|Country Report Parallel Research Programme - ICRC Worldwide Consultation on the Rules of War (International Committee of the Red Cross , 1999, 38 p.)|
At the heart of the ICRC consultation on war are national public opinion surveys of people in war-torn countries. These surveys were designed to be representative and administered in the most objective and consistent way possible. But they were not administered by independent, professional research organizations. They were administered by ICRC staff and volunteers of the National Red Cross and Red Crescent Societies. Through this project, during which more than 12,000 interviews were conducted with people in 12 countries, the ICRC aimed to give people a voice in the discussion on the limits in war. The project was guided by Greenberg Research, and local research organizations helped with the research plan, interviewer training, the participant sample design, and the questionnaire.
But the ICRC and National Red Cross and Red Crescent Societies are not professional research organizations, so their participation could reasonably be expected to produce subjective results. No matter how scrupulously objective they were in conducting the surveys, people might well adjust their answers to be more in line with the presumed positions and work of the ICRC and the Red Cross or Red Crescent Society in the country. It is possible that respondents might feel awkward or embarrassed to give certain responses - perhaps belligerent ones - to a Red Cross or Red Crescent interviewer.
On the other hand, on some issues in this context, people might be more open with a Red Cross or Red Crescent interviewer, who is presumed to be disengaged from the conflict. Thus, a Red Cross bias may come from peoples reluctance to be fully honest with a professional interviewer, who lacks the Red Crosss objective standing.
To determine whether there was a Red Cross bias in these surveys, parallel surveys were conducted in three of the countries being studied: Colombia, the Philippines and Bosnia-Herzegovina. In these countries, the ICRC performed the regular consultation, including a national survey; at the same time, Greenberg Research commissioned a local partner to carry out an independent professional survey - with the same specifications, sample design and questionnaire.
The parallel research programme included three pairs of nationally representative surveys, each composed of an ICRC survey and a parallel survey. The ICRC survey in Colombia is based on 857 respondents and was conducted between 8 November and 24 December 1998. The parallel survey is based on 1,000 respondents and was conducted between 8 November and 22 November 1998 by Centro Nacional de Consultoria (CNC) based in BogotThe ICRC survey in the Philippines is based on 1,100 respondents and was carried out between 26 January and 6 February 1999. The parallel research is based on 1,000 respondents and was conducted between 26 January and 8 March 1999 by Social Weather Stations (SWS) based in Manila. Finally, the ICRC survey in Bosnia-Herzegovina is based on 1,482 respondents (approximately 500 respondents each in the Republika Srpska, and the Croat and Bosniac parts of the Federation of Bosnia-Herzegovina) and was conducted between 14 February and 5 March 1999. The parallel survey for Republika Srpska is based on 500 respondents and was conducted between 14 February and 2 March 1999 by Medium (Belgrade). The survey in the Croat and Bosniac areas of the Federation was conducted between 14 February and 1 March 1999 by PULS (Zagreb) and is based on a total of 998 respondents.
The results presented in this report are mostly based on comparisons between the answers given in the ICRC and parallel surveys from a combined database of 6,937 respondents from all six surveys. Half of these (3,439 respondents) were interviewed by Red Cross staff or volunteers and the other half (3,498 respondents) by professional interviewers. Whenever the patterns within each country are significantly different from the overall pattern, the analysis explores the differences in greater detail.1
1 Findings will be elaborated when data patterns differ consistently from the overall patterns seen across countries.
In fact, the parallel research programme in the three countries found little evidence of significant or systematic discrepancies between the results of the ICRC and parallel surveys in terms of peoples views on international humanitarian law, acceptable practices during armed conflict, the distinction between combatants and civilians and the treatment of prisoners. On the great majority of questions and topics, it seems that the ICRC consultation attained results comparable to what would be obtained by an independent and professional research organization. There are a number of areas, however, where comparison of the parallel research results shows evidence of bias. These are outlined below:
· Greater knowledge and effectiveness of the Geneva Conventions. Compared with respondents in the parallel surveys, people interviewed by the Red Cross are more inclined to say they have heard of the Geneva Conventions and are more likely to see them as effective. That difference is not evident in peoples actual knowledge of laws that would prohibit certain behaviour by combatants in war.
· Greater insistence on prosecuting war criminals. Compared with respondents in the parallel surveys, people interviewed by the Red Cross are more likely to believe that there are important rules or laws in war that, if violated, should lead to punishment. The difference, however, is not very big and very large majorities in both the ICRC and parallel surveys favour prosecuting war criminals. The difference between the two sets of studies is not evident when it comes to the conceptual foundations of war crimes.
· Greater awareness of the ICRC. Compared with the respondents in the parallel surveys, people interviewed by the Red Cross are much more conscious of the ICRC. This leads to a substantial overstatement regarding its role in the protection of civilian populations and prisoners.
· More likely to help captured enemy combatants. In the ICRC surveys, respondents are more likely than those in the parallel surveys to say they would save or help a wounded or surrendering enemy combatant. This is a projection of behaviour in a simulated situation, but nonetheless suggests respondents may be inclined to offer more benign responses about their own behaviour when speaking to a Red Cross interviewer.
· More war-related injuries and incidents. Respondents in the ICRC surveys are much more willing to say they have been humiliated in the war and marginally more willing to say they have been wounded. The differences on most questions about war, however, are insignificant; the two studies offer identical results on the emotions felt during the war and in expectations for the future.
This parallel project certainly established areas of difference, though in each case the Red Cross bias seems highly contained. Almost none of the problem areas involve attitudes towards international humanitarian law; few involve the distinction between combatants and civilians; few involve actual knowledge of the Geneva Conventions and their prohibitions or understanding of their conceptual foundations. The bias is strongest for the ICRC itself and for the role of international institutions.
While the discrepancies between the ICRC and parallel surveys are small in the combined data for all three countries, there is evidence of some bigger discrepancies at the country level. Some of these differences are due to greater chance fluctuations as a consequence of smaller sample sizes, but some of them are real.
The idiosyncrasies at the country level are mainly brought about by the specific ways in which the people in each of these contexts relate to the ICRC/Red Cross. For example, in Bosnia-Herzegovina, which is marked by the pervasiveness of the war, the ICRC and its mission are well known. In this setting, there is greater awareness of the principles of international humanitarian law and of international organizations, including the ICRC. In Bosnia-Herzegovina, therefore, this may produce a greater bias on issues related to war crimes and international humanitarian law. On the other hand, in the Philippines, where the armed conflict is more remote, the ICRC and international humanitarian law are much less known. Filipinos questioned by Red Cross interviewers therefore seem under little pressure to provide responses favourable to the ICRC or its concerns. Finally, in Colombia, where the conflict is characterized more by random, episodic violence, the parallel research identified the smallest discrepancies between the responses given by respondents in the ICRC and parallel surveys.
The absence of consistent, cross-country discrepancies between the ICRC research and the parallel research indicates that there is little need for the great majority of questions asked to adjust the country findings to minimize a Red Cross bias. It is important, however, to keep in mind the problem areas: greater familiarity with the Geneva Conventions, greater insistence on the prosecution of war crimes, awareness of the ICRC, projected benign treatment of prisoners, and reported war incidents and injuries.