national policies. The data in Table 8 show that, except for the 12th Congress, far more local application bills (i.e., those intended to address matters specific to certain localities without nationwide implications) were filed and hence approved in absolute terms, indicating the legislators’ tendency to use national legislation to cater to their local constituents rather than addressing national issues. This contrasts with data on legislative 21 activities in the US state legislatures, where 80% of the bills dealt with state-wide (in 1997), as opposed to district-specific issues, even though local constituency services are considered hallmarks of legislators’ activities in the United States, where the term “pork barrel� politics was originally coined.24 An implication is that the Philippine Congress as a whole is less likely to take up a matter such as decentralization from a broad national perspective than from a parochial local perspective. If so, any legislative change, if it were to prosper against the collective action problem highlighted here, is more likely than not to take the form of aggregation of certain parochial matters (e.g., tax exemptions in specific localities as in the case of the entertainment tax exemptions mentioned above) that do not create zero-sum conflict among the LGUs. In Mexico the Executive has the option of submitting a bill directly, in addition to having its ruling party propose one. In contrast, the Philippine Constitution does not allow the government to submit bills for legislative considerations other than a budget proposal. This, in theory, limits the president’s agenda- setting power. In practice, when the executive wishes to present a legislative proposal of its own high priority, it can do so via a member of the House sympathetic to the administration and certify specific bills as priority bills. Available data indicate that president’s priority bills have unusually high enactment rates. The enactment rates of the priority bills in the Philippines are higher than that in Mexico under a fully democratic period (2000-06). This suggests that the Philippine president could exercise a reasonable level of leverage vis-à-vis legislators to have priority bills passed, perhaps by signaling the importance of certain bills and thus allowing individual legislators to overcome their collective action problem. This data, however, should still be taken with caution. Table 10: Enactment Rates of President’s Bills at the House of Representatives Congress /Administration President’s Priority Bills President’s Priority Bills Enacted Enactment Rate 8th / Aquino 93 54 58.1% 12th / Arroyo 20 8 40.0% Source: Kawanaka (2010) First, higher enactment rates may come at a price. The frequently made assertion about the politics of legislative process in the Philippines is that the president needs to “pay off� individual legislators to mobilize their votes in favor of his/her priority bills because of the absence of party discipline (Kasuya 2008, Kawanaka 2010). For that reason it is necessary for him/her to be selective of which bills to push. A related concern often voiced by both observers and some political actors, though, is that the process of negotiating legislators’ support sometimes entails not only side payments (e.g., preferential releases of the “pork barrel� budget) but also some substantive compromise in the content of the bill itself. On matters related to decentralization, some are weary that pushing for a technically sound bill, say, to reform the IRA distribution formula, might end up being altered into something far less acceptable in the end. 24 The data on the US state bills is from G. Gamm and T. Kousser (2010) “Broad Bills or Particularistic Policy? Historical Patterns in American State Legislatures,� American Political Science Review Vol. 104, No. 1, pp. 151-170. 22 Second, the higher enactment rates may simply reflect the president’s strategic choice to certify only those priority bills with reasonable prospects of garnering sufficient legislative support. After all, the legislators should not be expected to vote against their interests even if modest side payments are offered and if the president’s certification serves as a signaling device to overcome the coordination problems among individual legislators. This brings us to consider the second, and more straightforward, hypothesis that reforms of the LGC have been rare and sporadic because of the fundamental interest incompatibility among key political actors with say over the matter. Interest incompatibility among key actors Apart from the national legislature’s weak capacity to overcome collective action problems among its members, there are other more obvious reasons why legislative reforms are difficult to advance in the Philippines. Reforms may be rare simply because key actors whose consent (in the form of legislative votes) is required are not in agreement with each other about the content of the changes proposed (or whether any change is needed in the first place). A discussion above already alluded to the likelihood that some proposed changes would create conflict among the LGUs by, for example, redistributing the IRA resources away from a group of LGUs to another group. Barring some unforeseen extraordinary circumstance, we judge that the likelihood is low that the Congress will be able to overcome this fundamental interest incompatibility among the LGUs and enact a law that drastically changes the distribution of powers and resources among them. Depending on how the issue is framed, however, the LGUs collectively may be able to form a united stand in favor of certain reform options. For example, all LGUs may be in favor of a measure to increase the share of the IRA in the NG tax revenues from the current 40% to a higher level. As discussed above, bills have been introduced in both chambers of the congress to raise this share to 50% or 60% with support from both the League of Cities and the League of Provinces without changing the current distribution formula. However, the national government is expected to oppose such a measure and although the matter is ultimately up to the national legislators to determine, the executive will likely exercise its considerable influence over individual legislators to thwart such a measure under the prevailing political conditions. A key question here is whose influence or interest, the national executive’s and the LGUs’, matters more to individual legislators’ voting decision. The historical origin of the LGC enactment suggests that the national congress as an institution is cautious about empowering the LGUs’ resource bases. This is presumably because fiscally stronger LGUs depend less on individual national legislators for financial assistance and hence would result in loss of political leverage for members of the congress. The LGUs (i.e., the local chief executives) are interested primarily in expanding their resource base but national legislators, especially members of the House, are typically alleged to be interested in keeping the LGUs’ fiscal capacities constrained so as to maintain their own financial and political leverages over the LGUs within their districts. Especially if more resources were made available to provinces, governors could emerge as strong political rivals, 23 more so than they are already, at the provincial level.25 Exceptions may be those members of Congress who “have their feet in both camps� (i.e., those who themselves have been LCEs and may even expect to return to those posts once their congressional terms are over, those who have family members running LGUs). These congressmen may be more inclined to supporting legislation to strengthen LGUs’ fiscal capacities with the expectation that one day they themselves or their family/relatives may benefit from such a change. It turns out, however, that only 39 (less than 15%) of the house membership of the 14th Congress were former LCEs.26 In the Senate, 7 of its 24 members have been a local chief executive.27 In addition, 58 of the House members and 6 Senators had relatives serving as local chief executives.28 Table 11: Congressmen who are former LCEs or with LCE Relatives Category Representatives Senators a. Former LCEs 21 1 b. Former LCEs with LCE Relatives 18 6 c. With Relatives Only 58 6 Total 97 13 Finally, another source of support for LGUs’ interest could be found among those members of the Congress who contemplate running for local elected posts. For the 2010 elections, a total of 49 House members vied for local elected posts – of these, 22 were former LCEs and were counted in Table 11 already, leaving 27 as House members who were expected to be interested in empowering LGUs and did not have any family ties with current LCEs and had themselves not been LCEs. In the Senate, only one member chose to run for a local election. Therefore a simple tally of these House members who might have been expected to be sympathetic with a reform proposal favorable to LGUs (e.g., increasing the LGUs’ share in the national taxes from 40% to 50/60%) would be 117 (97 from Table 11 plus 27 HOR members vying for local elected posts) if we also include those vying for local legislative posts. These numbers constitute only around 40-44% of the total house membership, just shy of a simple majority. 25 In strongly federalist states, such as Brazil, state governors exert considerable influence not only over municipal mayors within the state but also over national legislators elected from the state (both the lower house and the senate in the case of Brazil). It has been asserted that the Philippine Congress that enacted the LGC was weary of such a possibility and deliberately kept the provinces’ resource bases weak. 26 Fourteen of them have been provincial governors and 23 city/municipal mayors. Two of them have served in both positions. 27 Four were provincial governors and three were city mayors. 28 For lack of historical data, we assume this distribution is more or less typical of the other previous and future congresses. 24 Table 12: Distribution of Congressmen by positions aspired for in 2010 elections Position Aspired for in 2010 Classification of Congressman LCE Post Local Legislative Senate President/ (Governor or Post Vice President Mayor) (Vice Governor or Vice Mayor) HOR Members a. First Term 9 1 1 - b. Second Term 2 1 2 - c. Third Term 30 6 3 - HOR Members Total 41 8 6 - Senate Members a. Mid-Term (elected 2007-13 - - - 4 b. End Term (elected 2004-2010) 1 - 4 4 Senate Members Total 1 - - 8 Source: Congress The analysis here has focused on the presumed career-related incentives of national legislators as a basis for imputing their likely stances on legislative proposals related to LGUs. But obviously there is no mechanistic correlation between these career-related variables and the actual voting behavior of the legislators. For example, 13 of the 62 co-authors of the HB No. 3708 have never been an LCE, have no relative running LGUs and did not run for local elections in 2010. Some individual legislators join in the co-authorship of a bill out of personal interest/conviction or political convenience and it is impossible to predict a precise number of such legislators ex ante.29 Likewise, legislators who are found to have career incentives related to LGUs cannot always be counted on to support a bill to strengthen LGUs’ fiscal capacities because in some specific circumstances doing so would amount to strengthening the hands of incumbents who happen to be their political rivals. Interviews with selected legislators and others close to the discussion (e.g., representatives of the League of Provinces of the Philippines, LPP) have revealed a range of political calculations the stakeholders make with respect to the relatively simple proposal of increasing the LGUs’ shares in the national taxes. Interviewees frequently talked about a tactic of getting the increase approved legislatively in exchange for some politically desirable support LGUs could offer, such as support for the President’s constitutional change initiative. The notion that national politicians, either the president or legislators, would cater to LGUs’ demand rests on the premise that local politicians hold key to assets or resources that are valuable to national politicians. Typically, it is argued that national politicians are dependent on local politicians for mobilizing votes because of the absence of a well-organized political party that can maintain vote-getting machinery. Therefore, to the extent a particular local politician (elected or not) can mobilize a large number of voters reliably, his/her ability to extract concessions 29 nd For example, Cong. Rufus Rodriguez (Cagayan de Oro, 2 Dist.) attributes his decision to support the bill to his learning about the plight of LGUs from his legal scholarship. Rodriguez, apart from authoring four bills related to th IRA in the 14 Congress, has also authored a number of books related to local governance. 25 from the national government should be higher.30 But a counter-argument is that local politicians are more dependent on national politicians (and mayors on governors in some cases) for financial help, not only for funding expenditures in their jurisdictions but perhaps more importantly their own electoral campaigns.31 Some members of Congress also feared that the increasing the LGUs’ IRA share might come at the expense of decreased PDAF allocation.32 Despite the obvious need to be cognizant of these subtleties, the crude count of legislators based on their ties to LGUs can still serve as an informative, though clearly imperfect, first cut at a stakeholder analysis on the prospects for legislative reforms of the decentralization frameworks. If the balance of forces in the 14th Congress was more or less representative of the distribution of preferences among members of both houses regarding matters related to inter-governmental relations in other time periods, then the “LGU alliance� appears to lack votes to push a legislative measure through even if they were able to overcome the collective action problem and take a united front. Summary and Implications The historical analysis of the origin of decentralization, the international examples, and the institutional analysis of the legislative dynamics all point to the conclusion that a major overhaul of the Local Government Code, or any of its fundamental features such as the IRA distribution formula, is highly unlikely under the prevailing political economy conditions in the Philippines. Like in the cases of decentralization reforms in some other countries including Indonesia in the region, an extraordinary political circumstance facilitated the launch of the decentralization process in the late 1980s-early 1990s. But such a circumstance is unlikely to repeat itself in a foreseeable future. In two decades since the passage of the LGC, the Philippines’ political system has demonstrated limited abilities to introduce incremental changes to the decentralization framework – of the more than 700 bills related to decentralization filed in both houses, only four have actually been enacted into laws (and they mostly dealt with issues tangential to the core of decentralization reforms). Quite a few of the bills have been filed on some fundamental features of the LGC, but a review of their content and the extent to which they have garnered support in Congress indicates that on most topics legislators are far from forging consensus with each other. The limited prospects of legislative reforms are due to (at least) two enduring factors of the Philippine politics. First is the atomistic nature of its national congress with the limited capacity to overcome collective action problems among its members. Strong presidential leadership is about the only basis upon which legislators could coordinate their actions to enact major legislation, but exercise of presidential leadership is likely to be selective both on the basis of the administration’s policy priorities (which may or may not include refinement to the decentralization framework) and on the basis of its 30 One of the best-known examples is the Amapatuans in Maguindanao who “delivered� 12-0 slate for the administration candidates in the 2007 senate race even though most of the administration candidates lost the elections in the nationwide contest. 31 A congressman asserted “some LGUs are even funded by the congressmen, by the governor. There are no certainties that they can win on their own. We cannot generalize it… Yes, there are so many mayors that are there because of their congressmen or their governors.� 32 An interview with a congressman. 26 perception/judgment on the political viability of certain legislative proposals. Coordination of legislative behavior by the president is said to rely heavily on selective use of patronage at his/her disposal (but the “selectivity� still needs to cover necessarily a large group of legislators to secure sufficient votes) and the president’s ability (or political capital) to deploy this resource is finite. In other words, the president could influence voting by national legislators, but he must also “pick a fight� he can win. Given the absence of strong societal demand for a decentralization reform – the process of the LGC passage appeared to be dominated by politicians, both national and local, and a few organized interest groups whose interests were directly affected by the proposed change such as the teachers’ unions, it appears unlikely that the administration will choose decentralization as its priority reform agenda. The other reason why a legislative reform of the decentralization framework looks unlikely is the discord among key political actors, particularly national legislators and local politicians. Even when the LGC was passed, national legislators (especially lower house representatives elected from sub-provincial electoral districts as opposed to senators elected nation-wide) showed reluctance to empower local politicians (especially provincial governors). The standard interpretation is that national legislators, especially members of the House, rely heavily on their ability to dispense patronage through “pork barrel� funds at the local level to maintain their influence, including by financing electoral campaigns of local chief executives. Giving LGUs greater resource bases would weaken national legislators’ relative influence vis- à-vis elected local officials. Besides, some national legislators, especially members of the House, see some LCEs, especially provincial governors, as their rivals and are loath to adopt measure to embolden their political (and fiscal) positions. Some national legislators “wear two hats� in the sense that they have been or are expected to have direct personal stakes in strengthening LGUs either because they themselves have been local chief executives and may intend to return to those posts or because they have members of the family running LGUs. Others may be supportive of a decentralization reform because of their personal conviction or for political opportunism. It is obviously not possible to predict exactly how many of the national legislators would vote in favor of certain reforms, but even if we assume that those House members with personal or familial ties to LGUs will vote in favor of a simple measure such as the proposed increase of the LGU share of the IRA from the current 40% to 50% or 60%, this particular group does not constitute a simple majority in the House. The likelihood that the president could muster a sufficiently large voting block for a more complex reform proposal that creates a zero-sum conflict for resource allocation among LGUs such as a revision of the IRA distribution formula is even lower. All these findings point to the conclusion that investing in advocacy of a major structural reform of the existing decentralization framework that requires a legislative change is likely to yield limited results even for a new administration with full political capital and eagerness to improve governance. 27 References Dillinger, William and Steven B. Webb (1999) Decentralization and Fiscal Management in Colombia. World Bank Eaton, Kent (2009). “The Political Economy of Decentralization in Peru,� mimeo. Diokno, Benjamin (2003). Decentralization in the Philippines After Ten Years? What Have We Learned? What Have I learned? UPSE Discussion Paper 2003-08 Gamm, G. and T. Kousser (2010). “Broad Bills or Particularistic Policy? Historical Patterns in American State Legislatures,� American Political Science Review Vol. 104, No. 1, pp. 151-170 Garman, Christopher, Stephan Haggard and Ezra Willis (2001). “Fiscal Decentralization: A Political Theory with Latin American Cases,� World Politics Vol. 53, No. 2, pp. 205-236 Hutchcroft, Paul D. (2004) “Paradoxes of Decentralization: The Political Dynamics Behind the Passage of the 1991 Local Government Code of the Philippines,� in Thai Politics: Global and Local Perspectives, KPI Yearbook No. 2 (2002/03) by Michael H. Nelson (ed.). Nonthaburi, Thailand: King Pajadhipok’s Institute, 283-332 Kasuya, Yuko. (2009). Presidential Bandwagon: Parties and Party System in the Philippines. Anvil Publishing, Inc. Kawanaka, Takeshi. (2010). Interaction of Powers in the Philippine Presidential System, IDE Discussion Paper No. 233 Khemani, Stuti. (2011). Buying Votes vs. Supplying Public Goods: Evidence from Public Health Services in Philippine Municipalities (mimeo). Leoni, Eduardo and Lucio Rennó (2006) “Reelection and Fiscal Responsibility Law: Reducing Populist Pressures in Brazil.� (http://svn.cluelessresearch.com/representation/reelection.pdf) O’Neill, Kathleen (2003) “Decentralization as an Electoral Strategy.� Comparative Political Studies 38(9): 1068-1091 Schneider, Aaron (2006) Governance Reform and Institutional Change in Brazil: Fiscal Responsibility and Tax. Institute of Development Studies Discussion Paper 384 Sidel, John. (1999). Capital, Coercion & Crime: Bossism in the Philippines. Stanford: Stanford University Press World Bank (2005). East Asia Decentralizes: Making Local Government Work. Washington, DC: World Bank World Bank (2007). Democratic Governance in Mexico: Beyond State Capture and Social Polarization. World Bank 28