A Single Comment — Permalink
© ICCforum.com, 2010–2024. All rights reserved. Policies | Guidelines
Featured Comments
- miltonlaw: Africa and the Concept of Positive Complementarity The answer to the allegation that Africa is inappropriately targeted by international criminal court could as well lie in the sui generis concept of positive complementary. My doctoral thesis research title is: The international criminal court and positive complementarity: Institutional and legal framework. I. Introduction It is the... (more)
- almariam: Saving the ICC: A Proposal for a Witness Protection Program Justice delayed, again? In late January of this year, I wrote a commentary entitled, “Kenyatta at the ICC: Is Justice Deferred, Justice Denied?” In that commentary I openly expressed my angst over the endless delays, postponements and backpedalling talk about “false evidence” and “lying witnesses” surrounding the Uhuru Kenyatta trial at The Hague. I felt there was perhaps... (more)
- almariam: Kenyatta at the ICC: Is Justice Deferred, Justice Denied? I am getting a little jittery over the repeated delays, postponements and all the backpedalling talk about “false evidence” and “lying witnesses” in the Uhuru Muigai Kenyatta International Criminal court trial. I don’t want to say I smell a rat but I feel like I am getting a whiff. Is the stage being set to let Kenyatta off the ICC hook? There has been feverish... (more)
- Marius_: How can we choose to hide behind claims of moral inappropriateness when - in fact - these crimes are indeed taking place on sacred African soil!? Yes! It is imperative that the ICC should, despite the influence of the 'Powers-that-be', focus on initiating proceedings on crimes within its jurisdiction taking place outside the African continent, so as to meet the dictates of fairness. But that is not to say that the ongoing cases in Africa are without their individual basis. The victims of those... (more)
- ecalmeyer: Mass African Withdrawal from the ICC: Far from Reality Introduction One hundred and twenty two countries are States Parties to the Rome Statute of the International Criminal Court (“ICC”).1 Thirty-four are in Africa, making African states the largest continental bloc of ICC signatory countries.2 Many African nations believe that the International Criminal... (more)
- John Litwin: The International Criminal Court and African Politics Introduction Given the recent vote by the Kenyan parliament to withdraw from the Rome Statute,1 it is necessary to examine the non-meritorious, political reasons that may be motivating the proposed African boycott of the International Criminal Court (ICC).2 Comprising over a quarter of all member-states,3 a withdrawal from the ICC by... (more)
- Jenevieve Discar: Potential ICC Responses to Kenya’s Proposed Withdrawal Introduction Kenya’s recent, precedent-setting vote to withdraw from the ICC highlights the critical nature of this debate; regardless of whether the ICC is actually unfairly biased towards Africa or not, the perceived bias is greatly affecting its reputation and its ability to operate effectively. Kenya’s withdrawal should... (more)
- emilygiven: Complementarity: Too Stringent a Test? While critics claim that the ICC’s focus on crimes committed in Africa is inappropriate, its defenders cite the Prosecutor’s preliminary examinations of non-African crimes as evidence to the contrary. Because the Office of the Prosecutor is evaluating situations outside Africa with an even hand, defenders argue, the Court exhibits no bias against Africa. Several preliminary examinations of non-African... (more)
- karen.kwok: Syria: a Case Study of the ICC’s Limited Jurisdiction Since its inception in 2002, all situations under investigation or prosecution have been in Africa. Critics have claimed that the ICC’s focus on Africa has been inappropriate. In particular, the ICC has been accused of having an African-bias in situation selection. However, such critiques regarding ICC’s unfair targeting of Africa... (more)
- kennygbite: The question “Is the International Criminal Court targeting Africa inappropriately?” is influenced obviously by the fact that all the cases so far being handled by the ICC fall within Africa as if crimes within the jurisdiction of the Court are not taking place in other continents. However, assuming Africans so far indicted by the Court actually committed these crimes, should the question still arise simply because their counterparts in other continents are not being investigated nor prosecuted... (more)
Comment on the Africa Question: “Is the International Criminal Court targeting Africa inappropriately?”
The Africa Bias and Official Development Assistance Disbursements: More than Just Prosecutorial Bias
I. Introduction
The International Criminal Court has increasingly come under scrutiny for operating under a so-called ‘Africa Bias.’1 Indeed, since the Rome Statute took effect, all eight cases opened by the ICC have been against African nations, of whom five countries were states parties to the Court at the time of referral.2 The obvious counter to the claim of an African bias is that the Office of the Prosecutor (“OTP”) is searching for crimes pursuant to its mandate; that Africa is repeatedly targeted is simply a logical corollary to the fact that the crimes to which the Rome Statute applies are being committed there.
However, I propose an alternative hypothesis: that there is in fact an African Bias, though it is not limited to the ICC. Instead, the prosecution of solely African nations is emblematic of an over-apportionment of resources toward the continent as a whole by the developed community; or, more specifically, the OECD. Therefore, the disproportionate focus on African nations by the OTP is not so much a concerted effort or predisposition against the African continent as it is a continuance of this more general focus on Africa by the international community.
In order to test this hypothesis, I conduct empirical and regressionary analysis of Official Development Assistance disbursement across five years to determine whether there is disproportionate attention being paid to Africa in general. Using ODA as a proxy for international attention, I show below that a disproportionate amount of resources are allocated toward Africa, even after controlling for the aid recipient’s GDP per capita. This attention might then explain the increased focus by the ICC on the continent as a whole.
II. Theoretical Framework
Given the clear tendency toward the prosecution of African defendants, I seek to explain not why these specific prosecutions are justified, but instead why they are unsurprising, and consistent with an implicitly accepted international norm. This implied norm is that there is an African bias in nearly every developmental sense, and that the law is no exception to this.
Stated more simply, the disproportionate representation by African defendants in the ICC merely mirrors the disproportionate disbursement of foreign assistance to Africa. Taking this a step further, this increased aid should then imply that the continent is less developed, generally. Therefore, in addition to increased economic development (by way of ODA), there is a need for increased legal development by way of criminal prosecution. Hence, given that African nations are the beneficiaries of a disproportionate amount of ODA, there should be the expectation that—or at least an understand as to why—these countries would be subject to increased legal liability, as well.
For this theory to hold any substantial weight, however, the theory must be grounded in some factual basis. To formulate this foundation, I begin with a discussion of the history and function of the Organisation for Economic Co-operation and Development (“OECD”), before discussing the data on development aid disbursements by this agency and then the statistical model to be used. I then present the results of this analysis as well as general conclusions.
A. The OECD and Official Development Assistance
The Organisation for Economic Co-operation and Development was founded in 1961 as an outflow of the Organisation for European Economic Co-operation (“OEEC”), which was formed to rebuild Europe’s economy after World War II.3 The OECD was subsequently established with the goal of promoting world trade and seeking to close the wealth disparity that existed between the developed and the developing countries of the world.4 One of the primary vehicles through which the OECD accomplishes this goal is through disbursement of Official Development Assistance. The Development Assistance Committee (DAC) is the branch of the the OECD devoted to disbursing the monies received to nations in need of economic development.5 The DAC is responsible for the apportionment, the granting of, and also the oversight of development assistance.6
III. Data and Model
A. Official Development Assistance (OECD)
The OECD publishes statistics tabulating the disbursements of ODA to recipient nations, disaggregated by country and by year. This data is publicly available on the OECD website, which contains numerous different statistics, including “ODA Receipts and Selected Indicators for Developing Countries and Territories.”7 This specific dataset contains statistics on ODA receipts dating back to 2007, Gross National Income (GNI) of recipient nation in 2011, the population of the recipient nation in 2011, GNI per capita in 2011, and ODA as a share of GNI in 2011. This data is then taken in conjunction with the economic indicators found in the Penn World Tables, in order to create per capita adjusted ODA disbursement figures.
B. Economic Data (Penn World Tables)
The economic data used is also publicly available international data services available through the University of Pennsylvania, via the Penn World Tables.8 This dataset includes a host of relevant variables, including GDP, population, and other economic indices for each country in the world, dating back to 1950. This data set is among the most widely used in the economic literature. Population is used to adjust ODA to a per-capita figure, and GDP is used as a control during analysis.
C. Statistical Analysis
To analyze whether or not there is an increased disbursement of foreign aid to African nations the first step is a top level analysis of that data. Thus, I begin by examining the means of the ODA disbursed by region to examine whether or not there are clear outliers. Table 1 shows that Oceana is a clear outlier. With an average ODA that exceeds all other regions by nearly fivefold, and a standard error that exceeds the majority of the other regions’ means, it is very likely an extreme set of observations, with the potential to skew regional tendencies. Therefore, Oceanic countries are discarded from subsequent analysis.9 Similarly interesting, is the size of the mean ODA disbursements to North America, as well as the variance. Yet this is likely explained by the relatively small number of countries in North America (predominantly Central American countries) which receive aid.
(in Millions of $)
After examining all of the regions to which ODA has flowed since 2007, the logical next step is a comparison of Africa, to all other regions taken together. Table A1 shows that there is a marked difference in ODA disbursements between Africa and the rest of the world.10 Though the difference is stark, a two way Student t-test does not lend enough confidence to reject the null hypothesis that there is no difference between ODA disbursed to Africa and ODA disbursed to all other nations. However, the regions within the dataset are even more granular. And, taking out North Africa, and looking only at Subsaharan Africa relative to all other nations yields an even more pronounced difference between ODA disbursements. Table 2 below shows this ‘Africa bias’ in ODA disbursements. Moreover, this difference is statistically significant. A one-sided t-test confirms that the disparity between the ODA doled out to Sub-Saharan African Countries and all other is significantly different at the 5% level.11
(in Millions of $)
This preliminary analysis makes an a priori case that there exists an ‘Africa bias’ in more than just prosecutorial discretion by the Office of the Prosecutor at the ICC. Indeed, these figures show a marked discrepancy in the focus paid to Sub-Saharan African Nations and other developing nations as measured by disbursements of aid.
1. Panel Regression Model
Given the seemingly strong evidence pointing toward an ‘Africa bias’ in more than just the ICC, a more robust model should be established to corroborate this theory. Using ODA data for the 5 year period from 2007 to 2011 combined with economic indicators, I created a 5 year panel dataset upon which regressionary analysis can be run. Beginning with this dataset, I generated a GDP per capita statistic, by adjusting GDP by population. I also generated dummy variables corresponding to each region.
A panel regression is appropriate when the data are disaggregated by year, and by region or grouping. The purpose of this form of analysis is to capture three specific trends: (i) region specific trends, (ii) time specific trends (for the whole sample) and (iii) region-time specific trends, or time trends that are exclusive to particular regions. The Ordinary Least Squares regression model for this equation looks as follows:
This simplified model projects ODA disbursements as a function of GDP per capita, as well as a regional vector, denoted Region, which consists of a host of regional dummy variables. The remaining four terms represent the trends mentioned above, and the error term. The μi term represents regional effects, ψt represents general time trends, ηit represents region-time specific trends, and εit represents the sample wide error.
Taking this model and applying it to the data, yields Table 3 below. This table chronicles various iterations of the above equation, adding and omitting regional dummies in an attempt to key into the effect of Africa on ODA disbursement.
i. Regression Analysis
Table 3 presents five separate regressions all against the baseline panel regression of GDP on the natural log of ODA disbursements.12, 13 In all five regressions, GDP per capita is not a significant predictor of ODA per capita. This of itself is troubling. It effectively implies that aid is not being doled out in accordance with need.14
However, there is a very strong positive effect attributable to Africa in each regression. Moreover, disaggregating Africa into countries North of the Sahara and South of the Sahara paints an even clearer picture of what is driving ODA disbursements. Regression (2) shows that when compared to other nations, African nations receive 63% more aid per capita even after controlling for GDP per capita. Regression (3) breaks this effect out over North and South of the Sahara; the percentage impact increases substantially for Southern African countries, to 76%. Interestingly, however, Northern African countries do not differ significantly from other nations. This effect is even more pronounced in regressions (3) and (4), where North Africa shows a significant negative effect relative to Africa generally. Similarly, Southern Africa appears to be driving the positive effect for the whole of Africa; Africa as loses its overall significance in this model, while South of the Sahara increases to 145% over the international average.
In regression 4, controlling for all regions, the trend is steeled. Relative to the North of the Sahara, Southern Africa experiences 145% more aid per capita. Also interesting is the size and significance for North America (120%). This statistic is possibly inflated by the fact that relatively few countries receive aid in this continent—after excluding the United States and Canada.
Interestingly, Northern African countries do not experience this corresponding over-emphasis. Thus, Libya might make some case for an Africa Bias. However, given the exacerbated circumstances out of which the investigation arose, as well as the fact that it was brought on by a Security Council referral, to point to the Africa bias as a factor in the prosecution of Libya would be a tenuous claim, at best.
In summary, this analysis presents convincing evidence that African nations South of the Sahara receive more Official Development Assistance than their African counterparts, and also relative to all other nations.
IV. Conclusion
Examining receipts of Official Development Assistance through OLS Panel Regression makes a cogent case that African nations, or more specifically, Subsaharan African nations receive an amount of ODA not commensurate with economic need. This disproportionate attention paid to African nations relative to others in need reflects a larger scale international emphasis on African nations. Thus, the fact that only African nations have been targeted for investigation by the ICC and the OTP is consistent with this international norm. The contention that this bias exists and should be ameliorated should therefore be accompanied by the contention that fewer official aid dollars should flow to Africa as well. However, few would support such a conclusion. Indeed, the most vocal admonishers of this prosecutorial focus are African leaders, who benefit—often personally—from increased aid receipts. Thus, the contention presented in this paper that ICC prosecutions are merely the stick accompanying the OECD’s carrot would likely dampen the volume of such criticisms.
Provided that the above analysis is taken as persuasive, the question changes from whether or not there is an Africa bias in the International Criminal Court, and instead shifts to whether or not the international community as a whole is over-apportioning resources toward Africa relative to demonstrated need. Granted, this argument and analysis miss less quantifiable determinants of international aid, including humanitarian relief, political interest, residual colonial effects, corruption, and a host of other variables. Yet, this comment casts doubt on the notion that the so-called ‘African bias’ being promulgated by the ICC is a concerted effort by the Office of the Prosecutor. Instead, I suggest that this ‘bias’ is in fact a subset of a larger, more pervasive trend. This comment is not meant to argue for decreased ODA to African countries. Instead, it is intended to show that this increased ODA is one facet of the developmental resources poured out into the African continent, and that legal development should be expected to track with economic. African leaders cannot expect ‘cafeteria’ development, whereby countries pick and choose which types of development to receive. An increased legal culpability for human rights abuses should instead be viewed as an implied pre-requisite to the receipt of international aid money.
Appendix
(in Millions of $)
Endnotes — (click the footnote reference number, or ↩ symbol, to return to location in text).
See Is Africa on Trial?, BBC News, Mar. 27, 2012, available online; Richard Lough, African Union Accuses ICC Prosecutor of Bias, Reuters, Jan. 30, 2011, available online. ↩
These countries, ordered by date of investigation are: Democratic Republic of Congo, Uganda, Central African Republic, Sudan (Darfur) [U.N. Security Council Referral], Kenya [proprio motu referral], Libya [U.N. Security Council Referral], Republic of Côte d’Ivoire, and the Republic of Mali. ↩
Organisation for Economic Development and Co-operation: Development Assistance Committee Mandate 2011-2015 (2010), available online. ↩
Id. ↩
About the OECD (Mar. 6, 2013), available online. ↩
Id. ↩
See Statistics on Resource Flows to Developing Countries, OECD, Table 25, available online . ↩
PWT 7.1 Alan Heston, Robert Summers & Bettina Aten, Penn World Table Version 7.1, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, Nov 2012 available online. ↩
Some possible explanations for the extreme values observed for Oceanic countries are: a lack of population density (as countries are predominantly islands) and also residual increased aid flows from the tsunami that struck the region in 2004. ↩
See Appendix, Table A1. Note, Oceana is omitted from subsequent analysis for the reasons discussed above. ↩
A one-sided Student t-test of means yields a p-value of 0.0428. ↩
Note, Table A2 displays this same regression. ↩
The natural log transformation corresponds to a percentage change in ODA disbursements per capita as a result of one unit changes in dependent variables. ↩
Insofar as we take GDP per capita to be an adequate proxy for need. ↩