Resumen

El presente trabajo muestra una valoración de la efectividad de la Ayuda Oficial al Desarrollo (AOD) destinada a promover el crecimiento y reducir la pobreza en los países receptores, así como su capacidad para influir en las principales variables macroeconómicas como la inversión, el ahorro, el consumo, los impuestos y la tasa de inflación. Con esta finalidad, utilizamos un análisis clásico de regresión y un modelo estructural. En el primer caso, usamos los mínimos cuadrados ordinarios para calcular una serie de ecuaciones en las que AOD actúa como la variable independiente. El segundo caso comprende un modelo de indicadores y causas múltiples (MIMIC) con tres variables latentes, una de las cuales, la AOD, es exógena.

Abstract

The present paper presents an assessment of the effectiveness of Official Development Assistance (ODA) in promoting growth and reducing poverty in recipient countries, as well as its capacity to influence the main macroeconomic variables such as investment, savings, consumption, tax burden and the rate of inflation. For this purpose, we use both classical regression analysis and a structural model. In the former case, ordinary least squares (OLS) are used to estimate a series of equations in which ODA plays the role of independent variable. The second case involves a multiple-indicator, multiple-causes (MIMIC) model with three latent variables, one of which, ODA, is exogenous.

1.- Introduction


During the 1990s, the value of Official Development Assistance (ODA), in terms of the GDP of donor countries, decreased from 0.31% at the start of the decade to 0.22% by the year 2000. Even the slight recovery of this ratio to 0.25% in 2003 is insufficient and the problem of "aid fatigue" remains unresolved.

Hunger and the fight against poverty among the least developed nations are still priority targets of the principal international organisations. Numerous summits and meetings have been held in recent years, at all of which promises and good intentions have abounded. In 2000, the General Assembly of the United Nations approved the Millennium Declaration, which included the Millennium Development Goals (MDG), the chief objective of which was to eliminate extreme poverty and hunger in the world.

However, in order to achieve these MDG, it is essential for more assistance to be provided and, moreover, that it should be directed towards the most needy countries, under the most favourable conditions for its implementation. In other words, the debate on the effectiveness of Development Assistance remains an open question.

In this study we assess the effectiveness of Official Development Assistance (ODA) in promoting growth in recipient countries and in reducing poverty, and evaluate its capacity to influence the main macroeconomic variables, such as investment, savings, consumption, the tax burden and the rate of inflation. To achieve these goals, we use classical regression analysis and a structural model. In the former case, Ordinary Least Squares (OLS) are used to estimate a series of equations in which ODA plays the role of independent variable. In the second case, a Multiple-Indicator, Multiple-Causes (MIMIC) model is used, with three latent variables, one of which, an exogenous one, is ODA.

2.- The effects of Development Assistance on the principal macroeconomic indicators


The debate on the effectiveness of development assistance has centred on three types of dependent variable: firstly, the most significant economic indicators, such as growth (normally GDP or per capita GDP), consumption (either total or divided into public and private), investment (also as a total value or broken down), savings, incomes and the inflation rate. Secondly, the various measures of poverty: the number of poor (those living below a given income level), infant mortality and life expectancy. Finally, the different areas of public expenditure, in order to study the interchangeability of the assistance provided. Let us examine the first of these types, starting with growth, the most commonly applied.

The literature describes many variables related to economic growth, but these can be grouped into three broad areas:

- Macroeconomics: ranging from specific statistics such as investment (Boone, 1994; Durbarry, Gemmell, and Greenaway, 1998; Lensink and Morrissey, 1999, and Levy, 1988) and savings (Mosley and Hudson, 1999; Mosley, Hudson and Horrel, 1987 and 1992, and Singh, 1985), to the construction of indices that seek to determine the influence of macroeconomic stability on growth. The approach most often cited in the literature is that of Burnside and Dollar (1998 and 2000), who included the public budget, inflation and the degree of openness of the economy. The same type of index is also used in Dalgaard and Hansen (2001) and in Hansen and Tarp (2000 and 2001), although with different results.

- Political or social: for example, institutional quality (Burnside and Dollar, 2000), which considers the security of property rights and the efficiency of governmental bureaucracy; state intervention (Singh, 1985), or rates of school attendance (Lensink and Morrissey, 1999; Lensink and White, 2001, and Sachs and Warner, 1995), all of which are used as approaches to assess work productivity.

- Regional: these are dummy variables. Most of them refer to all or part of the three least-developed continents: Africa, Asia and Latin America. For example, Sala-i-Martin (1997) found that geographical location in Latin America or sub-Saharan Africa was prejudicial to development.

In principle, we shall focus on the first two areas. The base model for estimations consists of the following variables:

o Dependent variable: the annual rate of growth of GDP (dY/Y)

o The three main sources of finance for a developing economy: internal savings (S), external investment, specifically, foreign direct investment (FDI), and official development assistance (ODA). To avoid the problem of the size of the economy, the data for these three variables are stated as percentages with respect to GDP. Expected sign: positive.

o The increased value of exports (EXP): the aim of this is to reflect the influence of external demand on growth. Some authors (e.g. Mosley et al., 1987 and 1992) have used it with satisfactory results. The data are stated as annual growth rates. Expected sign: positive.

o Literacy rate (LIT): this, too, was used by Mosley et al. (1987 and 1992), although with less good results. This ratio represents an approximation to that of work productivity and the objective is to measure the effect of the work factor on growth. Other authors, such as Lensink and Morrissey (1999) and Lensink and White (2001), have used the rate of secondary school attendance, and others that of primary school attendance (Sachs and Warner, 1995). The data are expressed as percentages of the literate population. Expected sign: positive.

o Population growth (POP): Singh (1985) used this with some degree of success. The data are expressed as rates of annual growth. Expected sign: positive.

- Initial GDP per capita (GDP): the aim is to reflect the process of convergence, according to which the less developed countries, with a lower per capita income, should grow at a faster rate. Boone (1994), Burnside and Dollar (2004), Lensink and Morrissey (1999), Lensink and White (2001), (Sachs and Warner, 1995) and Singh (1985) used similar variables. The data correspond to the initial years of three periods that were studied: 1990, 1994 and 1999. Expected sign: negative.

- External debt (EXDBT): this parameter is expected to represent the obstacle presented to an under-developed economy by excessive indebtedness. This variable has not featured widely in the literature. The data are expressed as percentages of debt with respect to the GDP: Expected sign: negative.

All the data are obtained from the World Bank, except those for school attendance rates and aid assistance, which were obtained from the OECD. The period studied is that following the collapse of the Soviet bloc (1990-2003), divided into three sub-periods: 1990-93, 1994-98 and 1999-03 and data were obtained for 96 countries. Therefore, the equation to be estimated by Ordinary Least Squares (OLS) is expressed as follows:

Table 1 shows the initial results, with two main estimates: for the whole sample (columns 2 and 3) and excluding the European countries (columns 4 and 5). The signs are as expected, except in the estimate for the whole panel, in which the LIT variable is negative, although not significantly so. In any case, its coefficient is small, as was also found in the studies by Mosley et al. (1987 and 1992). On the other hand, in Mosley et al. (1987), the sign of this variable was found to be significantly positive for the period 1960-70.

Another difference between the two estimates is in the sign of the FDI variable, which was significant for the whole sample, but not so when the European countries were excluded. This puts into question the relevance of foreign investment for growth in developing countries. The other variables that were also found to be significant for growth were internal savings, increased exports and population increase. The ODA coefficients were small and non-significant, although positive and with a considerable increase in significance when the European countries were excluded.

In addition to the above, estimates were obtained (1) by adding, separately, the dummy regional variables of Europe, America, Asia and Africa. Only in the cases of Europe and Asia did the explicative power of the model increase, with belonging to Europe found to be prejudicial for growth. Doubtless, this reflects the consequences of the dismantling of the former Soviet bloc.

Table 2 gives the estimation results (1) obtained for each of the four continents mentioned, showing that the model functions best in the cases of Asia and Europe. Again, it highlights the importance of export growth, except in Europe, probably because economic growth there is of a more endogenous nature, with a greater influence of internal factors.

The two most noteworthy features of Table 2 are, on the one hand, the significantly positive coefficient of the literacy ratio for America and, on the other hand, the significantly negative coefficient of the ODA coefficient for Europe. This latter might be caused by the less effective distribution of assistance within Europe, where greater weight is given to criteria of a political or geostrategic nature or to the economic interests of donor countries. The ODA coefficients for the other three continents are positive, but not significantly so.

In Table 3 we present the estimation results (1) for each of the three periods separately, excluding the European countries. Once again, the EXP variable is found to be significantly positive, for all three periods. Especially noteworthy is the third period, with five significant variables, including the negative variable for initial per capita GDP; this indicates that convergence did occur, although with a very small magnitude. Another significant variable was the one of negative external debt, which reflects the difficulties created for economic growth by excessive external indebtedness. The ODA coefficient was significantly positive, which might reflect an improvement in the effectiveness of assistance during the third period studied.

Let us now consider the rate of internal savings as a dependent variable. This time, the exclusion of the European countries does not improve the results obtained by the model, and so it was decided to restrict the study to the whole sample. The explicative power of the model is then substantially increased, and the adjusted coefficient of determination rises from 0.32 (Table 1) to 0.46. Except for literacy rate and population growth, the other variables are significant as regards savings. A positive influence is exercised on savings by growth, initial per capita GDP and external debt, although its magnitude is very slight in the latter two cases. On the other hand, the influence of foreign investment, export growth and external assistance is negative. In other words, we must consider the hypothesis of the fungibility, or replaceability, of assistance (see Pack and Pack, 1993; Griffin, 1970, and World Bank, 1998).

To further investigate the influence of ODA on investment, we shall introduce as a dependent variable the rate of gross capital formation with respect to the GDP. The data for this are published by the World Bank. Table 4 shows that although the coefficients of determination are not especially high, some variables are significant. The positive ones for investment are GDP growth, savings, foreign investment, work productivity and even ODA. The first four of these effects are, theoretically, acceptable. In the literature, opinions are divided concerning the effect of ODA on investment. Boone (1994) obtained a significantly negative coefficient on estimating by OLS, but this significance disappears when Instrumental Variables (IV) are used. For Collier and Dollar (2001), the influence is positive, but not significantly so. Boone (1996) reported a positive, significant impact when IV were used for the estimation. Finally, Hansen and Tarp (2001) obtained a significantly positive coefficient, but estimating with fixed effects per country and using the Generalised Method of Moments (GMM).

Estimation by OLS, as presented in Table 4, confirms the existence of a positive effect of assistance on investment. However, the sample from which the European countries are excluded leaves the significance of this effect open to some doubt. Moreover, we should note the importance of foreign investment. This promotes growth in three ways (Hansen and Tarp, 2001): as an indicator of correct political and institutional functioning, by contributing to the accumulation of capital and by raising the total productivity of the economy (the latter being achieved by technology transfer).

Moreover, the convergence effect is confirmed, although judging by the coefficients of the GDP variable, its magnitude is very slight. In the regression analyses with regional variables, as was the case with savings, forming part of America was found to be negative for investment, while the latter parameter was positive for countries in Asia.

Let us now consider consumption, both public and private, as a dependent variable. Few studies have been made of the effects of development assistance on consumption. Boone (1994 and 1996) obtained significant positive coefficients when total consumption was used as the dependent variable, but when this term was split into private and public consumption, only the coefficient for the latter remained significant. Boone considered most assistance to be aimed at public consumption, having the effect of increasing the size of the administration but making no significant difference to indicators of poverty. Burnside and Dollar (2000) only studied the effect on public consumption, with the explicative variable "assistance" being divided into bilateral and multilateral. Both coefficients were found to be positive, but only that for bilateral assistance was statistically significant.

Our results, shown in Table 5, basically coincide with those of Boone and Dollar, in that assistance has a positive effect on public consumption. However, in our case the coefficient found for private consumption, though also significant, was found to be negative. In other words, assistance promotes public consumption but discourages private consumption. Other important points revealed in Table 5 are, on the one hand, the strong negative relation between savings and private consumption, as predicted by theory, and on the other, the fact that the two types of consumption should present such diametrically opposed behaviour patterns, as is reflected in the signs of the variables found to be significant: FDI, ODA, POP and EXP (the first three were negative for private but positive for public consumption, and vice versa for the last one). This contrary behaviour is also reflected in the signs of the fictitious regional variables and in those of the regression analysis by regions and by periods (not shown here).

We also used as dependent variables the rate of inflation (INF) (taking into consideration both the GDP deflator and the rate of retail price inflation) and the tax burden (T). In both cases, the data were provided by the World Bank. The results are very similar, as was to be expected, given the high degree of correlation between the two variables (0.86). In no case were the ODA coefficients significant. This latter result agrees with that of Boone (1996), who obtained coefficients of both signs but which in no case were significant.

With respect to the tax burden, we used the level of taxes as a percentage of GDP, using figures supplied by the central government and by the World Bank). For this parameter, too, we failed to obtain significant coefficients for the ODA variable, as did Boone (1996) and Pack and Pack (1993).

This section on the macroeconomic effects of assistance is summarised in Table 6, which shows the coefficients and the corresponding t values of ODA as an explicative variable for each of the other variables described. In all the regression analyses we use the whole sample (including the European countries), and the INF variable is applied as the GDP deflator. According to the results presented, during the period 1990-2003 assistance did not promote growth, nor did it have any significant effect on prices or on the tax burden. It did, however, displace savings (the most marked result of all) and private consumption (significant to the 6% level) and promoted investment and public consumption (significant to the 8% level).

The contrary signs for savings and investment are paradoxical and might appear to be contradictory. However, this is not so, according to the reasoning of Hansen and Tarp (2000), under the assumption that assistance has no impact on other external resources. Indeed, in an open economy, the well known savings-investment identity can be expressed in the following way:

where A is the assistance received and F represents other external resources. By expressing these variables as income fractions, we derive

Assuming that = 0, i.e., that assistance has no impact on the remaining resources derived externally, the marginal effect of assistance on investment reduces to

According to (4), only if the impact of assistance on savings were less than -1 would a negative effect of assistance on investment be expected. Otherwise, even if assistance had a negative influence on savings, its effect on investment should be positive.

Nevertheless, the initial assumption, that of the null effect of assistance on other external capital flows, is debatable. If the increase in assistance is interpreted as a sign of economic difficulties, then one would expect a reduction in private external capital movements. On the contrary, if it is interpreted as a support for the macroeconomic stability that has been achieved, then one would expect such movement to increase. From the data currently available, the former situation seems to apply, because when FDI is introduced as a dependent variable, in all the above regression analyses, the coefficients for ODA are always negative, reaching a level of significance of 6% in one case.
Despite the above, Table 6 indicates the existence of a positive effect of assistance on investment. Therefore, the paradox is centred on the significance of this coefficient and on its non-significance in the growth equation. In other words, why is it that assistance promotes investment but not growth?

Our conclusions agree, in part, with those of the main studies reported in the literature. For example, for Boone (1994), assistance is not a suitable method for creating growth. According to this author, two thirds of the assistance is channelled towards public consumption and one third towards private consumption. Therefore, incentives to investment are low in countries receiving development assistance (the corresponding coefficient is slightly negative and non-significant). In other words, assistance promotes, above all, consumption, rather than investment.

Other authors are not so categorical, and relate the effectiveness of assistance to the fulfilment of certain characteristics in the recipient country. Singh (1985), for example, claims that foreign assistance as such cannot sustain a high rate of growth unless other factors, favourable for growth, are also present. Specifically, in this study the introduction of state intervention as a variable reduces the statistical significance of development assistance, which leads us to believe difficulties might arise in achieving effective assistance in environments where the level of state intervention is excessive.

Similar conclusions have been drawn by the World Bank (1998) and by various authors, including Burnside and Dollar (1998, 2000 and 2004), Collier and Dollar (1999) and Durbarry et al., (1998). Their main finding is that foreign assistance is only effective in a suitable macroeconomic framework. Among countries applying appropriate policies, those receiving large amounts of assistance have grown much faster (3.7% p.a.) than have those receiving small quantities (2.2%). For countries where policies are not so good, no positive relation has yet been established between the quantity of assistance received and the growth rate (Burnside and Dollar, 1998). This relation between the policies applied in a country and the effectiveness of the development assistance received seems to be more marked in countries where incomes are lower (Burnside and Dollar, 2000).

The views of other authors have evolved over time. Mosley et al. (1987) were sceptical about the effectiveness of assistance to promote growth. However, Mosley et al. (1992) recorded a partial regression coefficient that, though small, was significant and positive. The introduction of a dummy variable for policies showed that political orientation did not have a significant independent influence on the effectiveness of assistance. However, these authors did accept that there could be an indirect influence via its impact on exports, non-concessional funds and on the assistance itself.

Mosley and Hudson (1999) analysed two periods separately, the years 1969-80 and 1981-95. The effect of assistance on growth was found to be clearly non-significant for the first period, and positive and statistically significant for the second. With respect to its effect on investment, it was significantly negative for the first period, and positive, but not significantly so, for the second.

For other authors, the factors determining the correct functioning of development assistance are related particularly to the donors. For example, for Lensink and Morrissey (1999), the effect of assistance on growth may depend on the uncertainty concerning the provision of assistance. In this study it was shown that when uncertainty was controlled, the assistance had a robust effect on growth, via investment. This suggests that a stable donor-recipient relation may be a positive factor for the effectiveness of assistance.

Other studies have been more categorical, claiming that assistance has a positive effect on growth. For example, Lensink and White (2001) obtained a significant positive coefficient for the relation between assistance and growth. Levy (1988) concluded that assistance is positively correlated with investment and with economic growth, although this analysis was limited to sub-Saharan countries. Hansen and Tarp (2000 and 2001) shared this opinion; they stated that external assistance raised the levels of savings, investment and growth, and that this positive influence was not affected by the political index established by Burnside and Dollar (2000). Dalgaard and Hansen (2001) concur with these views, and consider it premature to apply selective political norms in assigning development assistance, as has been proposed by the World Bank (1998).

3.- The causal model



The structural model proposed above was designed on the assumption that there may be a causal link in the fight against poverty, as regards official assistance and aid to the private sector. Under the estimation method applied in the present study (multiple-indicator, multiple-causes; MIMIC), it is assumed that the changes produced in certain variables (causes) are related to changes in others (effects), but that the contrary is not so, i.e. the cause always precedes the effect.

Among possible effects, we shall distinguish direct and indirect ones, which are represented by means of a causal diagram. The variables shown in circles are latent, while those in rectangles are observable. The arrows indicate the directions of the effects.

Let us consider a structural model in which there exists a relation among three latent variables; two of these, Official Assistance and Assistance to the Private Sector, are exogenous, while the third, which we term "Effectiveness in the fight against poverty in developing countries", is endogenous. Each of these latent variables is measured by a set of indicators related to economic and social aspects of the countries receiving assistance. All these data are obtained from an OECD Database (Geographical Distribution of Financial Flows).

The following indicators are considered for the three latent variables:

1) Effectiveness (endogenous variable)

By means of the set of indicators comprising this variable, we seek to measure the effectiveness in the fight against poverty, from various viewpoints: on the one hand, economic aspects are taken into account, via the per capita GDP; on the other, the social aspect is considered, using the rate of school attendance; and finally, humanitarian aspects are introduced in the form of the rate of infant mortality.

- Secondary schooling or level of population attending school.
- Infant mortality for children aged under 5 years.
- GDP p.c. The per capita GDP is commonly used to compare poverty levels in the populations of developing countries.

2) Private Sector (exogenous variable)

In this study, the group of indicators that are included under the heading "Private Sector" are those that may have multiplier effects on the economy, namely:

- Savings
- Exports, used as an indicator of economic activity in the sense that greater participation by exports in the national economy could mean a higher degree of economic openness, more incentives for the private sector, etc.
- Net Flows of Private Capital. These are private investments made from abroad.

3) Official Assistance (exogenous variable)

This latent variable is measured from the Official Development Assistance received by developing countries, distinguishing between bilateral and multilateral assistance.

- Bilateral Assistance is that which is granted directly by a donor country to a recipient country. Currently, this is the most important mode of assistance. It has been the object of much criticism, it being argued that the assistance may be tied to the interests of the donor country.
- Multilateral Assistance is that which is granted by International Organisations such as the World Bank, the IMF and the EU. In these cases, the assistance provided is less tied to the interests of the donor.
- Donations. This variable is also part of Official Assistance, but relates to capital movements that are not related to international loans.

The equation system to be estimated is then as follows:

where is the latent endogenous variable Effectiveness, where are the two latent exogenous variables Official Assistance and Private Sector and is the random perturbation.

The measurement equations for the latent endogenous variable are:

The measurement equations for the latent exogenous variables are:

To estimate the possible effects, we now link the covariance matrix of the observed variables , the relations of which are given by the specification of the model on the sampling covariance matrix S. Then, by applying the maximum likelihood method,

The results of the estimations are shown in the causal diagram.

Causal Diagram




The values beside the arrows represent the sign and the intensity of the different effects.

Before going on to interpret the estimated direct and indirect effects, the quality of the results is assessed by application of the statistical contrasts normally applied to models with latent variables.

a) Individual Contrasts

The final column in Table 7 shows the Critical Ratio (C.R.). If its absolute value is greater than two, then the estimator is significantly different from zero.

In the estimates obtained in the present structural model, all the values of this parameter are greater than two, and so the estimates are statistically significant.

b) Global Contrasts

The CMIN/DF Contrast, which is shown in the final column of Table 8, is used as a measure of good fit. It is defined as the ratio between an and its degrees of freedom (DF). There is no general rule, but most practical studies apply one or other of the following interpretations: for CMIN/DF values less than 2, the fit is adequate; alternatively, for values between 1 and 3, the fit of the theoretical model to the sampling data is acceptable. In any case, the estimated value for CMIN/DF in the present study is close to 2 and so, in principle, the fit is a good one.

The values of the goodness of fit index (GFI) are less than or equal to one. A value equal to one means the fit is perfect. In our model GFI = 0.839. The adjusted goodness of fit index (AGFI) varies from GFI in that it takes into account the degrees of freedom of the model. If its value is close to one, the fit is good.

Table 9 shows, for the estimated model (Default Model) that the values of both GFI and AGFI are close to one, and so the model´s fit is good.

Interpretation of the effects

Interpretation of the effects is carried out at three levels: direct effects, which are those occurring between two variables without being influenced by a third one; indirect effects, those occurring between two variables but which are influenced by another; and, finally, the total effects observed between two variables, taken as the sum of the direct and indirect effects.

Table 10 shows the estimated direct effects. The direct effect of the Private Sector on effectiveness in the fight against poverty is zero, which seems logical, as the goal of the private sector is to maximise profits and not to fight against poverty. On the contrary, and as is to be expected, the effect of Official Assistance on effectiveness in the fight against poverty is positive.

The direct effects of the variables Savings, Private Capital Inflows and Exports are positive, i.e. these are variables that have multiplier effects on the rest of the economy, thus promoting growth.

The direct effects of the variables Bilateral Assistance, Multilateral Assistance and Donations on the variable Official Assistance are positive.

The direct effects of the variables Secondary Schooling and per capita GDP on Effectiveness are negative, while those of Infant Mortality are positive. These signs could be interpreted as meaning that, in the context of the currently established system of development assistance, effectiveness in the fight against poverty in developing countries is reduced when levels of school attendance are very low, when the population is very poor and when the rate of infant mortality is high.

Table 11, concerning indirect effects, shows that the Private Sector has a negative indirect effect on infant mortality. The indirect effect is positive on schooling and per capita income. The indirect effects of Official Assistance are negative on schooling and per capita income, and negative on infant mortality.

To sum up this section on the effects, it can be concluded that Official Development Assistance is not effective in the fight against poverty, in the sense that it does not enable developing countries to overcome their poverty, the rate of school attendance does not rise and that of infant mortality does not fall. However, it is important to note that a large proportion of Official Assistance is oriented towards extremely poor countries, with few resources in the fields of healthcare and education, and thus even if the amount of assistance to these countries is increased, their situation will continue to worsen. The donation of funds is not an effective instrument in the fight against extreme poverty, as domestic reforms are required.

With respect to the Private Sector variables with a stimulant effect, it is interesting to note the final effects of Exports (0.53*-0.43*0.93 = -0.212), reducing infant mortality, raising per capita income (0.53*-0.43*-0.7 = 0.16) and increasing the rates of school attendance (0.53*-0.43*-0.86 = 0.196). The other Private Sector variables, Net Private Funding and Savings, have a final effect similar to that of Exports. Therefore, it can be stated that, although the direct effect of the Private Sector on Effectiveness in the fight against poverty is negative, its final effects are positive, as the rate of infant mortality is decreased, the levels of individual incomes rise and the rate of school attendance is improved. Nevertheless, it should be borne in mind that developing countries in receipt of external private funding, that enjoy a significant volume of savings and that have a dynamic foreign trade sector are the ones where the standard of living rises and where better social and healthcare coverage is available.

4.- Conclusions

The main conclusion derived from this study is that the present system of international development assistance is not effective; not only as concerns economic growth but also, and this is what is really important, in improving certain indicators of poverty such as the rate of secondary school attendance and that of infant mortality.

We believe the principal cause of the failure of Official Assistance to provide more beneficial effects is the influence exercised by the political, economic and geostrategic interests of donor countries when assistance is provided. This phenomenon has been highlighted in the literature for decades (see, for example, Maizels and Nissanke, 1984 and McKinlay and Little, 1979), but the situation does not seem to have improved significantly.

However, despite the ineffectiveness of Official Assistance, we are of the opinion that more should be provided, but channelled mainly through multilateral organisations, to avoid as far as possible the distortions arising from the special interests of the donor countries. Such an increase in assistance should be carried out prioritising the criterion of effectiveness, by aiming it towards the most needy countries and social sectors.

Bibliography


Boone, P. (1994): "The impact of foreign aid on saving and growth". London School of Economics.

Boone, P. (1996). "Politics and the effectiveness of foreign aid". European Economic Review, vol. 42, No. 2, pp. 289-329.

Burnside, C and Dollar, D., (1998). "Aid, the incentive regime, and poverty reduction". Policy Research Working Paper (World Bank), No. 1937.

Burnside, C and Dollar, D., (2000). "Aid, Policies and Growth". American Economic Review, Vol. 90, No. 4, pp. 847-868.

Burnside, C and Dollar, D., (2004). "Aid, Policies, and Growth: Revisiting the Evidence". World Bank Policy Research Working Paper, No. 3251.

Collier, P. and Dollar, D. (1999). "Aid allocation and poverty reduction". World Bank Policy Research Working Paper.

Collier, P. and Dollar, D. (2001). "Development Effectiveness: What have we learnt?". World Bank Development Research Group.

Dalgaard, C.J. and H. Hansen, (2001). "On Aid, Growth and Good Policies". Journal of Development Studies, Vol. 37, No. 6 (Aug.), pps. 17-41.

Durbarry, R., Gemmell, N. and Greenaway, D. (1998). "New evidence on the impact of foreign aid on economic growth". CREDIT Research Paper, 8.

Griffin, K., (1970). "Foreign capital, domestic savings and economic development". Bulletin of the Oxford University. Institute of ECO, Vol. 32, pp. 99-112.

Hansen, H. and Tarp, F., (2000). "Aid Effectiveness Disputed". Journal of International Development., 12, pp. 375-398.

Hansen, H. and Tarp, F., (2001). "Aid and Growth Regressions". Paper to the Joint Development Centre/DAC Experts Seminar on "Aid Effectiveness, Selectivity and Poor Performers".

Lensink, R. and Morrissey, O. (1999). "Uncertainty of aid inflows and the aid-growth relationship". CREDIT Research Paper, 3.

Lensink, R. and White, H., (2001). "Are There Negative Returns to Aid?". Journal of Development Studies, Vol. 37, No. 6 (Aug.), pp. 42-65.

Levy, V. (1988). "Aid and Growth in Sub-Saharan Africa". European Economic Review, Vol. 32, No. 9, pp. 1777-1795.

Mosley, P., Hudson, J and Horrel, S. (1992). "Aid, the Public Sector and the Market in Less Developed Countries: A Return to the Scene of the Crime". Journal of International Development, vol. 4, nº2, pp. 139-150.

Mosley, P. and J. Hudson, (1999). "¿Ha mejorado la eficacia de la ayuda?". Información Comercial Española, 778, pp. 13-30.

Mosley, P., Hudson, J and Horrel, S., (1987). "Aid, the Public Sector and the Market in Less Developed Countries". Economic Journal, Vol. 97, No. 387, pp. 616-641.

Pack, H. and Pack, J.R., (1993). "Foreign aid and the question of fungibility". Review Economics and Statistics, 75, pp. 258-265.

Sachs, J.D. and A.H. Warner, (1995). "Economic Reform and the Process of Global Integration". Brooking Papers on Economic Activity, Vol. 1, pp. 1-118.

Sala-i-Martin, X., (1997). "I just ran two million regressions". American Economic Review, Vol. 87, No. 2, pp. 178-183.

Singh, R.D., (1985). "State intervention, foreign economiceconomics aid, savings growth in LDCs: someSome recent evidence". Kyklos, Vol.vol. 38, pp. 216-232.

World Bank, (1998). "Assessing Aid: What works, what doesn´t, and why".Work, What Doesn´t, and Why". Oxford University Press,. Inc.

About the authors:

Autor: José Boza Chirino
Dirección: Departamento de Métodos Cuantitativos en Economía y Gestión. Universidad de Las Palmas de Gran Canaria
Correo electrónico: jchirino@dmc.ulpgc.es

Autor: Juan Miguel Báez Melián
Dirección: Departamento de Economía y Dirección de Empresas.
Universidad de Zaragoza
Correo electrónico: jmbaez@unizar.es

Autor: Marta Wood Valdivielso
Dirección: Departamento de Métodos Cuantitativos en Economía y Gestión. Universidad de Las Palmas de Gran Canaria
Correo electrónico: mwood@fundescan.com

DOCUMENTOS DE TRABAJO EN ANÁLISIS ECONÓMICO (EAWP)
Derechos reservados 2002. El permiso para reproducir algún artículo está garantizado si Documentos de Trabajo en Análisis Económico lo acredita, las copias no son vendidas y es en acto de mayor difusión del documento.

Editor: Fernando González-Laxe. (Universidade da Coruña)
Director: Venancio Salcines. (Universidade da Coruña)
Subdirector: Andrés Blancas. Instituto de Investigaciones Económicas (UNAM)
Editor Asociado para America Latina: Luis Miguel Galindo. Facultad de Ecomomía (UNAM)