Macroeconomic indicators of DC’s indebtedness are supposed to measure or describe the degree of indebtedness of the concerned countries. They must bring out trends that may lead to economic, social or political difficulties; they must make it possible to correct potential deviations in a country’s state of indebtedness with regards to a particular reference state. Indebtedness indicators must also make it possible to measure the efficacy of a given action. In order to meet these goals, they must take into account the cause-and-effect relationships between a decision or action and its consequences (effect, impact, danger or risk) on the level of indebtedness and/or the burden it represents.
1. Necessary attributes of a “good” indebtedness indicator
The indispensable qualities of a good indicator
In general, it is highly desirable that an internationally used indicator:
(i) be as objective as possible, meaning that it is not designed to reflect the values or positions of the assessor (for example, comparing debt amounts with the salaries of civil servants working for International Institutions is not uninteresting, but there is not a link between these two measurements objective enough to make such an indicator indisputable),
(ii) measure results and not potential,
(iii) lend itself to an established distribution of obtained results,
(iv) be easily constructed and understood,
(v) make international comparisons possible.
On the other hand indicators must have a certain number of sometimes contradictory qualities:
(i) pertinence: the measurement must perfectly describe the desired phenomenon; it must accurately represent what is being measured and retain this representation over time
(ii) simplicity: the information must be easily and inexpensively obtained, so that the user can grasp it as directly as possible
(iii) objectivity: the indicator must be unambiguously calculable using observable variables
(iv) univocity: the indicator must fluctuate in a monotone fashion with respect to the described phenomenon to be able to unequivocally interpret these variations
(v) sensitivity: the indicator must move significantly in response to rather small variations in the phenomenon
(vi) precision: the indicator must be defined with an acceptable margin of error given the precision of the observable variable measurements
(vii) fidelity: if the indicator is biased with respect to the concept that it reflects, it must keep this bias constant over the spatiotemporal units of reference
(viii) auditability: a third party must be able to verify the correct application of the indicators’ usage rules (data collection, processing, formulation, distribution, interpretation)
(ix) communicability: the indicators must allow for a dialogue between populations with differing preoccupations
(x) acceptability: the indicator must be marketable and must not clash with the potential user’s culture
In short, the indicator must project an accurate image of the phenomenon in question to make possible a quick and easy evaluation of the data to be monitored.
An indicator possessing these various qualities can thus serve several functions:
(i) it can be used to measure the “performance” level of the studied phenomenon,
(ii) it can offer information that serves to maintain or improve this level,
(iii) it can allow us to detect faults, problems, irregularities and nonconformities in order to improve the observed “performance” level,
(iv) it can help us appreciate the progress that has been made and realize what is left to do.
Available and recognized data sources
When it comes to debt-related issues, there are, in the end, rather few data sources that can be called upon if we hope to strictly adhere to the criteria of simplicity and auditability. Objectively, the only “universally” available and recognized data comes from international institutions. Thus, despite the numerous objections to the collection methods and the pertinence of this data, we must base our indebtedness indicators on the data bases of the IMF and the World Bank. The vast majority of analysts use this approach, creating a sort of “shared statistical language” among them. In fact, the World Bank’s “World development indicators” and “Global development finance” will be used as the sole source of primary data. The homogeneity of data collection methods is the primary benefit of this choice, as it allows for comparisons between countries as well as over time. Regardless of the intrinsic quality of this data, it fulfils the important criteria of fidelity, which is no longer satisfied once we start using non-homogenised local data sources, which are often difficult to compare.
2. The meaning and the tone of the “official” indicators produced by the IFIs
There are three kinds of “official” debt measurements. On the one hand, a certain number of “absolute” measurements, evaluated in current dollars, provide annual summaries of total external debt and total debt service. The total external debt is divided into long-term debt and short-term debt, as well as private non-guaranteed debt and public and publicly guaranteed (PPG) debt. Total debt service is likewise divided into public and private debt service. On the other hand, the IFIs give us a number of ratios putting the aforementioned values in relationship either to the aggregate values of GDP or GNP, or to the aggregate value of goods and services exports (Present value of debt (% of exports of goods and services), Present value of debt (% of GNI), Total debt service (% of exports of goods and services), Total debt service (% of GNI), Central government debt, total (% of GDP), Debt service (PPG and IMF only, % of exports of goods and services), PPG debt service (% of central government current revenue)). Finally, a number of indicators characterize, describe, or reflect the composition of the debt and the funds flows, according to the type of lender, currency, terms and conditions, or the forms of loans concerned; debt redemption and reduction amounts are also exposed.
Concretely, it seems that all existing indebtedness indicators (the ratios) are destined to remain indicators of a country’s capacity to reimburse its debt, either by the wealth it generates (GDP or GNP), or more directly by the part of this wealth that allows the country to produce transferable currency (the value of their exports). These indicators are thus fundamentally aimed at lenders and only express the extent to which the net wealth of a concerned country will allow it to fulfil the “job” of a debtor.
3. The need for complementary approaches: a few proposals
Since available indicators do not say anything about the consequences of the debt on debtor countries, it seems necessary to produce additional indicators to assess the socio-economic impact of indebtedness and highlight the actual burden the debt represents for indebted countries. To do this, we first have to determine whether we measure the stock of the debt, i.e. the total amount of the debt as reassessed every year or the flow that consists of its annual servicing. Before we can decide on one or the other, we have to clearly identify the purpose served by indicators. If the purpose is to inform on the countries’ capacity to bear the burden of the debt or on socio-economic consequences, then we have to measure the debt service. Indeed, neither the accumulated stocks nor in fact the flows induced by the debt have any clearly established impact on a country’s development or on the burden it represents on its everyday survival. Only the service of the debt amounts to an actual financial burden that preys on available resources and prevents any other possible use. On the other hand, the flows involved in the debt service can easily be compared to a range of other financial flows during a definite lapse of time without there being any possible confusion of a flow / stock nature or conversely. The measure of the debt service flow should all the more serve as a first major reference as it is also seen as significant by the IFIs.
Indicators that can be suggested are of three kinds. The first category measures the actual "burden" of debt service on the people in indebted countries. In this respect we can assess the amount of the debt service per inhabitant, which highlights the debtors’ perspective. The second tries to establish the direction and the proportion of financial flows between debtors and creditors. We intend to assess the ratio between debt service and ODA. The third category aims at measuring the impact of debt servicing on the development of indebted countries. To this effect we will compare debt services with tax revenues, healthcare budget, education budget, total public service salaries, and finally capital expenditure. These various measurements will bring out how the debt burden is a handicap for development since it prevents other more productive uses of the available resources.
Let us finally add that measuring the ratio between debt service or debt stock and the degree of implementation of Washington consensus recommendations in the various countries could represent an interesting indicator of how successful structural adjustment plans are to counter indebtedness....
4. First statistical indications on the new indicators
The values for all seven proposed indicators have been measured in all countries in which data were available for 1970 - 2001. Here are the statistical results that can be deduced from these data:
— Debt service per inhabitant: in 2000, for 145 countries, the average amount is $ 112 per inh. Only for servicing the debt. It exceeds $ 36 / inh. in half of these countries and $ 149 / inh. in the most heavily burdened countries ( 25 %). The highest amount are to be found in Hungary ($ 1348 / inh.), Argentina, Croatia, South Korea, Mexico, Uruguay, and the Czech Republic where the debt service burden per inhabitant amounts to 5 to 10 % of the GDP/inh. Spreading to the right as it does, the distribution for this indicator shows that not many countries have very high values for this indicator (Mé = 36 $), but that these values are then extremely high.
— Debt service / net ODA: in 2000, for 135 countries, the average value of this indicator is 15 (!), but this value should be treated with caution owing to the significant variations in net ODA (Official Development Assistance) between countries. A more meaningful statement would be to say that in half of the indebted countries, the debt service burden is more than 30% higher than the amount of net ODA received. In the most heavily burdened countries (25%), the amount of debt service comes to more than 7.5 times the amount of foreign aid received. On average, for the period 1970-2001, it is impossible to arrive at an accurate indicator for certain countries, because the net aid is either globally negative, or infinitely small in comparison with debt service amounts - as is the case with Venezuela and South Korea. For the other countries, the ones with indicators showing the highest values (sometimes averaging 600!) are Turkey, Chile, Mexico, Argentina and Brazil. It should be noted, however, that for certain countries, especially the poorest, the indicator shows values below the figure of 1, indicating a net ODA amount higher than the amount of debt service (Togo, Madagascar, Bangladesh, Democratic Republic of the Congo, Ghana).
— Debt service / tax revenues: in 2000, for 57 countries, the average value of this indicator is 0.40, with a fairly normal distribution. On average, for the period 1970-2001, this indicator shows that for a number of countries, debt service represents more that half the annual tax revenues. The countries with the highest indicators are Bolivia (0.62), Ecuador (0.60), Papua New Guinea (0.59), the Philippines (0.58), Sierra Leone (0.54), Mexico (0.53) and Jordan (0.53).
— Debt service / healthcare budget : in 2000, for 133 countries, the average value of this indicator is 2.3, with a fairly normal distribution. On average, for the period 1990-2000, this indicator shows that for three-quarters of the indebted countries, debt service exceeds the annual healthcare budget. The countries with the highest indicators are Indonesia (16), Nigeria (13), Côte d’Ivoire (9), Angola (8), Morocco (7) and Cameroon (6).
— Debt service / education budget: in 2000, for 63 countries, the average value of this indicator is 1.5, with a fairly normal distribution. On average, for the period 1980-2000, this indicator shows that for 56% of the indebted countries, debt service exceeds the annual education budget. The countries with the highest indicators for this period are Nigeria (11), Indonesia (7), Zambia (5) and the Philippines (4).
— Debt service / total public service salaries: in 2000, for 56 countries, the average value of this indicator is 2.3, with an oblique distribution to the left (Mé = 1.3). On average, for the period 1990-2001, this indicator shows that for half the indebted countries, debt service exceeds total public service salaries. The countries with the highest indicators for this period are Democratic Republic of the Congo and the Philippines (5), Hungary (4), Colombia and Mexico (3).
— Debt service / capital expenditure: in 2000, for 58 countries, the average value of this indicator is 4.1, with an oblique distribution to the left (Mé = 2). On average, for the period 1970-2001, this indicator shows that for almost 60% of the indebted countries, debt service exceeds the amount of capital expenditure. The countries with the highest indicators for this period are of two types: very low capital expenditure, Democratic Republic of the Congo (9) or Haiti (3), or with - in relative terms - a very high debt burden, Argentina (6), Brazil (4), Uruguay (4).
Text written by François Combarnous, Economist, University of Bordeaux IV.