Advocacy, Colombo, Peace and Conflict

The Peace Confidence Index Survey: What the people think about peace, war and talks

An article sent to Groundviews by Rajan Philips, that is slightly revised from an earlier version published in the media in Sri Lanka recently.

Newspaper reports have given prominence to the latest Peace Confidence Index (PCI) survey carried out in February 2007, by the Centre for Policy Alternatives (CPA). The media reports have highlighted that 59% of Sinhalese agree with the proposition that “the government should expand its military action including all out war in order to weaken the LTTE.” There is also an apparent weakening of support among the Sinhalese for the ‘peace process’ evidenced by the drop in support for it from 57% in the PCI survey of November 2006 to 46% in the February survey. During the same interval, Sinhalese support for the alternative to the peace process – government defeating the LTTE – went up from 26% to 35%.

At first cut it would seem obvious to conclude that the attitude among the Sinhalese is hardening in favour of a military solution to the Tamil question. But a closer look of the PCI survey findings, posted on the CPA website, would indicate a more nuanced and ambiguous mood among the Sinhalese as well as among the other two survey target groups – the Muslims and the Up-Country Tamils. It should be noted that the support for expanding military action was already 55% in November 2006 and the increase to 59% in February 2007 should not be over-dramatized. At the same time, the support for the peace process remains substantial among the Sinhalese and consistently high among the Muslims and the Up-Country Tamils.

Since 2001 the Centre for Policy Alternatives has been periodically tracking the public mood on the ethnic conflict and the peace process, calling it the Peace Confidence Index (PCI). The latest measurements of this index are based on a survey conducted in February 2007 among the Sinhala, Muslim and Up-Country Tamil communities living outside the Northern and Eastern Provinces. The previous survey was undertaken in November 2006. The people in the two Provinces and the Sri Lankan Tamil community as a whole were not included in the survey. The qualification is that the results are not a national estimation of opinion.

Smelling the public mood
In many Western countries, public opinion polls have come to replace the good old political hunch that successful political leaders were known to possess. Philip Gunawardena, it used to be said, had the ability to “smell the mass mood months ahead.” Later day Sri Lankan political leaders have come to rely more on astrology than on opinion polls, which are still in their infancy in Sri Lanka. Although the CPA has to be complimented for its pioneering initiative in undertaking polls, the implications that a rudimentary exercise could have on so crucial an issue as war and peace should also be kept in mind.

CPA’s survey method, or model, considers ethnicity as the “sole factor” influencing individual opinions in the population. Thus, an individual’s opinion on the ethnic conflict and the peace process is assumed to depend on whether that individual is a Sinhalese, Muslim, or Up-Country Tamil. In reality it is not so; not all Sinhalese, Tamils and Muslims think alike on most matters and despite the heightened ethnic consciousness in the country there are significant differences especially among the Sinhalese and the Tamils as to how they view the ethnic conflict and how it should be resolved. Not all Tamils support the method and the madness of the LTTE, and not all Sinhalese support a military solution to the Tamil question.

The PCI survey acknowledges that while there are other factors influencing an individual’s opinion the survey model accommodates ethnicity as the “sole factor” because it is “the most important and influential.” I would hope that treating ethnicity as the sole factor has more to do with survey logistics and resources than with theoretical inflexibility. But having undertaken the PCI survey for six years, the CPA should now take it to a higher level and include in its survey model non-ethnic factors such as age, gender, education, marital status, type of employment, household resources etc.

Such factors help explain differences between what individuals belonging to the same ethnic group think about the ethnic conflict and peace. They could also be used to identify which sections of an ethnic group are more intransigent and who tend to be more accommodating. Above all, these factors are also useful to identify which sections in different ethnic groups share similar opinions in regard to conflict and resolution. Such disaggregate information is invaluable both for theoretical understanding, policy analysis and, more important, political decision making. In particular, women in all communities tend to be more supportive of peace than war, and their views should be emphatically brought into political decision making even though they are underrepresented in the political domain despite making up one half of every community.

Along with non-ethnic factors the survey method should also take into account the district variations of the Peace Confidence Index. The number of people surveyed (survey sample) is distributed among the 17 districts outside the Northeast Provinces in proportion to the district population, but the analysis does not indicate how the answers to the survey questions vary across the districts, or the larger units of Provinces. For example, is the support for military action higher in one district than another, or is the peace process favoured more in some districts than others? As well, the representation of Muslim and Up-Country Tamil opinions is likely to be skewed when presented on a national scale as opposed to a district basis since these populations are not equally present in every district.

The political leanings of the respondents, including party affiliations and voting inclinations, are also explanatory variables that should be included in the survey. It is no secret that the support for military action and the peace process correlates with the type of political organization that people belong to, or support.

Deciphering the Peace Index
The Peace Confidence Index is structured around people’s answers to a set of questions about the importance of the peace process and the ethnic conflict, how the conflict could be ended, the credibility of Government and the LTTE as the parties to the peace process, the benefits of the Ceasefire Agreement and the monitoring of ceasefire, and the need for foreign involvement in the peace process. Most of these questions would have been formulated during the Norway-facilitated ceasefire agreement and the peace process when Ranil Wickremasinghe was Prime Minister. Remarkably, if not ironically, these questions are not concerned about the parameters of a political solution, the failure to focus on which was also the failure of Wickremasinghe as the ceasefire Prime Minister.

This year’s February survey included a hotchpotch of new questions, which do not appear to have been included in November 2006. These questions try to capture the mood of the people concerning the emphatic shift, under the current Rajapakse government, from ceasefire to military action, and the not so emphatic search for a political solution through constitutional reform. The new questions are about the people’s attitude towards the Rajapakse Government’s ability to wage war and manage the economy, the UNP crossovers, the All Party Representatives Committee on Constitutional Reform, New Emergency Regulations, human rights abuses, killings and abductions, commission of inquiry on abuses, corruption in the public sector, and trade union strikes.

While these are all important political questions, not all of them are relevant to measuring to the peace index. The questions on the All Party Representative Committee and its Chairman Tissa Vitharana are poorly designed and not well integrated as part of the peace index. Hopefully, future surveys would include a more integrated and reasonably comprehensive set of questions to elicit answers not only about war and peace but also about the parameters of a political solution.

The February survey begins by asking the respondents to rank the importance of five national issues, namely, tsunami recovery, peace process, ethnic conflict, law and order, and the economy. Tsunami recovery gets ranked as the least important of the five issues. This is not surprising given that the majority of the respondents live in areas that were not affected by the tsunami. This is one instance where a breakdown of the responses by district would have given a more accurate picture of people’s priorities. In any event, the survey did not include the districts in the Eastern Province that were the worst affected by the tsunami.

Among the five issues offered for ranking, the peace process and the economy emerge as the leading issues in the survey responses, each scoring more than 27% in each of the three communities. It is not clear why the respondents are asked to rank between the peace process and the ethnic conflict although all three communities rate the peace process as more important than the conflict itself. Also, given that the peace process and the economy are ranked as the most important issues, it would have been worthwhile probing the respondents if they saw any link between peace and economic development. An additional probe for a future survey could be about the link between war and development, the line that the present government seems to be pushing aggressively.

Ambiguously, it would appear, a majority of the respondents think that all out war is likely to break out but believe at the same time that the end of the conflict is also near. Asked to choose among a number of approaches to end the war, the overwhelming majority of the Muslims (88%) and Up-country Tamils (95%) picked peace talks as the preferred solution. Among the Sinhalese, 46% (compared to 57% in November) opted for peace talks as the preferred approach to end the war, while 11% supported peace talks combined with military offensive, and 35% (26% in November) suggested that the government must defeat the LTTE to end the war. There is also a curious bunch of 1.4% (or about 12 out of 1158) of the Sinhalese surveyed, who want the LTTE to defeat the government to end the war!

The question is then posed about the government expanding military action to weaken the LTTE, to which 59% (55% in November) of the Sinhalese responded favourably, while 23% expressed opposition and an unusually high 17% (only 6% were not sure about how to end the war in the previous question) were not sure if this approach is a good one or not. The Muslims are evenly divided, with one half of the respondents agreeing with the approach and the other half disagreeing with it. Most (88%) of the Up-Country Tamils are opposed to expanding military action against the LTTE.

As if to countervail this apparent shift towards militarism, a majority of the respondents to a separate question about foreign involvement – Sinhalese (46%), Muslims (76%) and Up-country Tamils (76%) – have expressed their support for foreign involvement to help achieve peace in Sri Lanka. Asked as to which country is most suited to play the mediatory role, 45% of the Sinhalese could not name any country, while 21% named India and 16% named USA. Normal people have more commonsense than some pundits who have advised that the Sri Lankan government should seek the involvement of China or Venezuela!

When asked specifically if India could play a positive role in resolving the ethnic crisis, a majority of the Sinhalese (45%), Muslims (65%) and Up-Country Tamils (81%) respondents answered in the affirmative. On the other hand, there is no love lost among the Sinhalese (57%) for Norway, while other communities (Muslims 57% and Up-Country Tamils 55%) are kinder to Norway.

Deep dislike of the LTTE
In my view, the increasing rate of support for expanding military action against the LTTE is a reflection of the deep dislike of the LTTE among the Sinhalese and the Muslims, and the influence that government policy and associated propaganda has on public opinion. The dislike for the LTTE surfaces prominently in the answers to questions about the LTTE’s commitment and capability for achieving peace through talks. Identical questions – fanatical sovereignists will not tolerate this – are also asked about the commitment and the capability of the government to achieve peace through talks. Not surprisingly, the government gets very favourable responses among the Sinhalese, while the LTTE receives a highly unfavourable assessment. Interestingly, more Mulsims doubt the commitment, but not the capability, of the LTTE to achieve peace through talks. This assessment about the LTTE is reversed among the Up-Country Tamils.

The second reason for supporting military action is because of the government’s expansion of its military action over the last year as both the LTTE and the government opted to disregard the ceasefire and fight each other. The people’s support for this government at this time is evident from the answers to the new questions about the government and recent political developments. The survey responses indicate that there is greater confidence in Mahinda Rajapakse’s ability to lead the war than manage the economy; there are no tears shed for the UNP over the loss of some of its elephants; and there is apparent belief that corruption in the public sector has gone down after this government came to power. Mahinda Rajapakse’s personal popularity among the Sinhalese and the military’s battlefield successes against the LTTE have been mutually reinforcing and are reflected in the survey opinion supporting the government’s expansion of military action to weaken the LTTE.

It is not unusual for public opinion to be led by government action especially on the question of war. Public opinion would be favourable towards war especially during its early stages when things are going well in battles and the war propaganda is also particularly effective. Public opinion will swing against the war and its leaders when it runs out of ‘shock and awe’, and the cost of war begins to dispel the illusion of its benefits. The mood swings of the American people over the Bush Administration’s war in Iraq are a good example. Here at home, as President Rajapakse himself has said, there would be praise all around as each battle brings victory but only blame and criticism when the war turns uglier and starts to incur more loses than success.