Seminar on IMF Conditionality
July 10, 2001
Conditionality in Fund-Supported Programs -- Overview
IMF Seminars, Conferences and Workshops
I think this seminar has been useful in generating diverse inputs into the conditionality debate. From my side, I would like to make a few comments on the scope of conditionality and its relationship to ownership. The currently prevailing view is that extensive conditionality weakens ownership and diminishes the will as well as the ability of governments to implement programs. Policy undertakings should, therefore, be limited to a few key areas, which should be in the Fund's mandate, basically macroeconomic and financial.
I differ a bit from this approach. Based on the experience with programs for three Asian crisis countries (Thailand, Korea, Indonesia) as well as earlier crisis programs with India (1991-92) and the Philippines (1983-85), I would say that in all cases there was an obvious need for action across a broad range of policy areas, and that it was possible to achieve ownership of comprehensive programs.
In elaborating on these points, and illustrating them with examples of the five crisis programs mentioned, I will comment on three issues:
I will conclude by trying to draw a few lessons from experience.
The Need for Comprehensive Programs
Whether or not we like extensive conditionality in principle, in the five cases I mentioned, comprehensive programs were necessary for three reasons:
First, the crises were not one-dimensional, and overcoming them required action on all problems that had emerged. As Deputy Governor Li Rougu emphasized this morning, the overwhelming objective of a crisis program is to ensure a quick recovery (and I would add, at minimum cost). How comprehensive a program has to be to achieve this objective is a question of economic substance, independent of Fund competence or Fund mandate. I myself find it difficult to imagine that countries could have achieved a recovery and laid the basis for a lasting resumption of growth through a selective rather than a comprehensive policy response.
Second, apart from substance, there is also an issue of perception. Markets were watching governments carefully, and failure to address all major issues that contributed to the crisis would not have restored confidence - an essential ingredient for recovery. Whether they were right or not, markets would not have bought the idea to stabilize first and restructure later. They thought - and I think with some justification - that if needed structural measures were not taken now, they would never be taken.
Third, and this is often overlooked in the ex-post debate, at that time, only a comprehensive program had a chance to be accepted by the international official community - the only source of the massive inflows of funds, multilateral and bilateral, which were required to abate the BOP crisis. It is important to remind ourselves that conditionality is ultimately determined by the member governments of the IMF.
Let me elaborate a bit on the last point. In the end, programs had to be approved by the Fund's Executive Directors, who represent member governments and whose votes are weighted by the economic importance of the countries they represent. Programs, therefore, had to meet their expectations. Governments' views were obtained through regular informal contacts with the Executive Board during the program negotiations (in the case of Indonesia, including two video conferences), as well as in discussions with the Paris Club members. It was amply clear that the international community required comprehensive action.
To give some examples, I do not think the Board would have accepted programs which failed to dismantle the monopolies in Indonesia which were a main source of corruption; or in the case of Korea, failed to liberalize foreign ownership and take-over rules in order to attract capital into the financial sector and introduce effective competition, or to make changes in the labor laws to allow corporate restructuring. Similarly, programs failing to revise bankruptcy and foreclosure laws in Thailand and Indonesia, in order to allow effective debt restructuring, would have been considered inadequate. And in all cases, programs would not have been passed without including extensive privatization, an area that is not mentioned in the Fund's Articles.
I would even go further and say that the Indonesian program without conditions to strengthen the Jakarta commercial courts or to clean up the political corruption case involving Bank Bali - both of which had no direct bearing on the crisis - would not have been approved. The latter condition was highly popular in the country, but here the Fund had entered an area where I think more careful scrutiny and streamlining should take place in the future, even if this means resisting pressure by some member governments.
In any case, governments were aware that the country had suffered more than just a BOP crisis, as long-standing structural weaknesses and defects in the system of economic management had clearly come to the surface. They dealt with these problems simultaneously from the start of the crisis. And as a matter of fact, all five crisis programs were unanimously voted for by the Board, with no abstentions.
It is an open question where the international community will come out, after this debate, on the need for comprehensive programs, even if the focus of the Fund itself would be narrower. It would certainly be desirable to move towards deleting those conditions, which are peripheral for the overall program objective, but retaining all those, which are essential, whether they are regarded to be in the Fund's competence or not. In the end, as Jack Boorman said earlier, there must be a reasonable compromise.
Problems to be Addressed in Comprehensive Programs
Since there was a need for comprehensiveness in the crisis programs I mentioned, the IMF could not be parochial and restrict conditionality to those areas, which were perceived as its core competencies. This raised, however, two practical issues: how to deal with areas where the Fund staff had no or insufficient expertise and where it should not be directly involved; and how to ensure that the government takes ownership of a program involving a broad set of conditions.
We tried to resolve the first issue by involving the World Bank and the ADB as well as outside experts - with different degrees of satisfaction. The biggest effort in this respect was made in the case of the Indonesian programs (I refer particularly to the programs starting in May 1998). World Bank and ADB staff participated directly in all program negotiations, and the respective institutional responsibilities were explicitly indicated in the LOI. For certain areas, we involved outside experts, for instance to establish an AMC (IBRA), to set up the framework for corporate debt restructuring (Jakarta Initiative), to perform audits of public institutions, or to draft a new Central Bank law.
While these arrangements allowed us to negotiate comprehensive programs, it does not mean that they worked perfectly. Problems arose in particular, when other institutions were initially reluctant to be involved in a specific area and the Fund had to step in, since no time could be lost in a crisis, or when differences of view could not be resolved in an efficient way. Improving coordination will be a major challenge in future crises.
Nor does it mean that the resulting programs were perfect and did not involve some misguided and unnecessary meddling. Some conditions did not make much sense in retrospect and reflected inadequacies of program design, while some others reflected the narrow self-interest of governments of other Fund member countries. No doubt, there is need for improvement here. But justified criticism of flawed details should not automatically be extended to broad conditionality as such.
Before I go on, let me remind you that the broad approach to conditionality was in fact not new, and had not been objected to in the past. During the Indian crisis in 1991, the program included a major reform effort to dismantle the regulatory and licencing system, which had stifled commercial initiatives. Here, the crisis was used as an opportunity to push through reforms, which were considered impossible in normal times. Finance Minister Mamohan Singh backed by Prime Minister P.V. Narasimha Rao, took firm ownership of this truly revolutionary program, and the international community welcomed it in its entirety. No one, for instance, objected to removing small scale business reservations which had led to a major misallocation of resources, but had nothing to do with the crisis. India's high growth rate since then testifies to the success of these reforms. (We were much less successful in one core area - to reduce the public sector deficit. It did initially come down, but started rising again in the third year)
In the Philippine programs of 1983-84, the dismantling of the sugar and coconut monopolies, through which two Marcos cronies exploited the major part of the rural population, were major elements among several other reforms. The Aquino Government took full ownership of this and the international community applauded.
The second issue was to bring the government on board, and to keep it on board, with comprehensive conditionality. Of course, the more conditions there are, the more vested interests are threatened and, especially in pre-election periods, the more the government itself is put into a difficult position. Our efforts included being sensitive to government's concerns and exercising flexibility, but only to the extent that we were sure to obtain Executive Board agreement. This is a difficult area, and the results were mixed.
In the earlier Philippine crisis, it took almost a year to agree on a program with Marcos who was very reluctant to accept the proposed reforms. Suharto did sign off on the first two agreements, but then sabotaged their implementation. On the third, he and his team were on board, but then it was too late for implementation. President Habibi and Coordinating Minister Ginandjar with his team, on the other hand, took strong ownership right from the start, although the program was as comprehensive as its predecessors, and there was no official signing ceremony. Substantial progress was made despite the difficult political transition. Unfortunately, ownership under the Wahid administration evaporated quickly as factional infighting dominated the scene more and more. In Thailand's initial phase, the coalition government was not fully behind the program. The markets must have sensed this, but in the Fund, we were puzzled why confidence remained depressed despite such a wonderful program. However, the succeeding Leekpai government with Finance Minister Tarrin took full ownership, and things improved until a year before the election, when political maneuvering took center stage. In Korea, the outgoing and the incoming governments, after a very brief initial hesitation, took strong ownership of the program, and there was a smooth transition to the new government.
Ownership was achieved in each case in the end because governments and the international community were able to strike a reasonable compromise. Not that this went smoothly or that the outcome was ideal. But here, both sides had to accept some realism. Governments had to realize that in order to gain international support, they must meet the standards set by the international community which would not commit official funds otherwise. Ownership had, therefore, also to embrace some measures which governments would not have chosen to implement on their own accord. It would be naïve, therefore, to pretend that a program owned by a government would be purely "home grown" and a complete reflection of that government's preferences. The international community , on the other hand, had to recognize that governments cannot undertake certain measures, which would undermine their own political standing. It would be hypocritical to ram such measures down the throat of governments and then demand that they take ownership. They won't, anyway.
There is another issue related to multiple conditions which, however, in the countries I had worked, has been much overstated. Some people argue that wide-ranging conditionality overtaxes the capacity of the government to implement the program, and they point, in particular, to the case of Indonesia. I have seen little evidence of this. Some conditions, like dismantling a monopoly or closing an extra-budgetary fund, just required issuing Presidential decrees and their publication in the local newspapers to be effective, and the bureaucracy could handle this easily. Other conditions which involved extensive preparatory work, like bank closures or revising the bankruptcy and labor laws, proved also to be within the capacity of the bureaucracy. (Ironically, the area where the administration had real implementation difficulties was the - belated - switch from budgetary restraint to fiscal expansion. Expenditures could not be stepped up fast enough, and deficits remained substantially below the program ceilings)
Broadening of Ownership
While the importance of government ownership is obvious, I think the need for ownership goes further. Not only the government must be behind the program, but major parts in society must also back it in order to ensure its success. There should be something close to "national ownership" mentioned earlier by Khun Tarrin. To help build such a consensus during the Asian crisis, the IMF reached out to groups and personalities outside the government, and this involved frequent staff contacts as well as visits to the countries by senior Fund management. Indonesia and Korea are probably the best example. We talked to opposition parties, Parliamentarians, union leaders, religious leaders, bankers, students, faculties, and others. We also addressed the broader population through TV interviews and newspaper articles. We explained the rationale of the program and the constraints on both sides, and we responded to criticism. This did not produce agreement on every detail of the program - that was not the point - but it did raise the level of debate, separated genuine from imagined conflicts, and established a broad consensus in favor of the program. Publication of LOIs was a big help in this respect.
As Deputy Minister Kim said, there was a national consensus in Korea, which allowed effective program implementation. In Indonesia, despite the political turmoil and conflicts during Habibi's Presidency, the IMF program did not become a focal point of conflict or an election issue. Success was less in Thailand where the opposition became a bit unprincipled and turned hostile to the same program which they had negotiated while they were in government. Earlier in India, the emergence of a broad consensus for reform was mainly due to strong efforts of the government itself to change a deeply entrenched mindset.
To get the relationship between conditionality and ownership about right, my own suggestions would be to follow a few guidelines for negotiations of crisis programs: