The “Commitment to Development” Index 2006 of the “Center for Global Development”
The “Commitment to Development”Index (CDI) 2006, published last August, has received much publicity, and France’s poor ranking has been noted by numerous media in several countries. The premises on which this result is based call for a detailed analysis in order to explain this ranking. This week, the Notes du jeudi (Thursday Notes) clarifies the content and significance of this interesting document.
For the fourth consecutive year, the Center for Global Development (“CGDev”, a highly reputed think tank based in Washington) has published, in August, its “Commitment to Development Index” (CDI), supervised by David Roodman. The commitment to transparency, consistency and comparison demonstrated by this publication makes it possible to open a useful debate.
An Indicator of the Consistency of Donor Policies
The objective of the exercise is to grade the wealthy countries (21 countries, all international aid donors and members of the Development Assistance Committee (DAC) of the OECD) according to whether their public policies are favorable or unfavorable to the development of the poorest countries  (to their “capacity to live a free life and have access to a minimum of health and education”).
This way of quantifying the “commitment to development” not only enables to take into account international public aid, but also to approach questions concerning the consistency of different policies. The index is calculated by using seven indicators that are supposed to represent all aspects of the policies that have an impact on the developing countries: international aid, trade, investments, migration policy, environment, security and technology. No weight is assigned to the criteria, each one counting as much as the others. Development aid, considered under both quantitative and qualitative aspects, is only one of these criteria. Its weight is thus limited in the final result.
For the 21 countries classified, the overall score varies between a little more than 3 and a little more than 6, the average being 5.2 (see appendix 1). France has a score of 4.6, which has varied little (-0.1) since 2003 (if the comparison is made by applying the 2006 calculation method, which is different than the one from 2003). This score places France at rank 18 in 2006, a slight fall from the previous year. The first four are: the Netherlands, Denmark, Sweden and Norway. The last three are Japan, Greece and Italy. Germany is 9th with 5.3, the United Kingdom 12th with 5.1 and the United States 13th with 5.0.
The methodology has changed since the first editions of the classification and these changes have led the CGDev to recalculate earlier results and thus to modify the rankings a posteriori. Thus upon publication of the 2004 rankings, France was ranked at number 8, but today it is ranked 16th for that same year.
The main interest in the indicator, one of the only ones on the world scene concerning the consistency of policies in relation to the poor countries, is that it allows the situation of one country to be compared to others on the basis of clearly explicit indicators in order to question the various policies that this country carries out. But in order to do that, it is necessary to know the methods used to calculate the index because it is impossible to quantify a reality as complex as the “commitment to development” at the international level without making numerous hypotheses. No harm is done to the “scientific nature” of the index to recall that its significance is totally inseparable from the hypotheses that have made possible its calculation.
The Implicit Weighting of the Criteria: Example of Arms Sales
Among the numerous criticisms raised since the publication of the 2003 index (methodological weakness, age and lack of reliability of the data, obvious errors and misunderstandings, ideological biases), the principle of giving equal weight to all areas is often brought up. According to Roodman, any other weighting would be arbitrary, given the actual state of knowledge about the relative importance of each of the components as a factor of development. Nevertheless, it seems surprising to give as much importance to the Security and Migration components as to the Aid, Trade and Investment components. By giving 60% of the weighting to the Aid, Trade and Investment components, for example, and the remaining 40% to the other components, the classification is radically different. France returns to the first half of the rankings and the USA goes back to last place. The weighting is the major challenge of such an exercise and must be continually questioned.
The discussion of weighting is all the more important since, if each criterion in principle “weighs” as much as the others, the same cannot be said for the components and sub-components that make up those criteria. Analysis of the “security” criterion and the grade given to France in this area illustrate perfectly the “weight of hypotheses” and the necessity of not discussing the final score without relating it to the calculation methods that produced it.
France would be placed in an above average position in the general rankings for 2006 (7th or 8th) if it were not handicapped by a disastrous grade (only 0.5, dead last) for the “security” criterion. The latter is itself quantified by taking into account several parameters: (1) contribution to peacekeeping operations + (2) securing international shipping routes + (3) arms sales. The first two parameters together account for 75% of the grade, the third for 25%.
The grade obtained for the first parameter is good: France is ranked 9th out of 21. The grade for the second is even better: France is ranked 2nd out of 21. It is arms sales to “poor countries or those having non-democratic governments” that penalizes France heavily. This sector certainly occupies an important place in French exports, but other countries are quite active in these markets without that having the same consequences for their ranking.
This sub-criterion takes into account the volume of arms supplied, as a percentage of GDP, since 2000 for a group of 12 countries selected because their index of democracy (according to one of the indicators used by the World Bank) is below zero and because the portion of their GDP devoted to armaments is greater than 2.39%. The value of the arms supplied to each of the 12 countries is estimated by SIPRI of Stockholm (Stockholm International Peace Research Institute). It is multiplied by an index of poverty, by the index of (poor) democracy and by that part of the GDP devoted to armaments beyond the 2.39% mentioned above. The 12 countries that result from this method are Jordan (multiplier coefficient of 3.04 in 2004), Saudi Arabia (3.0), Oman (2.85), Pakistan (2.83), Morocco (1.10), Colombia (0.74), Algeria (0.64), Lebanon (0.57), Egypt (0.42), Turkey (0.15), Sri Lanka (0.07) and Singapore (0.03). No sub-Saharan country is involved. In this list, it should be pointed out that the bulk of the penalty attributed to France is linked to arms sales to Saudi Arabia (70%), Pakistan (20%) and Turkey (7%).
This component could be challenged and the choice of countries retained by recalling that, in relation to arms exports, France scrupulously respects the code of good conduct adopted by the European Union. This code includes, in particular, criteria related to respect for human rights, the internal situation and compatibility with the technical and economic capabilities of the client country. But it is the implicit weighting of this component that should be stressed here.
In fact, to make it so that the countries that export no arms receive a grade of 10 and that the average be 5, the CGDev makes additional, unexplained calculation hypotheses that result in several countries being given a negative grade, even a strongly negative one. It is the only criterion or sub-criterion where the grades depart so much from the normal range of 0 to 10, which completely increases the theoretically modest weight given to this sub-criterion in the “security” component by the CGDev.
The act of selling arms to Saudi Arabia (and to a lesser extent Pakistan), taking into account the calculation hypotheses used, gives France a score of -16.8 as arms supplier. The United Kingdom has -17.3, partially compensated by a better grade for the contribution to peacekeeping, although the operations in Iraq are not taken into account. 
Thus, as a result of an apparently harmless hypothesis of an average of grades standardized at 5, a simple sub-component that weighs 25% of one of the 7 criteria has a considerable weight in the final result. For this reason, only French arms sales to one or two countries cause France to fall from 8th or 9th place to 18th in the overall rankings. 
The Ambiguity of the Migration Criterion
France is ranked 16th for the migration criterion. The “strong point” reported by the CGDev is the significant proportion of foreign students from developing countries (5th out of 21). On the other hand, France is criticized for the low rate of welcoming unskilled immigrants from poor countries (17th out of 21).
According to Roodman, the migration component is essentially aimed at taking into account the magnitude of the potential transfers to the countries of origin. This in itself is questionable, because transfers from migrants can certainly be financial flows that contribute to development, but also represent the compensation for the departure of potential factors of production. What is more, in the absence of reliable measures of financial flows, the CGDev opts for indicators of human flows (foreign students, flows of unskilled immigrants in the course of the 1990s, the same for 2001). This approach has several limits: (i) It takes into account only recent flows of immigration and not the total immigrant population (as a percentage of the total population). In the case of France, for example, the large waves of immigration took place in the 1970s and 1980s and even if the flows have ceased over the course of the 1990s and 2000s, the present immigrant populations are not a less important source of financial transfers to their countries of origin. (ii) The impact of emigration flows on the economies and institutional systems of the countries of origin is debatable, in particular concerning the problem of the brain drain. (iii) The CDI does not take into account the specificities of emigrant populations and their impact on the integration capacity of the host countries.
The Other Criteria
As regards development aid, France, with 4.1, obtains 9th place. This grade and the ranking are respectable, taking into account the criteria that are clearly unfavorable to us. In fact, while France’s overall high and infrequently linked ODA (Official Development Assistance) is recognized, it is considered to be too dispersed and insufficiently selective (a smaller portion to countries with low incomes and too large of a portion to less democratic countries). A more detailed analysis is provided in Appendix 2.
It will be noted that France is in first place as regards “technology”, due to high public budgets for research. Also praised is the possibility in France to cancel patents that are not used by those who registered them. However, this technology criterion does not give rise to very significant differentiation between countries.
Concerning trade, France is treated much like all of the European Union countries, which places it around the median, while the United States is 2nd, which explains, in large part, its rise in the general rankings. France is also close to the median for the “investment” and “environment” criteria.
The Consistency of Policies: a Central Question
The CGDev index is an interesting attempt to make, in each country, the components of a general policy consistent in relation to the developing countries, within the context of globalization, on condition of never separating the result from its calculation hypotheses.
This subject of consistency of public policies is at the center of much thinking, in Europe as well as the United States. The CGDev index illustrates the American perspective quite well, with a very specific “vision” of development. By the way, this index was above all devised to be an instrument for internal lobbying in the United States. It was, in part, designed to promote certain reforms in American aid, whose adoption has resulted in a clear rise of the United States in the classification (20th in 2003, 12th this year, mainly due to trade). On the European side, upon the initiative of the Austrian and then Finnish Presidencies, the European Commission has launched an extensive study, in cooperation with the member States. It deals with the examination of 12 sectoral policies that have an effect on the countries of the South. It concerns both Community policies and the national policies of member States.
France, which is strongly involved in the European effort, sees the CGDev initiative as an additional source for very useful considerations. This is why, like several ranked countries, it supports this initiative and attentively monitors the methodological questions posed by the design of such an index in direct connection with the researchers of the Center.
France’s Rankings and Grades for the Seven Criteria of the CGDev Index
|Rank (out of 21)||Grade||Median||Average||Grade of the First||Grade of the Last|
|Aid||9th||4.1||3.3||4.4||10 Denmark||1.1 Japan|
|Trade||10th||6.0||6.0||5.5||7.6 N Zealand||-0.4 Japan|
|Investment||14th||5.9||6.2||6.1||8.6 UK||2.5 Ireland|
|Migration||16th||2.6||4.8||4.6||10.5 Austria||1.4 Portugal|
|Environment||11th||6.1||6.1||5.8||7.8 UK||3.2 United States|
|Security||21st||0.5||5.6||4.8||8.1 Norway/Aust||0.5 France|
|Technology||1st||6.9||5.1||5.1||6.9 France||3.0 Greece/Ire.|
|Average||18th||4.6||5.2||5.2||6.6 Netherlands||3.1 Japan|
Grades of the 21 Countries
Appendix 2: Analysis of the Grade Given to France on the Criterion of INTERNATIONAL AID
France is ranked 9th for the “aid” criterion, with a grade of 4.1 (10 for Denmark, 4.6 for the United Kingdom, 3.3 for Germany, 2.2 for the United States, 1.1 for Japan, the average for the 21 countries being 4.4 and the median 3.3).
This ranking results from:
1) the total amount of the aid paid by each country, following a critical analysis of the “true aid” paid;
2) a coefficient that measures the quality of the aid, as a function of three criteria:
a) the linking of the aid;
b) its selectivity, with two dimensions: the level of poverty of the recipient country (gross national income) and its governance index (the index used by the authors is the Kauffmann-Kraay index for the year 2000);
c) the distribution of the aid: small projects, which monopolize the human resources of the host countries, are justified for small donors, but not for the large, which receive a penalty.
3) the addition of multilateral aid, then private aid (“charitable giving”). For aid implemented by multilateral institutions, the total amounts are measured and adjusted according to points 1 and 2 above, then added to the bilateral aid of the 21 countries according to their portions in these organizations. Since multilateral institutions are, on the whole, well graded on all criteria, the countries that provide a large part of their aid through multilateral institutions are favored. For private aid, a complex calculation ends up raising American aid by 25%.
According to these calculations, France’s public aid, which was $8.5 billion net in the 2004 declaration to the OECD’s DAC, represented $8.1 billion gross bilateral aid for all the beneficiary countries taken into account, of which $4.8 billion is eligible according to the criteria of the CGDev (point 1). Taking into account the below par ratings (point 2), it would represent (without private aid) only $1.47 billion of “really useful” bilateral aid, reaching $3 billion with multilateral aid included.
|(in billions of $)||ODA 2004 (DAC)||Gross bilateral ODA,
lists I+II (DAC)
|Retained by the CGDev (point 1)||The same after below par ratings (point 2)||Multilateral contribu-
|Total bi+ multi||Private contri-
tion of GDP (%)
On the first point (total amount of aid), the experts of the Center for Global Development (CGD) base their analyses, in an unaccustomed way, on the DAC’s statistics that include the ex-“list 2”countries such as the former Soviet bloc European countries as well as Singapore and Israel. On the other hand, they seem to exclude the TOM (Overseas Territories, which do not appear in the lists). They logically deduct loan reimbursements, such as done by the OECD’s DAC, but also the interest paid on these loans, contrary to the DAC.
They exclude cancellations of debts concerning old loans or without any relation to ODA. Their calculations entail, for all the large donors, a strong diminution of their “real aid”. On this level, France is no more criticized than other countries and it remains in a very good place (before the United Kingdom, Germany and Japan) at the end of this analysis of the total net amount of its aid.
è It is to the second point that the authors’ principal criticism of France is directed, with the mediocre selectivity of French aid and its distribution, such as measured by the index. On the other hand, the grade that evaluates the degree of linking for the aid is satisfactory (unlike the United States).
Selectivity: the index is still based solely on the Burnside and Dollar study of 1997, concerning the relation between aid and growth, which served as the basis for the World Bank 1998 report Assessing Aid, as if this study had not been subject to vigorous criticism from the academic community.  France is heavily penalized (0.43 against 0.58 for the United Kingdom) because of legitimate preferences stemming from specific connections and a comparative advantage in assisting certain countries. A number of these principal partner countries (at least in 2004) are not LDCs (less developed countries) or have a poor governance grade.
Distribution (“proliferation of projects”): the selected criterion is complex and new. Its intention is to penalize projects that have a small total amount in relation to the size of the aid from the donor. It thus indirectly favors small donors that undertake large projects, the best grades being given to the African Bank and the Asian Development Bank. The qualitative aspect is thus strongly debatable, the idea being limited to avoiding overtaxing the administrative capabilities of partner countries. France’s grade (0.64), without being bad (it is equal to Germany’s), is below the United Kingdom’s (0.77).
All in all, the “aid” criterion of the CGDev index comes to be added to other rankings or critical analyses of public development aid, such as those of the OECD’s DAC, the British NGO, Reality of Aid, the French platform, Coordination Sud, or even the European platform, Concord.
It ends up stimulating debate on aid to countries having poor democratic governance: should donors lose interest in Haiti, Afghanistan or Liberia under the pretext that aid in those places would be two times less “useful” than in China, to take examples from the supplied technical appendix? Contrary to the somewhat simplistic perspective take up by the CGDev, the debate on the optimization of aid allocation is far from being closed.
 Understood not only as the developing countries listed by the DAC, but also those on the old “List II”, which includes countries such as Israel, Singapore and the countries of Eastern Europe, including those that are now members of the EU.
 The United States has “only” -4.3 because arms sales are weighted by being related to GDP.
 France has already informed the CGDev of its methodological concerns regarding the armament sub-criterion. The next editions should rectify this obvious bias and search for other distortions of this type among the numerous data making up this synthesized index.
 See notably “Aid Selectivity according to Augmented Criteria” by Jacky Amprou, Patrick Guillaumont and Sylviane Guillaumont, to be published in World Economy.