Difficulties in the IMF’s Financial Programming

Introduction

One of the most widely used and criticised macroeconomic models is the International Monetary Fund’s (IMF) so called ‘financial programming’ model. The stated objective of the model is to achieve desired macroeconomic targets in countries suffering a crisis or are about to receive debt relief. To pursue this objective through the financial programming instrument, in its macroeconomic programme the IMF utilises the Monetary Approach to the Balance of Payments (MABP) and fiscal identities. Through these identities the model is trying to have an effect on the level of foreign reserves and inflation and it is trying to calculate the amount of debt relief and import rate needed for growth.

The initial attempts were put together into one analytical model by Polak in 1957, setting the foundation of financial programming. Other succeeding approaches were all based on the modification and improvement of the Polak model. Basically the model can seen as an attempt to integrate the before existing framework of monetary and credit factors. The importance of this model is undeniable, as all models applied in the financial programming have their roots in the Polak model.

Nonetheless, the Polak model has its limitations as well. One of these is the assumption that changes in domestic credit have no effect on income and domestic interest rates, thus not having effect on money demand. Another drawback of the model is that it assumes a stable money demand function, assumption proved to be false in many cases. All these put together and the fact that the model is specified only in nominal terms may lead in many cases to unreliable macroeconomic projections.

Realising some of the model’s limitations, many other IMF staff members tried to improve and tweak the model including: Robichek (1971), Crockett (1981) and Gutian (1981).

This paper, through a brief description of the financial programming approach, will try to highlight some of these limitations and will discuss several criticisms regarding its effectiveness in targeted economies.

Financial Programming in action

The basic structure of financial programming is designed around the demand and supply of goods, money and foreign exchange. The money and foreign exchange markets are assumed to clear always, while long term aspects are introduced in the goods market where short term demand may be different from the long term supply.

The foundation of the IMF’s financial programming approach for designing economic adjustment programmes is manly consisting of country specific experiences. The advantages of such a model are quite evident. It is simple to use and it is very transparent. However, its biggest strength points to its biggest weakness: data reliability making it vulnerable to even low levels of data discrepancy.

According to Tarp (1993), the process of financial programming, through which an adjustment programme is formulated, can be structured in ten important steps followed through in most of the scenarios. Step one is choosing a desired target variable level. Generally this variable is R – foreign reserves, but inflation or change in private sector credit can also be a target. In the second step projections are generated for the exogenous variables (y – real output, X – exports, ΔF – change in capital flow). Based on the first two steps, in step three the value of import (Z) variable is assessed. In step four the necessity for devaluation is assessed, followed by the fifth step where the demand for money is established. This step includes an analysis of the need to influence the interest rate which would affect money demand. In the next step, sixth, the overall change in domestic credit expansion is set (ΔDC). In the seventh step the value resulted from the last step is confronted with the demand for domestic credit. This step is one of the most painstaking of all, as in most of the cases domestic credit expansion for the private sector is specifically targeted and any gap at this stage signals that the borrowing required by the government is exceeding the permitted level. Step eight comprises of intense negotiations regarding the way in which the gap from step seven should be closed. When this is done, in step nine the consistency of the measures are tested and in step ten, in the form of a letter of intent the performance monitoring criteria are set and the letter is sent to the IMF.

Built-in limitations and external criticism

The stated objectives of the IMF, that of ‘working to foster global monetary cooperation, secure financial stability, facilitate international trade, promote high employment and sustainable economic growth, and reduce poverty’ (IMF 2010), are somewhat idyllic considering the piling critiques towards the effectiveness of the approach in several cases. Although, IMF adjustment programmes are designed with the above objectives in mind, results and post evaluations tend to show a different picture as in many cases the implementation of the proposed policy is not associated with the expected levels of income and employment.

Criticism targeting the financial programming model can be divided into two major categories: internal and external. Internal criticisms are those questioning the model in its internal structure and logic of steps and assumptions. External criticisms are formulated targeting exact measures demanded by the programme in specific countries.

One important issue properly highlighted by Tarp (1993: 75), is that despite the claim that the IMF’s approach has gradually evolved through time, the theoretical framework has remained in principal the same since the 1950s. Accordingly, if prices are taken as given, exchange rate is fixed and imports depend on income, the ‘financial programming model is actually identical to the Polak model’. Continuing, Tarp raises further questions concerning the model. He argues that the model is not taking into consideration uncertainty and expectations and by being static it fails to adopt some of the new developments of macroeconomic theory like the role of risk and self-insurance in portfolio choices, the role of time consistency and pre-commitments in economic policy, the inter-temporal nature of the current account, the economics of equilibrium real exchange rates and several other more subjective aspects like the economics of contract and reputation and the theory of speculative attacks and devaluation crisis. Although it has to be said that due to the abstract nature of these new developments in macroeconomic theory, their inclusion in the model would clearly undermine Polak’s intention of keeping the model as simple and as transparent as possible.

Nonetheless, transparency has its price. As the model mainly relies on data supplied by the target country, Easterly (2006) effectively demonstrates, that the use of poor quality data in many scenarios may lead to different projections and can manipulate the underlying causalities.

It has to be said though, that all macroeconomic models contain identities which are used for the elaboration of projections. However, because of the way these identities are used and restrictions assumed, we might be faced with the fact that at the end, the data simply rejects the restriction.

Another data related problem with the identities is related to the interpretation of certain concepts. The IMF is not entirely consistent with the use of certain concepts as we can find different estimates for the same concepts. It happens quite often that in the IMF’s statistical publication (International Financial Statistics – IFS) an estimate is different from that in the IMF’s country report. Easterly (2006) demonstrates this with a set of randomly assembled sample data collected from recent country reports and compares it the IFS data for the same period, finding significant differences which can lead to serious misrepresentations and measures. Although it has to be said that in most of the identities, the IMF usually includes an item called ‘net errors and omissions’ meant to somewhat counterbalance possible inaccuracies. However, financial programming in practice does not usually attempt to resolve all the different identities from different data sources, making the uncertainty even greater about whether the identities really balance.

Buira, already back in 1982 highlighted some of the model’s weaknesses in three major areas. The first one is related to devaluation. Here Buira argues that the devaluation’s (possible) depressive effects should be kept in mind during the elaboration of the programme, also stating that the time period over which devaluation is applied should be lengthened. The second weakness mentioned by Buira is related to the balance of payments and the degree of monetary deflation making a more flexible credit ceilings and more prompt reaction to their modification desirable. According to him, in this case the problem arises from the concept of sustainable BOP position that involves the estimation of certain economic variables which are difficult to quantify, calling for ‘substantial judgement’. As such a significant disparity becomes visible, regarding the uncertainties surrounding a sustainable BOP and the exact aim in domestic credit expansion set by the model. His third concern is related to the length of the programme and costs of the adjustments. While he never questions the necessity of some kind of intervention, he does question the priorities associated with some of the targets and the adopted time frame, arguing that too much pressure is put on trying to circulate the IMF’s resources leading to damaging prescriptions.

The main source of external criticism, regarding financial programming is the fact that financial support is always bound to so called conditionalities. It is argued that these conditionalities in many cases undermine social stability leading to increased poverty thus, contradicting the stated objectives of the IMF. These preconditions in several cases include the adoption of ‘austerity measures’ even in countries where the economy is already weakened. These measures may include an increase in taxes with the intention of reducing budget deficit, restrains on public spending – healthcare, education, social welfare – bringing budgets closer to balance.

The IMF’s recent approach was also criticised by Stiglitz (2002), arguing that the IMF, by converting to a more monetarist approach, had no longer a valid purpose as it was mainly designed to provide funds for countries to carry out keynesian reflations, and that the Fund was starting to mainly reflect the interests of the western financial community.

The overall success of the IMF is perceived as limited. The fact that it couldn’t prevent crisis from happening and there are some suggestions that it even caused the eruption of several crisis (for example the one in Argentina in 2001 according to many was caused by IMF required budget restrictions) is strongly discrediting many IMF policies.

Due to the fact that, as stated by Stiglitz, the IMF is advocating a monetarist approach and frequently campaigning for currency devaluation, criticism is mounting from the side of supply-side economists as the proposed measures are considered inflationary and a higher tax rate is considered to lead to economic contraction.

The reality is that the challenges that the IMF financial programming approach is faced with are nowadays completely different in complexity and structure making any attempt to tackle them more and more demanding. The reason is simple: the model was originally intended to deal with temporary crises, yet present day crises in less developed countries (LDCs) and more recently even in developed countries, is presenting new challenges highlighting the need for changes in the way these situations are approached.

Conclusion

The main conclusion that could be deducted from this paper is best synthesised by Agenor & Montiel (1999):

‘Although all of the models to be examined have been applied frequently in policy formulation in developing nations, we shall argue that all of them are subject to limitations that constrain their usefulness for both policy guidance and analytical work as medium-term models.’

This paper has made an attempt to briefly present a few of the reasons why in some cases the financial programming approach is seemingly not working. Discussions on the topic are very intense made only more central by the current crisis the world is facing. However, in the lack of a better alternative the best advice available is that one has to be aware of the model’s limitations in order to minimise the possibility of undesired outcomes and constantly make efforts to improve and tweak the structure of the model without losing its simplicity and transparency.