Chapter

Chapter 10: Philippines

Author(s):
International Monetary Fund. Independent Evaluation Office
Published Date:
April 2003
Share
  • ShareShare
Show Summary Details

Introduction

1. The Philippines is probably the most extreme case of prolonged use of IMF resources, with 23 programs between 1962 and 2000. Over the 30-year period 1971–2000, the Philippines had programs for almost 25 years (Table 10.1) with credit outstanding from the IMF continuously since 1967.1 The Philippines’ status as a “prolonged user” was recognized by the Executive Board as early as 1984.2

Table 10.1.Philippines: Chronology of IMF Arrangements Since 19671(In millions of SDRs)
Type of arrangementDate of arrangementOriginal expirtion date2Actual expiration date or cancellationSize of arrangementPercent drawnParis Club debt restructringCommercial bank restructring
SBAJan. 1967Jan. 196855.0100.0
SBAMarch 1968Mar. 196927.5100.0
SBAFeb. 1970Feb. 197127.5100.0
SBAMar. 1971Mar. 197245.077.8
SBAMay 1972May 197345.077.8
SBAMay 1973May 197429.00.0
SBAJuly 1974July 1975May 197538.8100.0
SBAMay 1975May 1976Apr. 197629.199.9
EFFApr. 1976Apr. 1979217.0100.0
SBAJune 1979Dec. 1979105.086.9
SBAFeb. 1980Dec. 1981410.0100.0
SBAFeb. 1983Feb. 1984315.031.7
SBADec. 1984Dec. 1986June 1986615.065.5Dec. 1984Jan. 1986
SBAOct. 1986Apr. 1988Aug. 1988198.0100.0Jan. 1987Dec. 1987
EFFMay 1989May 1992Feb. 1991660.635.7May 1989Jan. 1990
SBAFeb, 1991Aug. 1992Mar. 1993334.2100.0June 1991Dec. 1992
EFFJune 1994June 1997Mar. 1998791.2100.0July 1994
SBAApr. 1998Mar. 2000Dec. 20001020.876.7

2. This study presents an assessment of the nature of the IMF’s prolonged involvement with the Philippines and attempts to draw lessons for the future. The evaluation is based on an extensive review of published and unpublished IMF documents and interviews with (i) current and former Philippine officials and a range of other stakeholders, including academics, NGOs, and representatives of the private sector, undertaken during an IEO mission to Manila in March 2002; (ii) current and former IMF staff; and (iii) staff of the Asian Development Bank, the World Bank, and some bilateral donors. The study is organized as follows: the second section presents a brief overview of the experience over three decades; the third section examines the factors that led to prolonged use; the fourth section focuses on implications for program design and implementation; the fifth section considers the implications of prolonged use for IMF surveillance; and the final section presents conclusions and recommendations.

Overview of the Philippines’ Experience with IMF-Supported Programs

3. The extent of continuous involvement by the IMF in supporting programs in the Philippines is reflected in Table 10.1, which lists the various standby arrangements (SBAs) and Extended Fund Facility (EFF) programs since 1967.3 Rather than provide a detailed evaluation of each program, we attempt a summary assessment of the experience in four distinct periods.

The Marcos period prior to the 1982–83 debt crisis: almost continuous programs that achieved no lasting adjustment

4. In this period there were a total of 10 SBAs and one EFF in 15 years but this near continuous involvement with the IMF did not prevent a widening of external imbalances and a sustained debt buildup that eventually resulted in a crisis.

5. The deterioration in economic performance and vulnerability is evident from Table 10.2. Economic growth was relatively strong during the 1970s, averaging 6 percent a year, but it fell to about 2½ percent in 1981–83. After approximate balance in the first half of the 1970s, a large current account deficit emerged from 1975—averaging 5½ percent of GNP in the second half and rising to 8 percent in 1981–83. The higher deficit was financed by rising external debt, most of which was directed to financing large-scale investment projects that subsequently proved to be of low efficiency. Programs struggled to deal with the aftermath of the liquidity expansion associated with the 1973 increase in commodity prices and a subsequent deterioration in the terms of trade later in the decade. The two oil price shocks of 1973 and 1980 also added to these challenges. A major structural weakness that developed over the period, and was to plague the Philippines for a long time, was a fall in gross national saving. This deterioration partly reflected a worsening fiscal balance that changed from a surplus of 0.6 percent of GNP4 on average in the first half of the decade to a deficit of 3½ percent in 1981–83.

Table 10.2.Philippines: Selected Economic Indicators(Period average, in percent of GNP, unless otherwise noted)
1971–751976–801981–831984–851986–881989–921993–971998–2000
Real GNP growth (in percent)6.16.12.6–8.05.33.54.92.9
Export growth (in billions of U.S. dollars, in percent)15.920.72.8–3.815.28.620.814.7
Inflation (CPI, annual average, in percent)17.012.310.934.45.313.07.96.9
External current account–0.4–5.4–8.1–2.10.2–3.4–4.97.91
External debt (in billions of U.S. dollars)2.510.823.026.028.329.340.350.5
Debt-service ratio (in percent, after rescheduling)23.433.533.039.734.722.916.814.8
National government fiscal balance0.6–1.3–3.4–1.9–3.5–2.2–0.3–3.7
Consolidated public sector balance–4.5–7.2–3.8–3.3–0.6–3.7
Tax revenue10.312.110.39.911.314.216.113.9
Gross national savingn.a.27.723.417.519.519.818.224.51
Gross investment16.615.619.823.417.9
Sources: IMF staff reports and WEO database.

6. Several sources of structural inefficiency also prevailed in this period. A system of monopolistic practices allowed political favorites to extract economic rents especially in key agricultural marketing areas (sugar, coconuts, bananas, etc.) that weakened incentives and harmed small producers, who were among the poorest groups. Inadequate governance and supervisory controls allowed widespread lending to insiders by financial institutions, including publicly owned banks. Several Article IV staff reports did discuss the distortions caused by the monopolistic practices, but—reflecting the general approach of IMF-supported programs at this time—there was little effort to address these issues in any of the programs prior to 1984. Some structural measures to make the economy more export oriented were implemented in 1970–72, including liberalization of exchange controls and tariff reforms. However, the main structural component of the 1976 EFF—an effort to improve the tax effort—failed.5

7. The rapid buildup in external indebtedness in the second half of the 1970s (Figure 10.1), which in part reflected excess liquidity in the industrial countries, eventually resulted in a serious economic and financial crisis culminating in a standstill on external debt service in October 1983. The standstill was accompanied by widespread failure of domestic banks and corporations, and a subsequent recession. The discovery that the authorities had misstated the international reserves position exacerbated the crisis and contributed to a program agreed in early 1983 going quickly off-track. Although staff reports show an awareness of the dangers of the debt buildup, programs were not able to address in a systematic way the factors that were leading to it.6, 7 Moreover, the risks to the quasi-fiscal deficit implied by a series of exchange rate guarantees and forward cover arrangements were not fully recog-nized and eventually contributed to significant central bank losses when the exchange rate depreciated sharply following the crisis.

Figure 10.1.Philippines: External Debt

Sources: IMF staff reports.

8. It is interesting to note that economic performance deteriorated and vulnerabilities increased despite the fact that most of the specific performance criteria in the various programs were satisfied and 94 percent of the amount committed in arrangements between 1968 and 1981 were disbursed. This suggests that the problem was not just one of program implementation.

9. Staff and former staff interviewed who were involved with the Philippines during this period, or in its immediate aftermath, were generally of the view that the extensive IMF involvement for much of the Marcos era had been a mistake, sending the wrong signals to markets and effectively helping to postpone adjustment. This was also the view of most of those with whom the evaluation team discussed the issue in the Philippines. A number of officials as well as some staff and former staff also expressed the view that, until 1983, political factors—that is, the close relationship between the Philippines under the Marcos regime and the United States as well as, to a lesser extent, some other major IMF shareholders—were important in explaining the lengthy IMF involvement during this period. We have not found any documentary evidence to support such a judgment, but the fact that such a view is held by a number of those who were close to the negotiations is, in itself, significant.

10. However, it has to be recognized that the IMF was dealing with a very difficult situation—epitomized by deep-seated governance problems and a lack of commitment to fundamental reform at the highest political level—at a time when it did not have adequate instruments, or even an explicit policy, for addressing such issues.8 While greater attention at an early stage to the structural and institutional problems discussed above might have helped, it is quite likely that program design was not the core issue. In these circumstances, it would have been better for the IMF to have refrained from lending.

The debt crisis and subsequent prolonged debt workout, 1983–93

11. In the 10-year period following the onset of the debt crisis, the Philippines had to undergo a protracted adjustment to correct the initially large external imbalances and to deal with the accumulated debt problem. There were five programs in this period, spanning the last year of the Marcos administration, the whole of the successor Aquino administration, and the early days of the Ramos administration.

12. As noted above, the 1983 SBA quickly went off-track and, after an interval of nearly a year, was followed by the 1984–86 SBA, the last of the Marcos administration, which prescribed tight macro-economic policies combined with floating of the exchange rate. The program also made a significant effort to address some of the major structural problems, most notably to dismantle the sugar and coconut monopolies, to establish a more elastic domestic tax system to replace taxes on trade, and to rehabilitate public financial institutions and strengthen control over public sector investment decisions. The program went off-track because of policy slippages in the run-up to the February 1986 election. Fiscal policy turned sharply expansionary and implementation of reforms to the tax system and the sugar and coconut sectors and import liberalization were delayed, with several measures reversed. The program did achieve a sharp reduction in inflation, while there was an even greater current account adjustment than envisaged under the program. This sharp adjustment was driven by the cutoff in private external financing and involved a large contraction in imports and in economic activity. Real GNP fell by a cumulative 17 percent during 1984–85, compared with the projected 7 percent fall.

13. The Aquino administration negotiated three arrangements with the IMF: an initial SBA in 1986, which was successfully completed; an ambitious, growth-oriented EFF covering 1989–91 that quickly went off-track as a result of major policy slippages amid considerable political instability and severe exogenous shocks (Box 10.1); and finally, in 1991, a less ambitious SBA—successfully completed—that focused on stabilizing the economy in the run-up to the May 1992 election, which was won by President Ramos.

14. The programs negotiated with the Aquino administration increased the focus on structural reforms with greater emphasis on dismantling the administrative and other restrictions that had hindered private sector activity and exports, distorted prices, and favored inefficient industries and vested interests. While continuing to aim at improving public sector revenue and the elasticity of the tax system, the tax reforms envisaged were more far-reaching than earlier reforms, paying greater attention to the structure of incentives and equity. There was initial success on reforms covering trade, taxes, and the agricultural sector, especially in those areas that represented a dismantling of the “crony capitalism” privileges of the Marcos era and the creation of a more level playing field. However, progress then slowed, as the administration tackled more politically sensitive structural issues.

15. The realignment of incentives for exports, arising from the exchange rate depreciation and trade liberalization in the 1980s, led to a sharp improvement in export performance as reflected in the growth rate of over 15 percent a year in the period 1986–88. Rapid export growth, combined with the series of reschedulings of commercial bank and official debt, contributed to a significant reduction in the Philippines’ external debt problems. Indicators of the debt and debt-service burden started to improve markedly in the late 1980s with the debt-service ratio (after rescheduling) falling to below 20 percent in the early 1990s and to 14 percent by the mid-1990s. The debt/GNP ratio fell from a peak of 100 percent in 1985 to less than 60 percent in 1992.9

The 1994 EFF: restoration of access to private financial markets and the planned “exit” from IMF programs

16. From the time of the accession of the Ramos administration in 1992, staff and management planned a final EFF explicitly intended to be an “exit program” from prolonged use. In the view of staff, a recurrence of external financing problems could not be ruled out if the economy were to resume sustained growth of 5 percent a year or more and further structural changes were needed to ensure medium- term viability. A new three-year EFF was agreed in June 1994 after lengthy negotiations with the Ramos administration. In the interim, the Philippines had regained access to international capital markets when it floated a $150 million Eurobond issue in 1993. It also resumed servicing Paris Club debt, although it would soon seek further rescheduling.

Box 10.1.Why Did the 1989 EFF Fail?1

One important development that prolonged the stantial-length of the Philippines’ adjustment was the failure of spect, the 1989 EFF to achieve its objectives. The three-year tratedprogram was intended to be growth oriented: (i) the targets for investment and saving were ambitious, while the current account deficit was projected to increase modestly rather than decline; (ii) the fiscal program allowed for a substantial increase in public infrastructure investment, to be financed by higher public saving but also by a modest increase in the fiscal deficit in the first year; and (iii) a wide range of structural reforms, including financial sector strengthening, trade reform, privatization, power tariff reform, foreign investment and exchange market reform, elimination of the Central Bank deficits, and land reform. Unfortunately, actual performance fell well short of these objectives. Rather than going through a consolidation phase with rapid growth, the economy slipped back into significant vulnerability: growth declined; inflation accelerated; and both the fiscal and current account deficits widened sharply.

Three sets of factors accounted for the disappointing performance:

  • Adverse exogenous shocks, in the form of considerable political instability with several coup attempts; natural disasters (including a drought, typhoon, and severe earthquake); and an unanticipated 10 percent weakening in the terms of trade. While these adverse shocks put the short-term balance of payments and growth objectives out of reach, they were temporary in nature and, by themselves, should not have caused the program to fail in its longer-term objectives.2 But the degree of political instability that occurred would have been hard to predict.

  • Program design issues, including: (i) A structural reform agenda that proved to be overambitious and insufficiently prioritized, especially in the context of a newly restored Congress that wanted a sub- stantial say in economic policy matters. In retrospect, it would have been better to have concentreated on a shorter and more specific agenda.3 (ii) An overoptimistic projection of the speed of recovery in private investment, reflecting an apparent underestimation of the adverse impact of the debt overhang on investment.4 (iii) The program rightly placed more emphasis on measures to improve tax administration and compliance rather than on discretionary measures. However, this posed challenges for program design since the revenue impact was uncertain in terms of both magnitude and timing—and progress proved slower than expected. (iv) The move to six-monthly performance criteria appears to have been a mistake. It permitted the buildup of large policy slippages that proved too large to correct by the time of the next test dates. Quarterly performance criteria were reinstated in June 1990.

  • Weaknesses in program implementation were probably most important of all, although the timetable of reforms would have been difficult to deliver even without the disruptions caused by the coup attempts. The Government’s ability to implement the program was hampered by its inability to pursue its economic policy agenda—particularly revenue and structural reform measures—in Congress, a constraint that had been insufficiently recognized when the program was prepared. Some actions by Congress that were opposed by the Government—notably approval of a large wage increase and a cut in oil taxes—seriously aggravated the fiscal situation.5

1 This box draws in part on internal ex post staff assessments, which were quite candid, but tended to put even more emphasis on implementation failures.2 IMF staff estimates suggest that the natural disasters accounted for about half of the deterioration in the external current account, mostly financed through foreign emergency assistance.3 For example, internal assessments by the staff suggest that the lack of specificity in the program on action to deal with the central bank deficit contributed to the lack of progress in this area.4 Overoptimistic projection of investment appears to be a common problem in program design. See Goldsbrough and others (1996).5 Implementation problems began at the outset of the program when the authorities indicated that the oil price increase agreed upon as a prior action would have to be delayed. Combined with an unanticipated increase in world oil prices, the delay had a significant fiscal impact. Oil price issues remained highly sensitive, since the December 1989 coup attempt was timed to coincide with the date of the postponed price increase, in order to exploit expected social unrest.

17. The 1994 EFF aimed at fiscal consolidation; tax reform; financial sector reform; and continued economic liberalization, particularly in the fields of the deregulation of oil pricing and the abolition of remaining quantitative import restrictions (Box 10.2). In order to meet program objectives, it was judged necessary to improve the saving/investment balance significantly, and this was to be achieved through a substantial fiscal adjustment. The program aimed at increasing tax revenue by nearly 4 percentage points, to 19 percent of GNP, over the three-year period by broadening the tax base, reducing exemptions, and improving tax administration.

Box 10.2.Original Aims and Outcomes of the EFF, 1994–96

AimsOutcomes
Macroeconomic targets
Increase investment by 3 percentage points of GNP.In 1996, investment was 23.1 percent of GNP, slightly lower than in 1994.
Reduce the current account deficit to about 3 percent of GNP from 6 percent of GNP.The current account deficit was 4.6 percent of GNP in 1996.
Increase national saving by 6 percentage points of GNP by 1996, almost all through fiscal adjustment.By 1996, the saving rate had risen just 0.8 percentage points.
Reduce consolidated public sector deficit (CPSD) from 2.7 percent of GNP in 1993 to 0.5 percent of GNP by 1996.Outturn in 1996 was a surplus of 0.2 percent of GNP.
Reduce underlying CPSD 1 from 4.2 percent of GNP in 1994 to 0.7 percent of GNP by 1996.Outturn was 0.2 percent of GNP.
Raise government tax revenue by 4 percentage points of GNP between 1993 and 1996.Tax revenue rose just 1 percentage point.
Increase real GNP growth from 2.3 percent in 1993 to 5 percent by 1996.Real GNP growth exceeded the target, rising to 7¼14; percent in 1996.
Key structural measures
Tax reform, including adoption of a minimum business tax, phased abolition of tax exemptions, and improved VAT administration.Comprehensive tax reform was eventually passed in December 1997 but major tax exemptions and incentives remained. Minimum business tax based on sales, not assets. Tax administration weaknesses remained.
Deregulation of oil pricing and importation.Deregulation of oil prices was completed in early 1998 and the Oil Price Stabilization Fund abolished.
Tariff reduction and elimination of nearly all quantitative import restrictions.Quantitative import restrictions were removed (except for rice). Average tariff was reduced to 15 percent.
Public expenditure reform.Legislative action on rationalizing civil service stalled.
Local government devolution.Devolution of public expenditure did not match large increase in transfers to local government.

18. An interesting feature of the EFF-supported program is that the fiscal targets of the program were met—mainly because of lower interest payments and lower public investment than targeted—al-though the aim of increasing tax revenue was not achieved. Tax revenues rose by only 1 percent of GNP over the period. Thus it was possible to meet the fiscal targets even though the structural strengthening on the tax side had only partially occurred, leaving the economy vulnerable on this front when economic circumstances changed.

19. Considerable progress was made on some of the structural measures included in the program. Under the Ramos administration, banking, telecommunications, domestic shipping, and the oil sector were opened to new competition, and limits on foreign participation were eased in a number of sectors. Import tariffs were reduced markedly, and quantitative import restrictions largely eliminated. The privatization program was accelerated, and the central bank recapitalized under a new statute. However—as discussed in the section “Implications for Program Design and Implementation"—progress in implementing tax reform proposals was slower and less complete than envisaged. The 1994 EFF was associated with improved export performance and a modest improvement in growth, but it did not bring about the sustained current account adjustment originally envisaged. Until the Asian crisis, however, the current account deficit was easily financed by buoyant capital flows.

20. In retrospect, it can be questioned whether the 1994 EFF was needed—in a balance of payments sense—and we take up this issue in the section “Why Did the IMF Remain Involved for So Long?” In any event, since financing problems eased subsequently, the authorities began to treat the EFF as precautionary after the initial purchase—indicating their intention to refrain from further purchases. As the fourth and supposedly final review approached in early 1997, staff briefing papers noted that the authorities were looking forward to a graceful exit from a long line of IMF arrangements, but delays in the progress on the tax re-form because of opposition in Congress meant that the review could not be completed as scheduled. Rather than let the program lapse, the authorities requested an extension as they wanted to “graduate” on a successful note, following passage of the Comprehensive Tax Reform Package (CTRP) by Congress. The original extension of the arrangement was not therefore driven primarily by a perception of balance of payments need, but rather the desire to end the program successfully. However, before this could happen, the program was overtaken by the Asian crisis.

Asian crisis and its aftermath

21. As the Asian crisis hit, the EFF was augmented and further extended first until December 1997 and then to March 1998 and the authorities again decided to draw on the program. The peso was floated in July 1997.10 Initially, the authorities showed no interest in a follow-on arrangement, in part because graduation from IMF-supported programs had become a political issue in the Philippines. However, as the regional crisis unfolded, both the authorities and IMF management became concerned about the risks of such an exit amid the turbulent financial market conditions. The IMF management was anxious to have a “follow-on” program to help ring-fence the Asian crisis and also to serve as a policy blueprint for the new administration which would take over following the May 1998 presidential election. Fund management therefore pressed for a precautionary Stand-By Arrangement prior to the elections.11

22. The new two-year SBA entered into with the outgoing Ramos administration centered on fiscal and monetary tightening, banking sector reform, improved tax administration, and other structural reforms in order to provide a framework for orderly adjustment to the reduced level of capital inflows. The program included a new emphasis on tax administration, as well as an effort to complete the original CTRP agenda, by reintroducing legislation to reduce tax exemptions and incentives. A revised version of the program12 was endorsed in October 1998 by the incoming Estrada administration. The program was originally intended to be precautionary but, at the time the revised program was approved, the authorities decided to make purchases in response to reduced access to international capital markets following the Russian default and collapse of Long Term Capital Management.

23. At first, the momentum of the policy envisaged in the program was maintained, but over time policy slippages grew and concerns over governance problems mounted. Policy slippages, related in part to the privatization of the Philippine National Bank, led to a delay of almost a year in completing the Fifth Review until mid-2000.13 This review was the occasion of a critical discussion in the Executive Board, which focused on persistent policy slippages, governance concerns, and the rationale for continued IMF program involvement. Some Executive Directors were highly critical of the IMF’s prolonged and, in their view, ineffective relationship with the Philippines and raised the issue of the future relationship after the conclusion of the arrangement. A few suggested that the program should become precautionary after the drawing that was the subject of the review.

24. The authorities noted that their earlier exit strategy had been frustrated by the Asian crisis and that they intended to develop a new one. They argued, however, that the reserve position remained somewhat vulnerable, implying a need for further IMF support. Staff took the view that, since the IMF’s presence in the Philippines had encouraged the implementation of sounder macroeconomic policies and had had an important catalytic effect, a close policy dialogue should be maintained with the authorities, while encouraging an exit from use of IMF resources. In the event, the 1998 SBA was extended to end-2000, but it expired without the last review being completed, as a result of further fiscal slippages, reflecting weak revenue collection and mounting governance concerns. Thereafter, the authorities requested post-program monitoring in line with the expectation of the Board for a country whose credit outstanding exceeds a threshold amount.14

25. In defense of the long involvement of the IMF, and particularly its continued involvement in the mid-1990s, it could be argued that, although not all aspects of the 1994 EFF were implemented as planned, it helped the economy to weather the Asian crisis fairly well. The Philippines did not suffer as severe a downturn as its neighbors; growth decelerated but remained positive. This was partly the beneficial consequence of the cumulative impact of structural changes that had been introduced over the years under successive programs, and which accelerated as a result of strong political leadership during the Ramos administration. The generally sound macro-economic management under the framework of the 1994 EFF15 probably also helped, although, as already noted, there continued to be differences of view about exchange rate policy. A number of the vulnerabilities that were central to the severity of the dislocation in other economies most affected by the Asian crisis were also present in the Philippines, nonetheless.16 However, imbalances did not build up over as long a period as they had in other Asian countries, in part because earlier problems meant that capital inflows were smaller and did not last for as long prior to the crisis as in the other countries.

Despite some progress, poverty remains high

26. The incidence of poverty has declined from 45 percent in 1985 to 32 percent in 1997, but remains high in absolute terms. Moreover, income distribution has not changed much in the last three decades and remains among the most unequal in South East Asia: in 2000, the richest 20 percent accounted for over one-half of total expenditure and the poorest 20 percent for about 5 percent.17 Significant progress was made in improving health and education indicators over the period of IMF program engagement, but these improvements did not match those in other Asian economies, in part neglecting fiscal policy shortcomings.

Why Did the IMF Remain Involved for So Long?

27. In this section, we explore the main reasons why the IMF became as closely and continuously involved as it did in the Philippines over the past 30 years or so and whether the involvement was consistent with Board guidelines on the subject.

Starting imbalances and structural problems were large

28. Part of the explanation for prolonged involvement lies in the fact that despite the IMF’s long engagement in the Marcos era, imbalances were large when the standstill on external debt service was declared in 1983. The current account deficit reached about 9 percent of GNP in 1982–83. External debt peaked at 100 percent of GNP in 1985, which was higher than in most of the other problem debtors of the time. Debt service, before rescheduling, was equivalent to about 50 percent of exports of goods and services and private transfers.

29. The Marcos administration also left its successors a daunting legacy of structural problems, including severely indebted state corporations and government financial institutions, including the central bank.18 A small elite controlled a large part of the economy and proved resistant to many reforms that threatened this control.19 As a result, the economic system remained quite closed, with patronage playing a dominant role with a limited scope for competition. Many industries benefited from a high degree of protection and there were barriers to the entry of foreign corporations. The Philippines was much more closed than other ASEAN economies even though the economy was smaller in size.

30. Moreover, staff reports were quite candid in indicating that it would take a long time to return to external viability. For example, the staff appraisal in the Request for SBA in 1984 notes that “effective adjustment will undoubtedly require time, stretching well beyond the program period.” Medium-term projections in subsequent program documents continued to imply that the restoration of external viability would take a long while, typically extending well beyond the projection period. Indeed, with hindsight the IMF was at times too pessimistic about the timing of the return to external viability, reflecting the difficulty of predicting when full access to private financial markets would be restored.20

“Seal of approval” influences on IMF program involvement

31. “Seal of approval” considerations, linking the provision of financing or restructuring of debt from other creditors to an IMF-supported program, were particularly important during the period 1982–92 as the Philippines sought to work out its external debt problems. IMF programs were formally required in order for Paris Club creditors to reschedule debt. Commercial banks also made it clear to the Philippine authorities that an active IMF-supported program was the sine qua non for the concerted commercial bank lending and restructuring that took place between 1986 and 1992. Apart from the various commercial bank restructurings in 1986, 1987, and 1990, the release of intermediate tranches of new money by commercial banks were also dependent on IMF-supported programs. This remained so until the final Brady restructuring in 1992, which was financed in part from “set-asides” under IMF arrangements together with support from JEXIM, the Japanese export credit agency, and from the World Bank—both of which were also tied to the programs.21 The timing of donors’ Consultative Groups was also closely related to the inception of IMF-supported programs.

32. Did the debt workout phase in the Philip-pines—and IMF involvement related to this factor—take longer than in most other countries that encountered debt problems in the 1980s? The answer appears to be no, in large part because the broader international strategy vis-à-vis the middle-income highly indebted countries took about a decade to evolve to the Brady plan. Therefore, most of these countries also had a series of programs lasting a decade or so, linked to debt restructuring (Box 10.3).

33. Even after the Philippines had regained market access in 1993 and IMF-supported programs were no longer required specifically for concerted debt arrangements, they were still seen to satisfy the demand for a positive signal on the macroeconomic framework to donors and private financial markets. In this situation, donors did not formally require an IMF arrangement to be in place in order to provide support, but they appear to have been more comfortable with such an approach. For example, former officials stated that one of the primary rationales for the 1994 EFF was that the authorities felt that, in the light of the Philippines checkered track record, an IMF-supported program would be helpful to reassure lenders as the country started to reaccess international capital markets. Both IMF staff and Philippine officials appear to have been concerned that a discontinuation of the prolonged relationship would be interpreted negatively by participants in the international capital markets. The EFF was therefore viewed as desirable in part because so many players in the system had become accustomed to the existence of a program.

Box 10.3.Many Highly Indebted Countries Took as Long as the Philippines to Resolve Their Debt Problems and Consequently Also Had Repeated IMF Arrangements

The Philippines is not unusual in the length of time it took to complete the debt workout after the debt crisis of the early 1980s. Most of the heavily indebted middle-income countries that were affected also took about a decade to work out their debt problems with the help of Paris Club reschedulings and commercial bank restructurings, culminating in the Brady restructurings of the early 1990s. Of the 17 highly indebted middle-income countries,1 Philippines was the eleventh to complete its series of reschedulings with commercial banks and the Paris Club, in July 19942 (Table 10.3). Six of these countries completed their restructurings in a concentrated period between 1992 and 1994, with five earlier and six later.

In most of these cases, countries had a succession of IMF-supported programs lasting a decade or more linked to the need for a seal of approval for restructuring. Up until the 1991 SBA tied to the 1992 Brady deal and the 1991 Paris Club rescheduling, the Philippines was fairly typical in the length of programs associated with restructuring. However, to the extent that the abortive Paris Club rescheduling was one of the rationales for the 1994 EFF, which lasted until 1998, IMF involvement linked in some way to restructuring lasted a little longer than in most other countries.

1 The “Baker 15” countries, plus Costa Rica and Jamaica. Colombia is not shown because it did not reschedule in the relevant period, although it did benefit from concerted lending arrangements with commercial banks in 1985, 1989, and 1991.2 If this Paris Club rescheduling, which was never implemented, is discounted the Philippines would rank as the sixth of the 17 to complete rescheduling. Nevertheless, the Paris Club rescheduling was one of the rationales for the 1994 EFF although a number of Executive Directors questioned its need.
Table 10.3.Heavily Indebted Middle-Income Countries: IMF-Supported Programs Tied to Debt Restructuring Following the 1980s’ Debt Crisis
CountryFirst program linked

to 1980s’ debt

restructuring
End of last IMF

program linked to 1980s’

debt restructuring
Last

commercial bank

restructuring
Last Paris club
ArgentinaJan. 1983Mar. 1996Apr. 19931July 1992
BoliviaFeb. 1980OngoingJuly 19921July 2001
BrazilMar. 1983Aug. 1993Apr. 19941Feb. 1992
ChileJan. 1983Nov. 1990Dec. 1990Apr. 1987
Costa RicaMar. 1980Feb. 1994May 19902June 1993
Ecuador3July 1983Dec. 1995Feb. 19951June 1994
JamaicaJune 1978Mar. 1996June 1990Jan. 1993
MexicoJan. 1983May 1993Mar. 19901May 1989
NigeriaJan. 1987Apr. 1992Jan. 19921Dec. 2000
PeruJune 1982Mar. 1999Nov. 19961July 1996
PhilippinesDec. 1984Mar. 1998Dec. 19921July 1994
UruguayApr. 1983Mar. 1992Feb. 19911...
VenezuelaJune 1989Mar. 1993Dec. 19901...
Côte d’IvoireFeb. 1981OngoingMay 19971Apr. 2002
MoroccoOct. 1980Mar. 1993June 1990Feb. 1992
Former YugoslaviaMay 1979Sep. 1991Sep. 19884July 19884
Sources: Paris Club; Global Development Finance database; and IMF Annual Reports.

34. Internal briefing papers and interviews with a number of staff suggest that the strong “gatekeeper” role implied by the close linkage of programs to other forms of financing has had a potentially adverse effect on the quality of programs, in two ways: (i) the severe consequences, in terms of loss of other financing, of the IMF’s potential disengagement probably caused it to be less selective in choosing to be involved only at times when a lending arrangement was likely to be most effective in advancing re-form; and (ii) since the seal of approval from IMF-supported programs has typically been linked primarily to agreement on a program, rather than to its successful implementation, it increased pressures on the authorities to agree to policy timetables that looked strong on paper, even if they were unlikely to be delivered within the agreed time frame.

35. These tensions were frankly discussed in internal briefing papers prepared when the 1994 EFF was under negotiation, at which time staff expressed a concern to resist being pressured into early agreement on a less than adequate program. The authorities were apparently keen to agree on a new program relatively quickly in order to pursue a Paris Club rescheduling and an early meeting of official donors, both of which were linked to such a program. This wariness on the part of the staff appears to reflect a number of earlier episodes where seal of approval considerations influenced IMF agreement on programs, and where, in the staff’s judgment, these programs suffered from weaknesses as a result.22 However, these internal concerns were not reflected in the Board papers for the 1994 EFF.

Fluctuations in political commitment and ownership

36. One of the reasons why adjustment took a long time is that there was strong political resistance by vested interests to implementing some key structural changes essential for success.23 The Philippines’ political system required a strong political commitment of both the executive and legislative branches to en-sure the passage of enabling legislation to support the economic policy agenda. However, the programs supported by the IMF were based primarily on agreements between staff and the executive branch, and important elements of the reform agenda often stalled in Congress or ran into judicial challenges.

37. Strong leadership, with ownership of reform, did at times help achieve the necessary political commitment and overcome vested interests—such as in 1986–87 and during much of the Ramos administration. During these periods, much was achieved under the IMF-supported programs, even if the need to mobilize a political consensus and push reforms through Congress meant that the ambitious timetables set out in program documents could not be attained. However, when political leadership was weaker, reforms stalled despite the commitment of the direct economic ministries and the Central Bank. Prolonged political uncertainties—including several coup attempts during the Aquino administration—added to these problems.

38. What could the IMF have done differently during these periods of weaker ownership? In many cases, there is no clear answer. In principle, a different approach might have involved a combination of three elements: greater efforts to build consensus, greater selectivity, and changes in conditionality.

39. On the first element, IMF missions did engage in extensive dialogue on economic policy with members of key congressional committees and both officials and staff interviewed thought this dialogue had been helpful. Some missions also reached out to “civil society” more generally, and resident representatives generally engaged in a policy dialogue with a broad range of groups, within and outside government, to explain the IMF point of view. However, such contacts could not create, or serve as a substitute for, broader domestic political commitment and ownership.

40. Greater selectivity in entering into, or extending, lending arrangements would have been highly desirable during those periods when it became evident that there was no commitment at the highest political level to reforms necessary for sustainable adjustment. This was the case for much of the Marcos era and should have been evident earlier in the Estrada administration. At other times, however, when the issue was more one of the Executive’s ability to implement measures they agreed to, the IMF was faced with a more difficult situation. On a number of occasions when policies went off-track, the IMF did halt programs—by not completing reviews—while seeking corrective actions (Box 10.4). However, partly because of the “seal of approval” issues discussed above, pressures often grew to resume the program and a compromise was generally reached that involved some erosion of the original policy measures.

41. Finally, could a different approach to conditionality have helped overcome the periods of weak ownership and thereby reduced the length of IMF program involvement? As will be discussed further in the section “Implications for Program Design and Implementation,” a greater prioritization of structural reforms—with greater flexibility on the timetable but not on the substance of those judged to be critical to macroeconomic sustainability—would probably have helped; however, as the later discussion of the efforts to improve the tax structure and strengthen tax administration will show, there are limits to what any specific modalities of conditionality can be expected to achieve, in the absence of strong domestic ownership.

Program design and implementation

42. Several key weaknesses were consistently identified by the IMF as at the core of the longer-term adjustment problem—notably weak tax collection and low public saving, which in turn contributed to low levels of gross national saving in comparison with the Philippines’ neighbors. This was long seen as a factor that limited the ability of the economy to invest and grow without running into periodic external financing constraints. Nearly all IMF-supported programs over the last 30 years identified increasing saving rates as a central aim, and sought to address this problem by improving public saving through a greater tax effort, which was also low in comparison to other economies in the region. However, programs were not consistently successful in achieving these goals, either because of overoptimism of the original projections or policy slippages.

43. As shown in Figure 10.2, programs generally targeted a large sustained improvement in tax revenues as a share of GNP as a means of boosting public and gross national saving.

Figure 10.2.Philippines: National Government Tax Revenue

(In percent of GNP)

Source: IMF staff reports.

44. Tax reform and efforts to improve administration did yield some benefits in raising the tax ratio in the late 1980s and early 1990s, although progress was intermittent. The tax ratio rose from 11 percent of GNP in 1986 to 15 percent in 1993 and peaked at over 16 percent in 1997. However, except for the 1989 EFF and to some extent the 1991 SBA, actual performance still fell far short of the sharp increases in tax revenues targeted under successive programs. Perhaps more important than this overoptimism was that part of the improvements that were achieved were the result of ad hoc measures. The increase in tax revenue in proportion to GNP also owes something to the recovery in activity in the late 1980s. In any event, the revenue gains were not sustained. Performance was particularly poor after 1998, when the tax ratio deteriorated significantly. These issues are discussed in more depth in the section “Improving the tax structure and strengthening tax administration: an example.”

45. The gross national saving rate was also projected to rise markedly in each program but the actual outturn usually fell short (Figure 10.3). Gross national saving as a percentage of GNP declined continuously in the first half of the 1980s, improved briefly between 1986, and then deteriorated again. The measured saving rate shows a sharp increase in 1999 and 2000, but there is reason to believe that this may simply reflect statistical weak- nesses.24 The failure of successive programs to increase the saving rate in turn left the economy vulnerable to stop-go cycles and to exogenous shocks.25

Figure 10.3.Philippines: Gross National Savings

(In percent of GNP)

Source: IMF staff reports.

Box 10.4.Examples of Major Interruptions to IMF-Supported Programs in the Philippines

As discussed in the main text, conditionality in the Philippines—especially structural conditionality—re-lied heavily on program reviews. On a number of occasions, program reviews were delayed for extended periods—sometimes leading to permanent interruptions—as policies diverged from program commitments. By interrupting programs, the IMF was sometimes able to secure desired policy changes. However, if such interruptions took place within a context where it was tacitly under-stood by both sides that outside factors—including “seal of approval” considerations—would be likely to ensure the eventual restoration of a program relationship, this may have affected the efficacy of efforts to impose greater selectivity. This box examines several such episodes to examine the major reasons for program interruptions and to assess the extent to which the policy slippages were corrected by the time programs were resumed, or new arrangements agreed. The broad conclusion is that the program interruptions were partly effective in correcting policy slippages, but that substantial “recontracting” also occurred, resulting in a weakening of original commitments.

Interruption of the 1989 EFF

The IMF took contrasting approaches to successive reviews under the 1989 EFF.

  • The first review of the 1989 EFF was completed on schedule in December 1989, despite significant slippage in monetary and fiscal policies that largely reflected the delayed implementation of politically sensitive oil price increases and a large public sector wage increase. The authorities eventually raised oil prices at the beginning of December, and promised to take remedial fiscal actions, including several discretionary tax increases, to permit the re-view to be completed. As noted below, these promised increases did not materialize. One motivation for completing the review on time was to free up IMF “set-asides” and linked financing from donors to help finance the “buy-back” of commercial bank debt that took place in January 1990.

  • The program went more seriously off-track during 1990 in light of policy slippages and a series of large exogenous shocks. Net international reserves targets and fiscal targets were missed, the latter largely because Congress failed to pass the tax measures promised earlier as a condition for completing the first review.1 Staff decided that the original aims of the EFF were unattainable and that the second review of the program could not be completed. This decision appears to have reflected the judgment that in the political environment following the earthquake and the previous year’s coup attempt, it was unlikely that the government would be able to take the comprehensive and strong policy measures that would justify the resumption of IMF support.

  • However, the IMF was keen to head off the risk of an intensifying crisis and also to free up needed financing from donors and creditors. Six months after the review was originally scheduled, agreement on a new Stand-by Arrangement was reached in January 1991. In order to meet the schedule for a donors’ group and avoid the damage to confidence that a further delay would cause, the IMF compromised on its initial requirements for a new program, accepting the President’s assurance that she would use veto powers and administrative means to keep expenditure in line with the program, as a substitute for full enactment by Congress of the budget. Reflecting political sensitivities, revenue measures were confined to a temporary import levy, rather than the measures promised under the previous program. Other prior actions were agreed, including the elimination of administrative foreign exchange arrangements that artificially supported the peso and the announcement of a power tariff increase. In the event, the politically sensitive power tariff increase was not implemented until a year later, and was subsequently partly rolled back. The band around the reference exchange rate for commercial transactions was eliminated as scheduled. More far-reaching reforms of the exchange system were introduced only in early 1992.

Delayed first review under the 1994 EFF

  • Policy performance in the first six months of the 1994 EFF deviated substantially from program expectations. In particular, the introduction of an automatic oil price adjustment mechanism, which was an original condition for completing the re-view, was postponed until late 1995, after it was determined that this would require legislation. Monetary policy also went off-track. There were also delays in establishing a simplified tariff regime, revising the National Power Corporation’s pricing formula, and liberalizing entry into the oil sector. To complete the review, the IMF therefore pressed for the implementation of these trade and energy reforms and for demonstrated adherence to the monetary program in early 1995. It also sought new understandings on compensatory measures that would be taken if there were unforeseen developments, such as weaker-than-expected capital flows, following the Mexican crisis.

  • The review was eventually completed in September 1995, after a ten-month delay. The IMF settled for the administration’s submitting of bills covering the measures required for completion of the re-view to Congress, rather than their actual implementation. The administration submitted a bill to remove quantitative agricultural import restrictions, which was passed in March 1996, and a comprehensive oil sector reform bill, as well as a tax reform bill.2 The decision to complete the re-view appears to have reflected the strong macro-economic performance, recognition that the authorities had limited influence over the Congressional timetable, and the judgment that the authorities’ efforts to advance structural reform deserved encouragement.

  • An Oil Deregulation Bill was passed in 1996, but was subsequently judged to be unconstitutional. Replacement legislation was eventually passed in February 1998, some two and a half years later. The Comprehensive Tax Reform legislation was passed in December 1997, but suffered some important shortcomings (discussed in the section “Improving the tax structure and strengthening tax administration: an example”).

Fifth review of the 1998 Stand-By Arrangement

  • Completion of the fifth review of the 1998 SBA, originally scheduled for September 1999, was delayed by ten months. As a condition for completing the review, the IMF sought renewed momentum in key structural reforms, with specific prior actions required in banking reform, tax administration, and power sector restructuring. Missions in September and December 1999 were unable to complete the review because of slippage in the fiscal program and because key structural reforms were delayed—particularly power sector reform legislation and the adoption of a plan for the privatization of the Philippine National Bank (PNB) by mid-2000. Growing governance concerns, including those related to the PNB privatization, also added to difficulty in completing the review.

  • The program was extended beyond its original termination date of March 2000 to permit completion of the review in July 2000. This followed agreement on a fiscal program for 2000, incorporating a slightly higher targeted fiscal deficit3 as well as approval of the power sector reform by the Lower House of Congress in April (a prior action), and by the Senate in June. In the event, approval of unified legislation was delayed for much longer—until late 2001. The review went ahead despite the failure of the auction of the government’s stake in PNB in June. Progress was slower than the IMF originally sought in some other areas of structural reform, including tax administration, but a General Banking Act was passed prior to the completion of the re-view. The arrangement was extended a second time until end-2000, but was ultimately permitted to lapse amid increasing governance concerns, without completion of the final review.

1 In the aftermath of the June earthquake, the authorities were pressing for the early provision of emergency assistance from the IMF while Congress threatened a debt moratorium. Congress also voted a substantial reduction in oil taxes in the face of the run-up in world prices, and a tariff reform was withdrawn in the face of congressional pressure.2 There was also a major revision to monetary policy, as the authorities adopted a limited form of inflation targeting.3 To 2.9 percent of GDP compared to an original target of 2.2 percent. In the event, the actual fiscal deficit for 2000 was 4.6 percent of GDP.

Adverse shocks

46. Bad luck—in the form of a long string of adverse shocks—also played a role in extending the period of prolonged use. A dramatic series of adverse supply shocks in the early 1990s (including an extended drought and related power shortages, a severe earthquake, and a damaging volcano eruption) and deteriorating terms of trade adversely affected the balance of payments, while political uncertainties—reflected in several coup attempts—affected investor confidence. Later in the 1990s, the timing of the Asian and Russian crises, and the associated contagion, disrupted what would otherwise probably have been a successful “exit” from use of IMF resources in 1997.26

The rationale for program involvement was sometimes too broad

47. The rationale for IMF program involvement in the Philippines appears to have been very broad. The traditional criterion of balance of payments need was clearly evident from the onset of the debt crisis in 1983 until the restoration of market access in 1993 and, once again, during the height of contagion from the Asian and Russian crises in 1997–98. During the 1983–93 period, IMF gross financing typically provided a relatively small proportion of projected financing gaps, but IMF involvement was necessary to catalyze other flows, mainly through reschedulings. Much of the IMF financing in fact represented a rollover of payments falling due.27

48. However, the balance of payments need was less evident from 1993, when market access was regained, until the Asian crisis. It is also doubtful for how long there was balance of payments need after 1998. Admittedly, an element of judgment was involved in such cases, since the justification for access to IMF resources can refer to a potential, as well as an actual, balance of payments need. In both periods, the Philippines probably could have raised from private markets the amounts that were borrowed from the IMF, but it could be argued that, at both times, this market access was not fully assured. However, two points are worth noting in this respect: first, both arrangements were extended at times when there appeared to be even less balance of payments need;28 second, there was very little discussion in internal documents of whether such a need existed. In the Board discussion in March 1993 relating to the augmentation of the then Stand-By Arrangement, some Executive Directors questioned the balance of payments need for the augmentation, and a few noted that they would want to carefully consider the balance of payments need for any successor arrangement. At the Executive Board discussion on the 1994 EFF, a year later, one Director noted that in spite of these reservations about a future program expressed earlier, there was no discussion of these issues in the Staff Report.

49. The broader rationale for continued program involvement at these times seems to have included three factors. First, a desire to send a positive signal to donors and creditors (discussed above). Second, the belief that it would foster “good” policies and bolster internal groups in favor of reform. A number of staff interviewed argued that the IMF’s continued presence in a program context, even when there was no pressing balance of payments need for IMF financing or a catalytic role, was still an important pressure in favor of the “right” policies in key periods. In their view, policy implementation and economic management would have been weaker without a program. A number of staff and former officials also noted that the existence of a program had helped to increase the leverage of internal groups that favored a strong macroeconomic policy framework and continued reforms.

50. Finally, one of the major motivations for the 1998 SBA according to internal briefing papers and interviews with staff and former Philippine officials, was to sustain earlier gains and avoid “backsliding” during a change of administration. In the event, while the program helped macroeconomic policies remain prudent during the transition period itself, this was not sustained and the approach ultimately proved unsuccessful. Although the incoming President signaled his agreement to the program, ownership eventually proved weak and implementation suffered, especially after the first year. (The discussion of efforts to strengthen the tax system, in the section “Implications for Program Design and Implementation,” gives a specific example of how a detailed set of measures negotiated with an outgoing team was not implemented because they were not effectively “owned” by the incoming team.)

Sponsorship by key shareholders

51. International political factors probably also have their place in explaining the IMF’s long program engagement in the Philippines, particularly in the 1960s and 1970s. In the view of a number of officials and staff, up until about 1983, the close relationship of the United States—and some other share-holders—with the Marcos regime in the context of the cold war and regional tensions appears to have been reflected in the favorable attitude of IMF management and the Executive Board to the authorities’ requests for support. Subsequently, tensions in the political relationship between key shareholders and the then Philippines leadership may also have contributed to delays in agreeing on a program in 1983 and an increase in the threshold of what would be viewed as an acceptable program. The interests of key shareholders in regional stability and their support for the efforts of President Aquino to restore and maintain more democratic institutions were probably also a factor influencing decisions to continue program involvement at times when weak ownership of some core economic reforms might otherwise have argued for greater selectivity.

Borrower incentive because of low cost

52. Some external observers have argued that continued use of IMF resources by countries that have access to private financial markets is driven by the borrowing countries’ incentives to reduce their borrowing costs, by substituting lower-cost IMF financing.29 It is not possible to test this proposition directly, but the impact on the Philippines’ overall borrowing costs, if it had ceased to use IMF resources in 1994 and from 1999 onward, would have been small. During the 1994 EFF the Philippines only made the initial SDR 36 million drawing before the Asian crisis, treating the program as precautionary for the most part. The relative cost of borrowing from the IMF and the markets might have been a more relevant consideration in 1999–2000. During this period, the Philippines borrowed SDR 491 million from the IMF. At the time, issue spreads on the Philippines’ sovereign international bonds ranged from about 350 to 425 basis points over U.S. treasuries of comparable maturity—implying an interest saving of about SDR 20 million a year.

53. Officials and staff interviewed were all of the view that a more important factor driving the author-ities’ interest in continued programs, both during and after the restoration of market access, was the belief that there was a catalytic role of an IMF-supported program—that is, the signal it sent to other creditors and donors.

Did the IMF’s approach in the Philippines match its guidelines on prolonged use?

54. Whatever the reasons explaining prolonged use, and there are many, it is relevant to ask whether such use in the Philippines was in accord with the underlying policy governing prolonged use as articulated in Board discussions. This policy consists of the following elements (see Chapter 3 of Part I for a more detailed discussion): (i) front-loading the adjustment effort, including greater emphasis on prior actions, and closely monitoring program implementation; (ii) seeking a net reduction in prolonged users’ outstanding liabilities to the IMF, including through arrangements with limited access but that would still serve a catalytic role, and back-loading disbursements; (iii) an analysis of the factors underlying a country’s prolonged use and a candid ex post assessment of performance under previous programs at the time of new program requests; and (iv) an “exit” strategy including ex ante assessments of the time needed to complete the adjustment process and of the time frame to disengage from IMF lending along with strengthened surveillance at the post-program stage.

55. In the Philippines’ case, some elements of this strategy were implemented and some were not.

  • (i) It is hard to argue that the adjustment effort was “front-loaded” in practice and, as will be shown in the next section, prior actions were used sparingly. For example, fiscal adjustment and the program of tax reform in the 1994 EFF were substantially back-loaded, although expansion of the coverage of the VAT was made a prior action,30 with most of the planned fiscal adjustment to occur in 1995 and 1996, after the Task Force on Tax and Tariff Reform had formulated more detailed plans for tax reform. However, as the discussion below of the various efforts to strengthen the tax effort will show, the precise structure of conditionality, including prior actions, does not seem to have been a critical factor.

  • (ii) Access under arrangements did in general seek a net reduction in the Philippines’ liabilities to the IMF, while preserving a catalytic role, but disbursements were not always back-loaded.31

  • (iii) There was not a systematic ex post assessment of each program, although the staff did “step back” on several occasions to conduct such assessments (e.g., in 1991, 1993, and 2000)—more so than in the other country cases examined in the evaluation. However, internal assessments were much more candid than those in final Board papers.

  • (iv) While parts of an “exit” strategy were set out, and early Board papers were reasonably forthcoming about the time frame involved, the rationale for the IMF’s continued program involvement once market access was restored was not fully presented in Board papers, nor does it seem to have been subject to much internal debate. In any event, the strategy was overtaken by the Asian crisis. A comparison with Morocco, which exited from 10 years of IMF-supported programs in 1993 after extended adjustment and debt restructuring, is in-teresting. The key macroeconomic indicators for the two countries, discussed in more detail in Part I, were quite similar by the early 1990s. While it would not be desirable to pre-establish simplistic quantitative “exit” criteria for a prolonged user, it is not obvious why such different approaches were taken in the two cases.

56. The latter two points raise a more general question about whether the internal review process played its role sufficiently. While any summary of a review process spanning several decades is bound to involve some generalization—especially since guidelines on the review process changed over time—an assessment of internal briefing papers, country strategy papers, and review department comments on those papers suggests the following points. First, the diagnosis of the main problems and the content of the proposed policy framework was typically the subject of a detailed internal assessment and often considerable debate. Second, several of the ex post assessments mentioned above did draw potential lessons for program design (e.g., the 1991 assessment pointed to the need for greater prioritization and a more realistic time frame on structural reforms), but these lessons were not always fully absorbed or debated in the subsequent review process (see the next section for more discussion of the time frame of program design). Third, many of the often quite sharp differences of view about the overall strategy that emerged during the review process were not reflected in subsequent Board papers, which tended to focus on an explanation and justification of the agreed consensus approach and the program as negotiated.

57. The period leading up to the 1994 EFF provides a good example of the process. From the early 1990s, the IMF did have an exit strategy for the Philippines, including an ex ante assessment of the time it would take to complete the adjustment process. From the time the 1989 EFF went off-track, the consistent game plan was that an SBA, primarily focused on stabilization, would be followed by an EFF, which would include the major structural changes that would permit the Philippines to “graduate.” There was considerable internal debate, ahead of agreement on the EFF, as to how strong—particularly in terms of fiscal adjustment—any successor program would have to be to be to warrant IMF support. However, there was not a full discussion of time frame and ownership/implementation issues, even in the draft 1993 country strategy paper that aimed to draw lessons from earlier experience. The area department was of the view that one lesson was that an unrealistic time table and overcrowded agenda had contributed to implementation difficulties, whereas review departments generally took the view that strong up-front actions were needed in light of the Philippines’ perceived poor record of program implementation. Nor was there an explicit discussion of whether any follow-up program was needed at all.32 Although PDR implicitly questioned the balance of payments need, by drawing attention to the tenuous nature of the small financing gaps assumed, it apparently did not argue against an arrangement per se. In retrospect, the country strategy paper—which was never completed—was a missed opportunity for a broader debate on the lessons from past programs and the IMF’s role.33

58. As noted above, the critical nature of the Executive Board discussion at the time of the fifth re-view of the EFF did trigger a more comprehensive assessment of the IMF’s role and approach in the Philippines, in the form of the internal (2000) country strategy paper.

Implications for Program Design and Implementation

59. This section discusses issues relating to program design and program implementation, which arise from our review of the Philippines’ experience and which suggest lessons to help avoid some of the problems associated with prolonged use.

Program design

60. A recognition of the longer time frame required for adjustment could have led to a greater prioritization of structural reforms.

61. As noted earlier, the IMF recognized at an early stage that the Philippines would require a relatively long time to restore external sustainability, but program design remained focused on a relatively short time horizon. The short-term focus was partly a reflection of the institutional constraints requiring programs to offer evidence of concrete progress within the time frame of the program itself. In the view of some IMF staff interviewed by the evaluation team, this excessively short-term focus caused programs to address structural issues across too broad a front and to target timetables for various structural reforms that were often unrealistically ambitious. Tax reform and tax administration, examined in detail below, is one example where, in the view of many staff and officials, program design would have benefited if a longer-term perspective had been adopted from an early stage.

62. Program design may also have paid insufficient attention to political implementation capacity. The case of the 1989 EFF which went off-track is one example (Box 10.1). Several IMF staff noted that the programs on which they worked would likely have been more effective if the program had been focused on a smaller number of core structural reforms. For example, in the latter stages of the 1994 EFF, the IMF was simultaneously pushing for reforms to the oil-pricing system and to tax policy, each of which required congressional approval, as prior actions for the completion of program reviews. In the view of some staff, this may have been overambitious, exceeding the capacity of the political system to digest several major reforms at the same time.

Difficulties in strengthening key institutions were critical

63. The issue of tax administration has featured in virtually all IMF-supported programs for the Philippines, dating back to the 1970s.34 As discussed below, several different approaches were tried in successive programs, including efforts to promote ownership of these reforms through local working groups to develop detailed measures. With hindsight, while programs incorporated a wide range of measures to improve tax administration and policy, based on extensive technical assistance, they failed to achieve the strengthening of institutions, particularly the Bureau of Internal Revenue, which was critical to achieving program aims.

64. The effectiveness of the Bureau of Customs was greatly strengthened, at least temporarily, under the Ramos administration, as a result of reforms—and strong leadership of the agency—initiated in 1993.

65. Could the IMF have done more to strengthen tax institutions? It did make repeated efforts to tackle the issue—as the discussion below will illustrate. It would likely have been more effective if the IMF had focused on a limited number of measures that were critical to achieving success, along with greater on-the-ground follow-up to technical assistance missions to monitor and facilitate program implementation, and a greater willingness to halt programs when implementation of these core measures was inadequate. However, the experience may indicate the limitations of what conditionality from an outside agency can reasonably be expected to achieve, unless there is strong commitment from the country in question.

Dealing with uncertainty in programs

66. Some staff interviewed also believed that program design would have benefited from more explicit attention to contingency planning. The Philippine economy has remained vulnerable to exogenous shocks. Moreover, the revenue impact of certain tax administration measures was highly uncertain and the program would have benefited if contingency actions to be introduced in the event of revenue shortfalls had been specified at the outset.

67. In practice, program reviews were generally used to adjust programs in the light of unexpected developments, which was probably the correct approach, but the adjustments often did not occur in a sufficiently timely fashion. It would likely have been more effective if programs had paid more attention to identifying up-front the major risks—including implementation risks—along with a more systematic stress-testing of how the program framework would be affected by such risks and a discussion of how policies would respond.

68. However, better contingency planning does not require a detailed ex ante quantification in the letter of intent of responses to specific shocks. For example, the Philippine authorities did request access to contingent financing under the Compensatory and Contingency Financing Facility (CCFF) in association with the 1989 EFF. The view of both staff and officials was that the CCFF experiment had not been helpful. The effort consumed considerable time and administrative resources since the formulas determining eligibility were very complex and hard to check, and the mechanism proved excessively rigid and overconstrained as it simultaneously tried to satisfy a number of inconsistent considerations. As a result, the focus was more on rigid formulas rather than on the broader logic of policy adjustments.

Program implementation

Implementation of IMF-supported programs has been mixed

69. The effectiveness with which programs have been implemented has varied markedly, with phases of strong implementation—for example 1987–88, and to some extent 1991–92—alternating with periods when implementation has been much weaker. There have also been phases, such as during the 1994 EFF, when economic performance was strong and significant progress was achieved on some structural reforms, even though overall program implementation was inconsistent.

70. Over the period 1970–2000, 8 of the 16 arrangements were fully disbursed, while 77 percent of commitments were disbursed on average, a relatively high proportion by comparison with other countries (see Table 10.1).35 Four of the seven programs since 1983 were not completed. Even those programs that were completed suffered implementation problems as program reviews were subject to frequent long delays, sometimes of as much as ten months.36 This reflected in part the heavy reliance on reviews to provide structural conditionality. For the same reason, the Philippines has not been a heavy user of waivers.37

71. The implementation of the fiscal component of programs has been patchy. For example, deficits were narrowed substantially in the first half of the 1990s—by more than the program targets—but have subsequently widened again (Figure 10.4). Moreover, as noted earlier, revenues were generally overestimated except during the early 1990s. As a result, fiscal deficit targets were often achieved at the expense of shortfalls in planned public spending, particularly on infrastructure investment. Given the typically limited scope of discretionary expenditure, such ad hoc adjustments distorted expenditure priorities and at times prevented planned increases in capital expenditure. For example, under the 1986–88 SBA, public investment remained at 3.5 percent of GNP throughout the program rather than rising to 5 percent of GNP as targeted. Public investment also remained well below program projections under each of the three arrangements in the 1990s.

Figure 10.4.Philippines: Implementation of Fiscal Programs—Consolidated Public Sector Balance

(In percent of GNP)

Source: IMF staff reports.

Impact of prolonged use on domestic policy formulation

72. The close and continuous relationship between the Philippine authorities and the IMF also needs to be evaluated in terms of its impact on institutional development. Current and former officials and staff noted that IMF involvement had a strong influence in shaping the macroeconomic framework as well as the technical and data systems used by the Philippine authorities, which were essentially a mirror image of the IMF’s financial programming framework. In the view of most officials interviewed, IMF training and technical assistance played a constructive role, most notably in enhancing the skills needed for the formulation of a comprehensive and consistent macroeconomic framework.

73. While it is impossible to judge how policy formulation processes would have developed in the absence of almost continuous programs, formulating policies around negotiations with the IMF for more than 30 years must have had an impact. Both the 1993 and 2000 country strategy papers noted that there was a strong risk that the Philippines had become overdependent on the IMF and that this was an important reason why the prolonged program engagement should eventually be discontinued. The 2000 CSP stated explicitly that “Institutional development has likely been hampered in some respects by the dependence of the authorities on the IMF to articulate a consistent policy package, monitor its implementation, and provide discipline in the short-term. This may not have left room for the develop-ment of the authorities’ own institutions or procedures to ensure internal accountability and discipline in policy-making and implementation.” Nevertheless, the Philippines does have considerable strengths in some of its major policymaking institutions—including economic ministries, the Central Bank, and key congressional committees. The fact that, under the post-program monitoring (PPM) arrangement in effect since 2000, the government has continued to set goals and establish targets on key macroeconomic and financial aggregates, to ensure a forward-looking aspect to monitoring, without being required to do so under the PPM, suggests that domestic economic managers have been able to adapt quickly and prudently to the absence of an IMF arrangement.

Prolonged use has probably eroded the effectiveness of conditionality

74. The nature of conditionality in programs for the Philippines has evolved markedly over time, as it has for programs in general. Conditionality in programs during the 1970s centered on quantitative performance criteria related to the financial programming framework and to quantitative limits on contracting medium-term external debt. The breadth of conditionality increased markedly from 1983 onward as programs increasingly addressed structural issues. Conditionality on these issues was exercised mainly through prior actions and program reviews (often in combination). There was limited use of structural benchmarks and virtually no use of structural performance criteria.38 In deciding whether to complete a review, the IMF typically exercised a high degree of flexibility in relation to original program targets, at times introducing new conditions to complete reviews in light of evolving circumstances. Reviews were often substantially delayed, but generally the program remained in existence and reviews were eventually completed.

75. This flexible approach was developed at least partly in response to the Philippines’ congressional-style political system, whereby the Executive has limited control over the legislative agenda and therefore often cannot credibly commit to a specific timetable for measures that require legislation.39 Moreover, the IMF frequently had to decide whether to compromise over shortcomings in measures passed by Congress that deviated from those that were originally envisaged (see Box 10.4 for examples).

76. While the reliance on reviews seems to have been broadly appropriate under the circumstances, there was inevitably a trade-off between the flexibility gained from exercising a high degree of discretion, and the consequences in terms of a lack of clarity for the authorities, markets, and the wider public over the “bottom line” on conditionality. Each review became an occasion for “recontracting” the underlying commitments, which weakened their focus and credibility. In retrospect, greater specificity on a small number of critical outcomes for completion of reviews may have been desirable. In the words of one Executive Director, “including the most important policy commitments as formal conditions... reduces the risk that, ex post, measures used to justify completing reviews are the ones the authorities completed rather than the ones they needed to complete.” One example is the Comprehensive Tax Reform Package (CTRP), the structural centerpiece of the 1994–98 EFF, which is discussed in the next section.

77. Both IMF staff and former officials acknowledged that such factors contributed to a tendency for the authorities to “overpromise” with regard to the timing of structural measures, even when they knew that these commitments were politically unrealistic, knowing that failure to meet their commitment would be unlikely to prompt program cancellation.40

Improving the tax structure and strengthening tax administration: an example

78. As pointed out earlier, virtually all IMF-supported programs have aimed at raising the tax ratio as a key to raising domestic saving. The relatively low tax ratios have in general not been the result of low statutory tax rates, but rather of a narrow tax base and poor compliance in relation to that base. Over the past 30 years, several attempts have been made to increase the tax effort through broadening the tax base and improving tax administration. Since these efforts have taken place against a background of moves to open up the economy, an increasing emphasis has been placed on taxes on domestic economic activity as trade taxes have been reduced.

79. Successive tax reforms have achieved some important gains, including a simplification and improvement of the tax structure by reducing distortions and increasing its progressivity, and have also reduced the dependency of revenue on trade taxes. The tax revenue/GNP ratio was also improved markedly between 1989 and 1997. However, continued weak tax administration and large exemptions and incentives mean that revenues remain low by international standards and inelastic with respect to economic activity (Table 10.4). Moreover, as noted earlier, revenue projections in many programs proved too optimistic and the considerable gains achieved in the first half of the 1990s were eroded after 1998. As will be discussed below, some of these later slippages were related to weaknesses in tax administration and governance and some to shortfalls in the tax reform program.

Table 10.4.Tax Efficiency in Selected Asian Economies
Statutory

VAT rate
VAT effort1

(1994)
Efficiency

ratio1
Indonesia104.80.48
Thailand73.180.45
Korea104.270.43
Philippines103.330.33
Sources: IMF Fiscal Affairs Department and IEO calculations.

80. The Philippines has received extensive technical assistance (TA) in this area, with nine TA missions on tax between 1984 and 2001. Moreover, efforts were made to develop reform strategies suited to the Philippine environment. A number of times,41 working groups were established within the government to develop detailed proposals for tax reform with the help of IMF technical assistance, in the hope of developing proposals adapted to local circumstances and with ownership by the authorities. However, as pointed out earlier, the results have been disappointing.

81. Could the IMF have done more to improve the performance on the tax front? The Philippine experience on this issue suggests that it is not possible to correct for weaknesses in the domestic political commitment through variations in the form of conditionality. However, in retrospect one can identify several issues.

82. One set of problems arose from the authorities’ strong reluctance to accept any form of direct conditionality specifying tax collection as a concrete quantitative performance criterion. Staff proposed to establish a formal performance criterion on tax revenues of the National Government under the 1989 EFF, but this was strongly resisted.42

83. As a result, when tax revenue did not materialize as expected, either because the required tax structure or tax administration measures were not implemented or because other macroeconomic assumptions turned out to be inaccurate, there was a tendency to squeeze expenditure in an effort to meet fiscal targets. Under some programs, stop-gap measures to raise revenue with undesirable efficiency and equity implications—such as the special import levy in 1991—were imposed pending action on more desirable structural reforms.

84. The implementation of conditionality through program reviews was also not fully satisfactory. For example, some of the original targets for structural tax reforms under the CTRP during the 1994 EFF were achieved, but many were not and the IMF progressively agreed to a number of modifications as the authorities encountered difficulty in getting the program through Congress. For example, there was little progress in reducing fiscal incentives for investment, which was an important element of the EFF. The excise bill signed in November 1996 was expected to account for most of the revenue gains of the CTRP, but in its ultimate form it was seriously flawed because in shifting from an ad valorem to a specific basis, in order to make collection more effective, the provision for full indexation was eliminated. This was one of the reasons for the subsequent sharp fall in collection relative to GNP in later years as Congress proved reluctant to increase excise tax rates in line with inflation. Similarly, a minimum business tax based on gross assets for all companies above a certain asset size (which could be credited against corporate income tax due) was originally a central component of the CTRP. The version that was ultimately passed did include an alternate minimum corporate income tax, but it was based on sales rather than assets. Implementation of this tax has reportedly been very weak, partly because of this change in its structure.

85. Weak ownership presented special challenges for the design of conditionality during a transition in administration. The 1998 SBA included a renewed emphasis on tax administration, as a “structural pillar” of the program.43 The Memorandum of Economic Policies, based on the advice of an IMF technical assistance team, contained a matrix of extremely detailed commitments to reorganize the Bureau of Internal Revenue (BIR) and improve its operations, focusing on increasing control over large taxpayers, strengthening audit arrangements, and completion of a computerization project. However, according to interviews with IMF staff, the package was agreed primarily with officials of the outgoing Ramos administration. With the change in administration, and the appointment of a new Commissioner for the BIR, the incoming officials did not necessarily accept what had been agreed earlier and were not necessarily committed to implementing the measures.

86. The detailed “matrix approach” was also not very effective because it was relatively easy to find superficial ways to meet the commitments and because judgments on effective progress could not rely on a checklist of items achieved. For example, the staff report for the Third Review of the SBA judged that “good progress” had been made with tax administration in 1998, whereas a subsequent mission noted that “The implementation of reform measures had slowed down in the second half of 1998, after the new administration took office” and that, while progress had been made in some areas, “it had become increasingly clear that the new administration has not bought into key reform elements such as the control of large taxpayers and the need to strengthen BIR’s audit and enforcement programs.”44

87. Another approach that was tried subsequently to develop a more “owned” program involved a three-day workshop with IMF staff, key officials in the Philippines, and representatives from other agencies, including the World Bank and USAID, on ways of improving tax administration in the Philippines. While this is a potentially promising approach, the action plan developed in the workshop was not effectively implemented because commitment from top political leadership was lacking.

88. A more difficult question to address is whether IMF-supported programs could have done more to address the pervasive corruption that is widely seen to lie at the root of problems with tax administration in the Philippines. A World Bank study published in 200045 identified the BIR as one of the agencies that should be targeted as a priority for anticorruption efforts. IMF-supported programs were not able to come to grips with the impact of corruption on tax administration although programs have from time to time touched on issues of corruption, including through pressing for tax structures that limit corruption by reducing tax collectors’ scope for discretion. More direct targeting of conditionality on governance concerns in the field of tax administration—perhaps accompanied by more specific conditionality on the collection of tax revenue, based on more realistic and conservative projections of the impact of policy changes and administrative measures—might have yielded better results. However, it should be recognized that the IMF’s decision to be more directly involved in governance issues is relatively recent. There are also difficult questions about how much any outside agency such as the IMF can realistically be expected to achieve in such areas.

Continuous Programs Have Limited the Independent Role of Surveillance

89. An important issue is whether surveillance functions are crowded out during periods of prolonged program involvement. In order to examine this issue systematically, the quality of surveillance in the Philippines was assessed in relation to nine “key elements” of surveillance, derived from the IMF’s most recent surveillance guidelines.46 Since the guidelines evolved over time, such a comparison does involve, to some extent, judging previous sur-veillance exercises by current standards. However, the functions highlighted have been part of surveillance requirements, or related guidelines, since at least the early 1990s. One area that has received much stronger emphasis since the 1995 Mexican and 1997/98 Asian crisis is the analysis of potential risks and vulnerabilities. In what follows, we note those areas where more recent surveillance reports have addressed these issues in greater depth.

90. Since 1986, nearly all the staff reports on Philippines’ Article IV consultations have been combined with papers relating to the use of Fund resources. Given the predominance of combined papers, and the almost continuous nature of IMF-supported programs, it is difficult to generalize from the Philip-pines’ experience about the relative adequacy of surveillance in program periods and outside them, and the relative merits of joint and stand-alone consultations. In general, the content and assessment of programs seem to have dominated joint reports. Surveillance reports that were combined with requests for new programs have tended to be relatively stronger in reviewing performance under previous programs, compared with those prepared along with program reviews.

91. All the Article IV staff reports included substantive medium-term scenarios for the economy. However, the related sensitivity analysis and discussion of risks to the projections were often limited to presenting the impact that relatively small changes in export growth, global interest rates, or the oil price would have on the projections, rather than on presenting alternative scenarios that focused attention on the major vulnerabilities. Even in one case (1996) where the text of the paper did a good job of identifying these vulnerabilities, the formal sensitivity analysis only examined the impact of a marginally slower growth in remittances. More recent reports have improved in this respect. For example, the stand-alone staff report for the 2000 Article IV consultation examined the impact of policy slippage explicitly in an alternative medium-term scenario.

92. In general, however, there was little candid discussion of implementation capacity of the authorities in any of the surveillance reports, although from time to time, uncertainties over a program’s prospects for success have been noted. Staff papers typically emphasized in general terms the importance of implementation of the agreed program for the medium-term health of the economy. By contrast, a number of internal documents did contain a substantive discussion of implementation issues.

93. For the most part, presentation of collaboration with the World Bank was confined to the standard proforma appendix setting out World Bank activities. However, the most recent papers have given greater consideration to the role of the World Bank and coordination with its activities—especially in the financial sector.

94. With a few notable exceptions, most of the staff reports on Article IV consultations do not give much sense of the long history of IMF involvement with the Philippines, and any lessons that this experience may have had for the assessment of the current situation. Many of the staff interviewed by the evaluation team believe that explicitly standing back and reviewing the history of the relationship, as was done for example in preparing internal country strategy papers (CSPs) in 1993 and in 2000, had been very useful. However, even these exercises had their limitations. As noted earlier, the preparation of a draft country strategy paper in 1993, which was never completed, prompted a lively internal debate about the lessons from previous programs for the design of the new program. However, little of the candid flavor of this internal stocktaking showed up in the subsequent 1994 Article IV consultation report. Even the 2000 Article IV report, which had considerable strengths (see below), does not reflect much of the candid flavor of the 2000 CSP.

95. Discussion of policy alternatives and tradeoffs was typically limited, although such issues have more recently been covered in greater detail particularly when there were clear divergences between the staff’s view and that of the authorities.

96. There is some sign that recent Article IV reports, even when combined with UFR issues, have become somewhat more effective at “standing back” from immediate program issues to assess the totality of the relationship and the vulnerabilities inherent in the existing situation. Some reports have also become a little franker in airing areas of disagreement between the staff and authorities. To give a few examples:

  • The 1991 Article IV Report/Request for Stand-By included a critical review, unusually candid for the time, of performance under the preceding EFF. It drew upon an even franker internal assessment. However, not all of the lessons—including the dangers of overloading the structural reform agenda—were fully taken into account in subsequent program design.

  • The Staff Report for the 1996 Article IV consultation identified a number of key vulnerabilities facing the economy and was used to put a number of issues, which were not central to the original program, onto the policy agenda. In particular, the report noted, inter alia, the risks of the de facto exchange rate peg that the authorities had maintained since late 1995 in encouraging large short-term capital inflows; the dangers of the rapid rise in private credit in 1995–96 and its concentration in real estate and consumer lend-ing; and the risks in the way in which the large credit expansion was being financed, through a buildup in banks’ net foreign currency liabilities and foreign currency deposits, offset imperfectly by domestic foreign currency lending. The report highlighted the potential market risks of this financing structure. Circumstances at the time, particularly the program’s then precautionary nature, likely made it easier for surveillance to provide added perspective.

Conclusions

97. Before discussing potential lessons for the IMF from its prolonged involvement in the Philippines, the first question to ask is: Was this involvement a failure? In the sense that having lending arrangements in place for 25 years in a 30-year period is clearly not what the IMF is trying to achieve, the answer must be, at one level, yes. But this is not to deny that a number of IMF-supported programs achieved important successes: between 1984 and the second half of the 1990s a substantial transformation of the Philippine economy took place, and the programs assisted that transformation. So a more nuanced response to the question would contain the following elements: (i) the IMF-supported programs with the Marcos administration prior to the debt crisis clearly failed to achieve their objectives and a much more selective IMF involvement, conditional upon better policies, would have been desirable; (ii) the long IMF involvement from the 1983 debt crisis until the restoration of market access in the early 1990s could probably have been shortened with a program approach that insisted upon stronger implementation of a smaller set of core measures. However, considerable progress was eventually achieved—some of it masked by a tendency of programs to overpromise. Global systemic factors—especially the evolving approach to debt workouts—also extended the IMF’s program involvement in this period; in this respect, the Philippines’ experience was not un-usual; and (iii) even after allowing for the unlucky timing of the Asian crisis, which prolonged the Philippines’ use of IMF resources even further, the IMF maintained its program involvement for too long in the 1990s. In particular, the IMF should probably have disengaged earlier during the Estrada period, by not completing reviews, once it became clear that there was insufficient support from political leadership for core elements of the program, including stopping corruption.

98. With this background, the main messages arising from the Philippines’ experience with prolonged use are as follows:

  • (i) The rationale for continued lending arrangements needs to be spelled out clearly in each case and pressures to expand this rationale resisted.

99. In the Philippines, the rationale for continued arrangements—well after market access had been restored, and any actual or potential balance of payments need was questionable—was too broad. The underlying motivation to support reforms and reformers or to provide a framework to guide an incoming administration was understandable. But if applied IMF-wide, such an approach would lead to many more prolonged users. While it is probably not possible or desirable to specify ex ante precise quantitative exit criteria, a more rigorous discussion and justification of the reasons for continued lending arrangements is needed than occurred in this case.

  • (ii) The choice of the appropriate boundaries between programs and surveillance will be one of the critical determinants of the extent of prolonged use.

100. The decisions to continue with programs in the Philippines in 1994 and again in 1998–2000 in effect reflected judgments that surveillance alone would not be adequate to the tasks of encouraging macroeconomic discipline, promoting structural re-form, and providing some “seal of approval” for investors and lenders. In preparing “exit strategies” for prolonged users, the IMF should consider making greater use of alternative mechanisms, which do not involve IMF financing, to ease the transition from program arrangements. Greater transparency in recent years has increased the potential for surveillance to signal the IMF’s views on the adequacy of the macroeconomic framework to the private sector as well as to official creditors and donors.

  • (iii) IMF-supported programs since the 1983 debt crisis did encourage macroeconomic discipline and structural reform, but problems associated with the time frame of programs and implementation failures during periods of weak ownership prolonged the length of the adjustment.

101. Institutional constraints on implementation, particularly the Philippines’ congressional-style political system, posed special challenges for program design and required structural conditionality to be exercised flexibly. Moreover, programs tended to “overpromise” on both the number and timing of structural reforms—reflecting systemic pressures to show substantial progress within the relatively short time period of the program. Combined with the tacit understanding that the program relationship in the Philippines was likely to continue for a prolonged period, and that continual recontracting of conditions was possible, this overpromising led to a weak-ening of the credibility of conditionality. During those periods when it had become evident that there was no political commitment at the highest level to the reforms and that governance problems were themselves a major threat to macroeconomic stability, the IMF should have been more selective in agreeing to, or extending, programs. However, during other periods when the issue was more one of the authorities’ political ability to push through a wide-ranging reform program, the IMF was faced with more difficult and finely balanced judgments. In retrospect, programs would likely have been more effective if they had focused on a narrower set of critical reforms; for some elements of reform, notably those involving deeper institutional changes such as tax administration, it would have been preferable if a longer time frame had been adopted from the start, combined with an increased emphasis on strengthening implementation capacity and more direct attention to addressing governance concerns. Nevertheless, it should be recognized that the IMF’s role in the latter area has been clarified only relatively recently and there are clearly limits to what outside agencies can be expected to achieve.

  • (iv) The IMF’s role as a gatekeeper for many other sources of financing, through the “seal of approval” element in programs, can be a two-edged sword.

102. It can potentially increase the IMF’s leverage on policies, but in practice, it also appears to have influenced the IMF’s reluctance to disengage, since to do so would have had severe consequences in terms of financing from other creditors and donors. As a result, the IMF was unable to be as selective as would have been desirable in focusing its efforts on periods of crisis management or when political circumstances were ripe to advance reform.

(v) The Executive Board’s guidelines for dealing with prolonged use were only partly followed.

103. Although there was no general common agreed definition of prolonged use, the Philippines has been unambiguously identified as a prolonged user since the early 1980s. In spite of this, the policy governing prolonged use as articulated in Board discussions was not fully implemented. In particular, there was too little systematic candid assessment of program experience in papers for the Executive Board. Opportunities to reconsider the overall strategy and learn from experience were too limited. The staff did step back a few times to review strategy, informed, inter alia, by fairly candid ex post assessments of programs. This stepping back went further than in the other countries subject to case studies, but the most candid elements of the assessments and of the internal review process generally were often not communicated to the Board. Consequently, they were not used by the institution as a whole to critically reconsider the IMF’s overall strategy in the Philippines. A more regular and systematic approach to appraising the overall strategy, on the basis of frank assessments of previous programs, and in a manner that involved the country authorities and the Executive Board would have been desirable.

Appendix I Philippines: List of People Interviewed in Connection with the Evaluation of Prolonged Use of IMF Resources

Government Officials

Mr. Jose Isidro N. Camacho, Secretary of Finance

Mr. Rafael B. Buenaventura, Governor, BSP

Mr. Amando M. Tetangco, Jr., Deputy Governor, BSP

Mr. Nestor A. Espenilla, Jr., Managing Director, BSP

Mr. Gil S. Beltran, Assistant Secretary, Department of Finance

Ms. Emilia Boncodin, Secretary, Department of Budget and Management

Ms. Laura Pascua, Under Secretary, Department of Budget and Management

Mrs. Maribel D. Ortiz, Director, Economic Re-search Department, Social Security System

Former Senior Officials

Mr. Fidel V. Ramos, former President of the Philippines

Mr. Cesar E.A. Virata, former Prime Minister and Secretary of Finance

Mr. Gabriel Singson (former Governor, BSP), Chairman/President, JG Summit Capital Markets Corporation

Dr. Jesus P. Estanislao (former Secretary of Finance), President & CEO, The Institute of Corporate Directors

Mr. Victor Macalincag, former Under Secretary of Finance

Dr. Felipe Medalla, former Director-General of NEDA

Mr. Roberto De Ocampo (former Secretary of Finance), President, Asian Institute of Management

Mr. Jose L. Cuisia, Jr. (former Governor, Central Bank of the Philippines), President & CEO, PHIL-AM

Mr. Romeo L. Bernardo (former Under Secretary of Finance), Managing Director, Lazard Bernardo Tiu & Associates, Inc.

Ms. Wilmida Guevara (former Under Secretary of Finance), Ford Foundation

Academics

Mr. Mario B. Lamberte, President, Philippine Institute for Development Studies

Mr. Ponciano Intal, Professor, de la Salle University

Mr. Ruperto Alonzo, Professor, University of the Philippines School of Economics

Banking Sector

Dr. Placido L. Mapa, Jr., President, Metropolitan Bank & Trust Company

Dr. Vaughn F. Montes, Senior Vice President, CitiBank, N.A.

Mr. Don Hanna, Salomon Smith Barney Mr. Tim Condon, Chief Economist, Asia, ING

Business Community

Mr. Raul Concepcion, Chairman, Federation of Philippines Industries

Nongovernmental Organizations

Mr. Carlito T. Anonuevo, Action for Economic Reforms

Ms. Jessica Reyes Cantos, Action for Economic Reforms

Ms. Maria Teresa D. Pascual, Freedom From Debt Coalition

Mr. Jose Luis Gascon, National Institute For Policy Studies

Mr. Ed Tongson, World Wildlife Foundation

Mr. Carlos H. Aquino Jr., Philippines Peasant Institute

Representatives of Donors

Mr. Khaja H. Moinuddin, Director General, Southeast Asia Department, AsDB

Mr. Ricarda Rieger, Deputy Resident Representative, UNDP

Ms. Jennifer Navarro, UNDP

The mission also met with a large number of current and former IMF and World Bank staff involved with these institutions’ work on the Philippines.

Amounts outstanding were initially small as a percentage of quota, in line with general IMF policies at the time, but then rose quickly to the 250–300 percent range by the time of the 1982 debt crisis. They declined by the mid-1990s, but picked up again in the aftermath of the Asian and Russian crises.

See “Prolonged Use of Fund Resources,” SM/84/91, April 27, 1984.

There were also several precautionary SBAs prior to this starting in 1962. The background paper to the 1984 review of prolonged use of IMF resources contains a detailed review of the IMF’s involvement in the Philippines since the 1960s. See also Boughton (2001).

National government balance.

The EFF initially targeted an increase in the ratio of tax to GNP from 13½ percent in 1975 to 16 percent in 1978; the actual outturn in the latter year was under 14 percent. This largely reflected shortfalls in indirect taxes but does not appear to have been related to adverse exogenous shocks since real growth was broadly on target and the terms of trade were better than expected under the program.

Moreover, at the time there were significant IMF-wide shortcomings in the design of limits on the external debt contracted by the public sector, in particular the exclusion of short-term debt from these performance criteria and insufficient attention paid to the impact of public guarantees on external debt contracted by the private sector. These problems are now well recognized, and external debt ceilings generally now have a more comprehensive coverage.

The internal 1984 review of prolonged use was quite candid in its assessment of the reasons why the long series of programs with the Philippines had failed to achieve their targets. For example, it noted that, while growth had been quite high for much of the period, “these developments reflected, in part, an unsustainable accumulation of foreign debt as the external position and domestic saving performance remained unsatisfactory.... Subsequently, unfavorable external developments, inadequate structural policies, and inadequate demand-management policies, in particular in 1981–83, resulted in a sharp worsening of the balance of payments position.” It also noted the problems with wasteful public expenditure but shied away from an explicit discussion of governance-related factors: “A significant part of the public investment program, which was largely financed abroad, consisted of projects with doubtful economic justification.” See “Prolonged Use of Fund Resources,” SM/84/91, April 27, 1984.

A guidance note on governance calling for a more comprehensive treatment of governance issues and a more proactive approach in advocating policies that promote good governance were only approved by the Executive Board in 1997. See IMF (1997 and 2001b).

Public debt service also fell—from an average of 48 percent of central government revenue in 1986–88 to 26 percent in 1994–97.

Exchange rate policy was the subject of some disagreement between the staff and the authorities in the mid-1990s. The Philippine peso was pegged de facto to the dollar in late 1995, as the authorities tried to prevent a real appreciation, encouraging short-term capital inflows. The risks of this policy, particularly if capital flows were reversed, were highlighted in the Article IV consultation in December 1996. Following the float of the peso in July 1997, the authorities initially engaged in substantial one-way intervention to support the peso and took other measures designed to prevent depreciation, including limiting access to the foreign exchange market by some foreign banks, while rapidly reducing interest rates contrary to commitments in the letter of intent.

The existing EFF could only have been further extended until June 1998.

The program was revised to take into account weaker economic activity and a deteriorating external environment, in particular through a somewhat looser fiscal stance. It also reflected the announced policy agenda of the new government, with greater attention to corporate governance issues, power sector reform, and medium-term public sector reform in addition to existing planned reforms for the banking sector and tax administration.

The program aimed at the full privatization of the PNB by mid-2000. However, the situation was complicated by the non-transparent acquisition of a controlling stake by a local group, headed by a close associate of then President Estrada, which owned several companies that were nonperforming debtors of PNB, and by the revelation of increasing concerns about the group’s financial position.

In September 2000 the Executive Board had agreed that, when a member’s credit outstanding exceeds a threshold of 100 percent of quota, there should be a presumption that the member will engage in post-program monitoring (PPM) by the IMF of economic developments and policies after the expiration of its arrangement. To this end, the member engages in discussions with the staff on its policies, including a quantified macroeconomic framework. The staff then reports formally to the Board, normally twice a year, on the member’s policies, the consistency of the proposed policies with the objective of medium-term viability, and the implications for the member’s capacity to repay the IMF.

The question addressed in the section “Continuous Programs Have Limited the Independent Role of Surveillance” is whether surveillance could also have provided the necessary framework.

Over the years of IMF-supported programs, various efforts were made at reform of the financial sector, including the rehabilitation and restructuring of the Philippine National Bank and the Development Bank of the Philippines in the late 1980s. After the opening of foreign direct investment in 1991, most capital account restrictions were removed in a wide-ranging reform in 1992. Measures to strengthen the supervisory and regulatory regime for commercial banks and to increase competition in the banking sector were taken at about the same time. In particular, the move was accompanied by a change in the prudential formula determining the maximum net open foreign exchange position for each bank. Restrictions on entry and operation of foreign banks were eased in the early 1990s to increase competition, and other measures were taken to tighten prudential regulation, including a tightening of minimum capital requirements and the imposition of a liquid asset requirement on foreign currency deposit unit loans. Nevertheless, private credit grew very rapidly ahead of the Asian crisis, and bank supervision and regulation continued to exhibit serious shortcomings compared with best international practice. Since 1998, the authorities have adopted further financial sector reforms, with the help of the IMF and a Banking Sector Reform Loan from the World Bank (Rodlauer and others (2000), Chapter VI: Banking System Reform).

According to the World Bank’s World Development Indicators, 2001.

The Central Bank incurred large liabilities, due to the impact of sharp exchange rate depreciation on the cost of the exchange rate guarantees and forward cover arrangements, provided in the early 1980s. The authorities took a policy decision not to monetize these liabilities—which was of critical importance in helping the Philippines avoid high inflation in the aftermath of the crisis—but interest payments on central bank bills issued to cover the losses gave rise to continuous quasi-fiscal deficits and a highly negative net worth. It also affected the Central Bank’s ability to conduct monetary policy. The Central Bank was eventually recapitalized and restructured, under a new statute, in 1993. This was a prior action for the EFF, which was agreed the following year.

Projections for the 1991 SBA foresaw financing gaps throughout the decade, noting that this pointed to the continued involvement of the IMF and further Paris Club and commercial bank financing packages through the mid-1990s. The 1989 EFF request incorporated financing gaps of $1.7 billion a year in 1993–95, and suggested that an end to exceptional financing and a return to normal market access might not occur until after 1995.

More recently, full activation of bilateral swap arrangements with other Asian economies have been tied to the existence of an IMF-supported program. In 1994 the program was also associated with a Paris Club rescheduling but, in this case, the Philippines (uniquely in the history of the Paris Club) eventually decided to cancel the agreed rescheduling, apparently because of the implications for cover from export credit agencies and the questionable need for a rescheduling.

For example, in 1992, the tight links between Paris Club rescheduling and IMF-supported programs caused the program to be extended through a change of administration, risking compromising program ownership.

The 1993 and 2000 country strategy papers—and the 2000 Occasional Paper (Rodlauer and others, 2000)—were quite candid in identifying these factors as key obstacles to faster adjustment and growth.

Private saving, which is calculated as a residual, apparently rose from 19.9 percent of GNP in 1998 to 29.2 percent in 2000, while the current account surplus rose to 11.8 percent of GNP from 2.2 percent over the same period. However, a staff analysis of partner country trade data suggests that the current account surplus may be overstated by as much as 9–10 percentage points of GNP, mainly due to the underrecording of imports. Given the method of calculation, this would imply that the saving ratio is correspondingly overstated.

More recently, additional vulnerability concerns have focused on public sector debt dynamics and continuing weaknesses in the financial sector. Exports remain heavily concentrated in the electronics sector.

This does not imply that policies in the Philippines—including exchange rate policy, remaining financial sector regulatory weaknesses, and relatively shallow domestic capital markets—had no role in magnifying the effects of the crisis.

Between 1982 and 1993, there were small net repayments to the IMF, which was in line with the guidelines that called for a gradual reduction in exposure to prolonged users.

As noted earlier, the first one-month extension of the 1994 EFF took place in June 1997 before the floating of the Thai baht, and was related to the desire for a “graceful exit” after completion of the (postponed) tax reforms.

It was not ultimately implemented until 1996 as a result of a judicial challenge.

For example, disbursements under the 1994 EFF were evenly loaded. The guidelines appear to have been more closely followed in some respects in the case of the 1998 SBA. The arrangement was back-loaded with 5 percent of quota available on approval, while the first four installments each for 14 percent of quota and the remainder each for 25 percent of quota. This back-loading was not maintained when the program was rephased in its later stages.

This may in part reflect the commitment of management to a follow-up program communicated to the Philippines authorities as early as June 1992.

Similarly, a number of internal comments prepared by reviewing departments at the time of the various reviews of the EFF did question some of the underlying rationale for the program. In particular, the weak correspondence between the Philippines’ economic performance and its compliance with program conditionality was noted, since strong capital inflows had continued despite repeated delays in completing reviews and the breaching of a number of performance criteria and structural benchmarks, suggesting that any “signaling” rationale was weak.

Tax administration was not featured prominently during the 1994 EFF, which focused more on tax policy issues, but, according to staff, even then efforts continued on the ground to improve tax administration.

Over the same period, the average disbursement ratio for all arrangements using general resources was 64 percent.

For example, of the 22 program reviews scheduled under the seven arrangements since 1983: just 6 were completed on schedule, with a further 3 completed with a delay of three months or less. Nine were delayed for longer, including 2 reviews delayed for ten months each. Four of the scheduled reviews were never completed, as arrangements were cancelled.

Over the 1987–99 period, the Philippines ranks fifty-fourth out of 105 countries that had programs in terms of waivers per program year with just eight waivers. None of these waivers related to structural performance criteria and all but one of these waivers related to quantitative performance criteria. In addition, the Executive Board granted three waivers of applicability over this program period, related to delays in completing reviews.

Programs varied greatly in the degree of detail with which conditions on structural measures have been specified. For example, some programs incorporated highly detailed agendas for action against which progress on structural reforms could be monitored, while others were simply specified in terms of reviewing general progress in a particular area of structural reform.

In the mid-1990s, the Philippine government formed the Legislative-Executive Development Advisory Council (LEDAC) as a vehicle to push through Congress certain bills identified as priori-ties by the Executive.

The following quote illustrates the problem: “Contrary to common belief, the IMF is not really that inflexible.... The Philippines has had 25 programs covering 35 years. The first lesson I learned is you do not have to do everything the IMF wants you to do. Secondly, even if you agree with the conditions and you do not comply, you can always ask for a waiver” (Senior Philippine official, 1998).

This approach was followed in 1986/87, 1994/95, and 1999.

In the 1991 program, an unusual type of “implicit” structural benchmark was used. In order to meet pressing short-term fiscal needs a temporary (9 percent) import levy was introduced and this was to be phased out as Congress passed other tax measures submitted by the government. Completion of the reviews was made dependent on progress in phasing out the import levy and hence, implicitly, on progress in Congress in passing the alternative tax measures. However, the levy was in fact eliminated ahead of schedule in 1991, despite considerable slippage in the authorities’ revenue efforts as some of the tax administration measures were delayed.

The tax administration weaknesses were due to major problems with compliance and widespread corruption; weak control over granting refunds and tax credit certificates, which were major loopholes in VAT collection; a failure to implement fully the large taxpayer scheme launched in the early 1990s; and lack of an effective tax information system.

Both missions included FAD tax experts and overlapped with FAD technical assistance missions on tax administration, although a lack of continuity contributed to the differences in judgments. One of the major area of differences between the authorities and the staff (which was complicated by the change in administration) proved to be on the scope and functions of the large taxpayer unit. In the event, many of the institutional changes took much longer to implement than originally envisaged.

These elements and their derivation are discussed in Chapter 6 of Part I. The quality of surveillance was assessed by systematically rating the performance of each surveillance report for nine functions viewed by the IEO as “key elements” of surveillance in a program context. These nine functions draw on the “minimum requirements” of Article IV reports as listed in internal guidance notes. They are (i) provision of realistic medium-term and alternative scenarios; (ii) provision of meaningful sensitivity analyses; (iii) discussion of risks to the assumptions and projections; (iv) discussion of the risks and impact of policy slippages and of vulnerabilities; (v) balanced reporting of the authorities’ views, including any significant differences with staff; (vi) cogent presentation of proposed policy course; (vii) discussion of policy alternatives and trade-offs; (viii) critical and frank review of previous UFR performance; and (ix) presentation of collaboration/interaction with the World Bank.

    Other Resources Citing This Publication