Document - Colombia: Amnesty International response to Andrés Ballesteros, Jorge A. Restrepo, Michael Spagat, Juan F. Vargas, The Work of Amnesty International and Human Rights Watch

Colombia:

 

Amnesty International response to Andrés Ballesteros, Jorge A. Restrepo, Michael Spagat, Juan F. Vargas, The Work of Amnesty International and Human Rights Watch: Evidence from Colombia, CERAC, Colombia, February 2007

 

While the paper by Andrés Ballesteros et alasks some valid questions about the use of Amnesty International (AI) figures by CIRI and PTS, we are at a loss to understand clearly why the paper chooses to focus its main criticisms on AI. AI has never sought to construct a hard data base nor has it sought to make any detailed quantitative claims. This is not AI’s mandate and never has been. If the focus of AI’s work were based on statistical analysis, rather than on describing and trying to affect a complex and evolving situation using complex and evolving sources of information and analytical tools, then such criticism might indeed be perfectly valid. However, the focus of AI’s work, and as such the basis for its strength, as well as of its reputation for accuracy and impartiality, is based on the organization’s qualitative (primarily case-based) rather than quantitative work, as the paper rightfully acknowledges.

 

This is not to say that AI does not on occasions illustrate its analysis with the use of statistics. But these on the whole are general quantifiers. AI tries on the whole to avoid time-series or cross-country data sets. This is because of the difficulty of finding methodologies and data which would make comparisons over time and across all countries in any way meaningful. One of the benefits of AI’s long and intense engagement with the human rights situation in a country like Colombia has precisely been to contribute to changing the quantity and quality of the information available. Local human rights NGOs have grown in number and become stronger and the number of international actors on the ground increased significantly. Over time, this has led to a greater availability of information, while changes in technology have significantly affected the speed and means of information gathering and dissemination.

 

AI also does not accept, as the paper implies, that human rights NGOs should be expected to cover the same issues consistently over a period of many years, simply to be able to provide some kind of statistical benchmark. Both country situations and human rights organizations change. It is thus necessary for AI to change priorities in response to both the local and the international situation. If AI were to seek to cover the same information via the same sources over a long period of time it would not be doing its job as a human rights advocate.

 

We refute the accusation that AI’s work on Colombia is anti-government. Most importantly, it must be made clear that prior to 1992 AI was only mandated to work on state violations and therefore had no authority to document or condemn breaches of IHL committed by non-state actors such as the guerrilla in Colombia. This is why, as your report states, "in the early years of our sample it is true that AI focused exclusively on the government". Once the mandate was changed AI has documented such abuses of IHL by guerrilla groups with as much vigour as abuses committed by state actors. This is acknowledged in the article which states that the guerrilla accounted for 24% of all cases highlighted by AI (although this percentage is probably much higher given that the data from the early years will have distorted the average). But it must be stressed that when AI began using an IHL analysis, the primary aim was not to count abuses by armed groups but to look more specifically at the large numbers of civilians affected by conflict who suffered abuses that did not fall into a traditional human rights paradigm. It was a victim-driven innovation. IHL standards, methodologies, policies and practices (particularly regarding approaches to armed groups) had to be developed over years.

 

The paper acknowledges that AI is primarily a government watchdog, and most of its campaigning and lobbying work is directed at governments. This does not mean that AI is asserting that governments are the main perpetrators of human rights abuses, simply that for strategic reasons AI’s focus has to be on changing government policy. They are the signatories to international human rights treaties and should hold a monopoly of power. As such, they will continue to be the main, but not exclusive, focus of AI’s work.

 

As regards allegations about AI’s "lack of sourcing". Many of its case studies and figures come from material provided by contacts in the field. Their safety would be put at risk were AI to identify them publicly. Information is also often confirmed from a variety of sources rather than a single source. In the case of Colombia, AI tends not to use official statistics, with the exception of kidnapping figures, primarily because official statistics have proven deficient in including human rights/IHL-related categories. For example, as the 2006 report on Colombia of the UN High Commissioner for Human Rights stated, "[t]he Human Rights Observatory of the Office of the Vice President does not include extrajudicial executions or arbitrary detentions, nor certain categories of breaches of international humanitarian law".

 

Another concern regarding statistics in relation to Colombia is the historic under-reporting of human rights violations. AI does use statistics when these are relevant to a particular Colombia report or annual report entry, or when it believes that there has been a significant and indisputable rise or drop, as has been the case with the rate of kidnappings in recent years. When AI does cite statistics it often refers to " more than" or "no fewer than" or "numerous" "or "repeated", or similar terminology, since AI is seeking to give a flavour of the type and degree of human rights violations/abuses in the country, not to be quantitatively precise.

 

Over the last few years AI has avoided statements which classify the human rights situation as either improving or deteriorating, since it believes the situation is more nuanced. AI has thus sought to be more specific, highlighting in its recent annual reports those areas where it believes there has been an improvement (such as security in urban areas, kidnapping, overall killings) and where there has been a deterioration (reports of extrajudicial executions and, until very recently, in terms of new cases of those forcibly displaced by the conflict).

 

In terms of defining the categories AI reports on, the organization believes that whether, for example, AI uses the term "conflict-related killings" or "politically-related killings" our readers understand the point AI is trying to make (all those killed as a result of the conflict, whether civilian or military), as when we refer to "civilians killed out of combat". However, the fact that in early reports AI only referred to EJEs (which in legal terms can only be committed by state actors) reflects the fact that AI’s mandate only covered state actors. The phrase "political killings" refers to all conflict-related killings and was incorporated to reflect the new mandate, which allowed AI to work on issues relating to non-state actors. If AI figures aim to include only civilians killed as a result of the conflict then we try to make this clear in the text (for example by referring to "non-combat politically motivated killings").

 

In terms of human rights versus IHL, AI’s position, made clear in its recent publications on Colombia, is that non-state actors "abuse" human rights and breach IHL, while state actors "violate" human rights (because only states are party to human rights treaties) as well as breach IHL. For short hand AI sometimes refers to state and non-state actors "abusing human rights" or refers to "human rights abuses and violations". Although such language may appear legalistic, and may not mean much to the lay reader, it does correspond to legal definitions in international human rights law which is the basis of AI’s work.

 

The claim in the paper that AI is seeking to downplay human rights abuses, such as killings, by the guerrilla and overplay those of the security forces and paramilitaries is inaccurate. The article even acknowledges that accusations that AI rarely criticizes the guerrilla are unfounded. In recent annual reports AI has also broken down the figures for kidnapping according to perpetrator and has clearly identified the guerrilla as the main culprits. AI has also sought to be consistent with its use of statistics for IDPs and to differentiate between accumulated numbers and new IDPs. The accumulated number, of course, can be difficult to gauge and AI has therefore often resorted to the use of phrases such as "at least" or "over".

 

AI does often band together the security forces with the paramilitaries. AI’s extensive research over the years has exposed the close links between the security forces and paramilitary groups. Successive Colombian governments have persistently sought to deny these links despite overwhelming evidence from Colombian and international NGOs, as well as international organizations such as the Inter-American Commission on Human Rights and the Office of the UN High Commissioner for Human Rights. AI rejects claims in the paper that it has failed to provide any evidence of such links. It has done so consistently over the years in numerous reports. Regardless, it is AI’s role only to expose claims of such links and it is the responsibility of Colombia’s judicial authorities to investigate them. This they have persistently failed to do. Recent revelations about the degree of paramilitary infiltration in various state institutions, including the security forces, the legislature, and the DAS appear only to have confirmed what AI and other NGOs have been saying publicly for years. In many cases, AI has found it difficult to separate security force from paramilitary responsibility in killings, since there is evidence of the involvement of both. In fact, in many cases in which the article might describe the evidence for collusion as being "soft", subsequent research and/or judicial investigations have corroborated AI’s initial claims.

Amnesty International books (65)