Of Mountains and Molehills

1

September 15, 2008 FINAL

Of Mountains and Molehills:

“The” Medium-Term Expenditure Framework (*)

by:

Salvatore Schiavo-Campo

Paper presented at the Conference on

Sustainability and Efficiencyin Managing Public Expenditures

Organized by the East-West Center and Korea Development Institute
Honolulu, Hawaii

24-25 July 2008

(*) The author, currently an international consultant, is a retired senior staff member of the World Bank and the International Monetary Fund and former Professor of Economics at the University of Massachusetts. Comments on an earlier draft by Hyungpyo Moon, A.Premchand, and Daniel Tommasi are gratefully acknowledged.

Introduction

While some of the observations in this paper may have some relevance to OECD countries, the focus is on the application and applicability to developing and middle-income countries of the different variants of a multi-year perspective—commonly known since the early-1990s as “the” Medium Term Expenditure Framework (MTEF).

Limits on space and experience make this a selective and personal account—although this experience includes long-standing concern with and research on the main issues as well as direct involvement in public expenditure management reform in some 40 countries, several of which are used to illustrate the points made here. (Country names are withheld in cases when my involvement carried an obligation of confidentiality.) Also, because the paper is intended mainly as a contribution to rebalancing the MTEF approach, it is discursive and has a strong operational bent.

The introduction of the MTEF concept and its early application are now some 15 years old, and the time for a candid and fact-based assessment is long overdue. Given the hype the MTEF has enjoyed, its rapid expansion in the last decade, and the disregard of some fundamental considerations of institutions and capacity, a little extra emphasis in the interest of understanding the actual issues it timely.

It is well to remember that the problems the MTEF is intended to address are among the oldest in the development literature. How to reconcile short-term urgencies with longer-term priorities; prevent fiscal stability from degrading into economic stagnation; find paths to sustainable growth with financial stability; complement growth with equity—these issues were at the center of the debate on development planning during the 1950s and 1960s, the years of what Paul Krugman has called the “High Development Economics”. A few references to those intellectual roots help place the current arguments in better context.

After a brief account of the genesis of the MTEF and its widespread adoption by the donor community—led by the World Bank—the paper summarizes the current state of play and the meager results of the MTEF so far in most countries; proceeds to unbundling the concept into its several variants; gives some guidance on how to implement the specific variant that is suitable to different situations; and concludes with a scorecard of positive and negative aspects of “the” MTEF.

1. How did we get here?

The coronation of the MTEF: A stroll down memory lane

A quick recapitulation may be useful to place the discussion in context. As a broad generalization, by the end of the 1970s, developed country governments had become too expensive, too big, and too intrusive. The mid-1970s to the late 1980s consequently witnessed a first wave of reforms focusing on expenditure reductions, public sector downsizing, and privatization. These reforms for “smaller/cheaper government” took place in response to fiscal necessity and thus under strong direction and control from the center of government, particularly the ministry of finance. From the late 1980s to the 1990s, after fiscal deficits had been reduced in most OECD countries, a second phase of reform addressed the goal of “better government”, by improving service delivery, deregulation, devolution of responsibilities to subnational levels of government, greater transparency and better access to information.

During about the same period, the World Bank’s approach evolved from an exclusive focus on projects to consideration of the entire investment portfolio, then expanded to the profile of overall public expenditure, and to an assessment of the soundness of the overall policy framework. In particular, the need for a “public investment program” (PIP), essentially a Bank innovation, was justified by the reality that money is fungible and aid to finance a large project needed the assurance that the investment portfolio as a whole was acceptable.[1] This was even more necessary in order to justify untied budget support. Unfortunately, in most developing countries PIPs quickly degraded into shopping lists to attract aid, theatrical scripts written by external consultants to be performed at international donor meetings, and occasionally even fig leaves for giving budget support to thoroughly corrupt regimes. [2]

By the early 1990s, with the credibility of these “first generation” PIPs in ruins, the World Bank was ripe for a “new paradigm” allowing it to get away from the pesky realities of improving the nuts and bolts of investment programming and public expenditure management in general--as in the Turkish folk tale of a man looking for his purse under the street lamp because it was too dark in the alley where he had lost it. Enter The Medium Term Expenditure Framework to save the day—assertively propagated throughout the institution by a few enthusiasts convinced that medium-term expenditure programming methods introduced in highly advanced countries could be transplanted throughout the developing world like so much breadfruit from Tahiti to the Caribbean.

In the overused cliché’, it was a perfect storm: the previous “model” had failed; the World Bank was perfectly placed to push the new model; many of the Bank’s operations staff fell in love with the irresistible cutting-edge, state-of-the-art, best-practice “new paradigm”; budget support was acquiring momentum as preferred form of aid; and most aid-recipient governments were in no position to push back—partly because their own public financial systems were in shambles to begin with. The Bank engine went into overdrive and, in the post Cold War world, other donors quickly followed suit.

The vast economic, social and institutional differences between developed and developing countries were glossed over, “the” MTEF spread like hot butter and by the end of the century it had become well-nigh impossible to find any aid-supported program of public management reform that didn’t call for development of an Em Tef, or consolidation, or strengthening, or widening, or deepening of it. The original valid purpose of placing the annual budget decisions into a multi-year perspective in a manner suitable to country conditions was obliterated by a blizzard of transaction costs, which then, predictably, were used to justify further interventions to improve the “reforms” that had produced them. And, since comparatively few donor or governments’ staff had actually done the homework needed to find out just what the tool implied and required, it was pushed everywhere in cookie-cutter fashion. What had started as a necessary programming tool to cut government spending in advanced countries acquired shamanistic status as cure-all for weak budget preparation, discretionary budget execution, inoperative financial controls, systematic theft of public money and progressive deterioration of social services in developing countries.[3]

Father of MTEF: Australia’s forward estimates

As discussed later, the need for placing the annual budget process in a medium- and long-term perspective goes back at least to the 1950s literature on development planning. In more recent times, medium-term expenditure forecasting was pioneered by the United Kingdom, as explained in Premchand, 1983. The current paradigm, however, can be traced mainly to Australia, a leader among developed countries in reforms to control expenditure growth. Australia’s “forward estimates” approach is too well known to describe in detail here.[4] Briefly, the approach aimed at strengthening the link between government policy and expenditure programs and improving the affordability of policies, by combining projections methods with institutional arrangements to enforce the outcomes. The initial goal of constant revenue/GDP and expenditure/GDP ratios in the late 1970s was tightened into a goal of constant expenditure in real terms. Between 1985 and 1990 this resulted in turning a fiscal deficit of 4% of GDP into a 2% surplus, while at the same time reorienting the composition of expenditure and providing incentives for greater efficiency. The approach replaced the earlier practice of reviewing only the budget requests for the coming fiscal year (paying no attention to the expenditure estimates for the subsequent years) with a practice of rolling forward estimates, by which the ministries agree with the Department of Finance on baseline projections of expenditure on ongoing programs for three future years and the Department of Finance thereafter updates these projections according to changes in economic parameters or to government decisions affecting program costs. Given the original agreement, the process provides reduces uncertainty over future funding levels. The same process is followed regarding new programs, which must project their full costs over the three-year period in order to be considered for funding. This eliminated the time-consuming bargaining over the base for each fiscal year and allowed focusing only on the budgetary implications of policy changes or strategic decisions.[5]

High-level political guidance and buy-in was ensured by an Expenditure Review Committee (ERC--consisting of the Prime Minister, the Treasurer and Minister of Finance and a number of major line ministers), responsible for approving the overall fiscal framework and managing strategic policy changes, as well as setting the resource ceiling for each sector ministry for the preparation of the annual budget. If the aggregate ceiling for the ministry is lower than the coming year’s cost of existing programs, the ministry concerned would need to find savings or take other efficiency measures; if higher, the ministry could use the fiscal space to introduce new initiatives. In any event, under the system it is the responsibility and authority of the competent ministry to determine the allocation of resources among different programs in the sector, consistent with overall government policy and within a hard expenditure constraint. Analogously, within each ministry, line managers would have flexibility in regard to both staff and money, again within the budget constraint applicable to their program.

The Australian MTEF has worked well, as have similar mechanisms in several other OECD countries. In addition to the many technical pre-requisites, the reasons: include the quality of governance; the high propensity to rule-compliance; the political discipline of a well-organized executive apparatus; the prodding and vetting by the legislature; the contestability arising from a vibrant civil society possessing both exit opportunities and voice channels; the availability of a large pool of highly-competent government economists, accountants, econometricians, sector specialists, others; administrative and social accountability mechanisms; the high level of public integrity; a high degree of flexibility given to line ministries and budget managers within them regarding both personnel management and financial resource allocation; a large and diversified economy producing a foundation of predictable government revenue not dependent on external sources; and a population relatively homogeneous in cultural, religious and ethnic terms. [6]

To preface the discussion that follows it is sufficient to ask how many of the above requirements for a well-functioning MTEF are present in the average developing country, and thus how likely it is that the same approach can be successfully exported to such country. That question, unfortunately, has too rarely been at the forefront of the policy dialogue and international activities to strengthen public financial management in developing countries.

2.Where are we now?

The multi-year expenditure estimates practices developed in advanced countries were relabeled as “MTEF” and simply exported, with some changes in terminology and justification but essentially under the same assumptions concerning accountability, transparency, public integrity and, above all, capacity. The chasm between assumptions and realities did not take too long to become evident.

A first glimmer of recognition of these problems appeared at the World Bank’s “PREM Week” meeting of late 2000,[7] when it was noted, among other things, that if a country cannot put together a sensible annual budget and execute it in minimally acceptable fashion it is very unlikely to have any use for a medium-term expenditure framework. One panelist--Alister Moon--gave a long list of preconditions for introducing an MTEF, including macroeconomic stability; revenue predictability; early political commitment; core capacity of the finance ministry and central agencies; supportive donor behavior; capacity to enforce a hard budget constraint at the ministry level; executive commitment to having a transparent budget process; and capacity in sector policy analysis. While an MTEF process can probably begin even without one or another of these conditions, as a whole they are indeed required. [8]

But the MTEF fashion juggernaut kept rolling along, and the international financial institutions were satisfied enough with the “Potemkin Approach” to multi-year programming.[9] Thus, for example, despite the serious problems of the MTEF in Ghana, already evident by the late 1990s, the country was still cited as an example of good practice. And, three years after the meeting mentioned above, a review of MTEF experience in eight countries (Holmes and Evans, 2003), still projected a fairly rosy picture, concluding that MTEFs were progressing in all the review countries. [10] Although an increasing number of economists and aid practitioners kept pointing to the emperor’s skimpy attire, it was to take another four years before the disconnect between MTEF rhetoric and the practice would become unavoidable, based on a survey of actual results.

In a recent fact-based analysis, Brumby (2008) first recalls the seductive potential of the MTEF construct to “help create [a] textbook world of public finance”, and then proceeds to list the problems that have emerged in practice, and mainly:

• the intense activities of the MTEF units set up in most countries to develop the new processes have not led to improvements in annual budget preparation;
• budget behavior has not actually changed;
• the political leadership has little understanding of the MTEF;
• MTEFs can become a means to present an unrealistic budget.

And, after a reminder of the well-known (but routinely disregarded) prerequisites for an effective MTEF, Brumby gives the following scorecard of MTEF results in Africa: [11]

• virtually no evidence of improved macroeconomic balance;
• some limited evidence of reallocation to priority subsectors;
• no evidence of a link to greater budgetary predictability;
• no evidence of efficiency gains in spending.

This is what the donors and the developing and middle-income countries have got in return for the billions of aid dollars, mountains of red tape, heavy burdens on local government staff, and literally centuries of full-time-equivalent technical experts expended on “the” MTEF. [12]

The realization of the magnitude of wasted resources and dashed expectations is sobering. One hopes that at least the main lesson will not be forgotten: A public management innovation cannot be transplanted as is to a different institutional soil (McFerson, 2007), nor implemented successfully except gradually and over a long period of time.[13] In public management reform, fashion is a dangerous thing. The increasing formalism and futility of MTEF introduction in so many countries is now generating a predictable backlash—partly because that “reform” had been forced on the locals without sufficient comprehension, let alone ownership. The perception pendulum is shifting, and there is mounting skepticism of all MTEF concepts—seen as exhausting and expensive initiatives pushed by donors, and carried out as supply-driven self-propelled exercises conducted mainly by external consultants.[14]

It would be a great pity, however, if the proverbial baby were to be thrown out with the bathwater. The international community should not repeat the mistake made some 15 years ago when it latched on to the “new paradigm” of an MTEF instead of fixing the problems evident in the practice of PIPs. The lesson from the discouraging MTEF experience so far is certainly not to forget the need for a medium-term perspective for the annual budget, but to re-size, redefine and reformulate the MTEF approach in a manner suitable to the possibilities and constraints of the different countries. If the appropriate variant of a multi-year expenditure perspective is introduced with an eye to conditions on the ground, in close cooperation with the host government, with the right sequencing and in deliberate and realistic manner, it can substantially improve the control, allocative efficiency and use efficiency of a government’s financial resources. This leads to the next section.