It is not surprising that the Chatham House Report authored by Robert W. Bradnock, Kashmir: Paths to Peace is receiving serious attention of scholars and policy makers in the western world, given that the report has been propagated as the first ever poll conducted in the two parts of Kashmir. Based on the assumption that “Kashmiri” opinion matters, the poll tries to suggest alternate scenarios for the resolution of the conflict based on attitudes of both sides of the LoC to these scenarios. The results of the poll have been rather cryptically presented in the form of a 31-page report. However, an analysis of the Report and the 86 data tables available on the website of the polling agency Ipsos MORI (http://www.ipsos-mori.com/), reveals serious problems.
First, the findings, as presented in the report, are misleading. Moreover, the report does not mention several significant results that data tables showcase. Thus, in the section related to “Perceptions of Key Problems”, the report identifies unemployment, corruption, poor economic development, human rights abuses and the political conflict over Kashmir as the top five concerns of the sample surveyed (Table 3). The problem with the results as presented is that while these issues get the top scores at the aggregate level for both sides of the LoC as well as at the disaggregated level in AJK and J&K, at the disaggregated level of the districts, among the 32 issues on which people were polled, the ranking of issues is at variance with the aggregate results, and is more reflective of local concerns.
Several interesting findings based on the responses to this question do not find mention in the report. For instance, it fails to bring out that terrorism is not perceived as a major problem in the entire state. In fact, excepting for one respondent in Srinagar, no respondent has even mentioned terrorism as a major problem confronting J&K. Also, on the issue of human rights, in five of the 11 districts of the State, namely Rajauri (5 %), Punch (2 %), Kathua (8 %), Udhampur (0 %), and Jammu (3 %), human rights abuses are not among the top five concerns. The low percentages in Punch (mistakenly marked as 92 % instead of 2 % in Figure 5 of the report) and Rajauri, which were once hotbeds of terrorism, are especially significant, since even till date the area is witness to counterinsurgency operations by the armed forces. What is surprising, however, is that the issue of human rights is perceived as important in Leh (18 %) and Kargil (42%). Another noteworthy fact is that inability to move across the LoC is a serious issue for the people of Leh (21 %) and Kargil (88 %). This is not surprising considering that the cross-LoC routes in Jammu (Punch-Rawalkote) and Kashmir Valley (Srinagar-Muzaffarabad) cater mainly to the local population. More worrying is the fact that despite the opening of the Punch-Rawalkote route, in Punch (43 %) and Rajauri (74%) respondents feel that they cannot move across the LoC. Such a feeling also persists in Srinagar (12 %).
While the report also depicts 55 per cent of people feeling safe as a result of the talks despite the low levels of awareness on details of the talks between India and Pakistan, what the report misses out on is that there is an equally high number of people polled who felt that the talks have made no difference or those who felt less safer as a result of the talks. On an average, approximately 23 per cent people in the Valley feel safer as a result of the talks, whereas in Jammu this percentage is considerably higher at 64 per cent. In the districts of the Valley, where the percentage of those feeling safe is less, a greater proportion feel that talks make no real difference. Also to be noted is that, in Baramulla the number of people feeling less safe is much higher than those feeling safe. Anantnag and Baramulla have nearly a fourth of those polled feeling less safe due to the talks. It is important to note here that excepting in Kathua and Jammu, the percentage feeling less safe is significantly lower than those feeling no real difference. It is also to be noted that whereas in Kargil a greater number feel safe due to talks, in Leh a greater number feel that the talks make no real difference, nearly to the same levels as those in the Valley districts. It also misses out in highlighting the high degree of correspondence between those who personally feel safe as a result of talks and those who feel talks will bring peace.
While the Report concludes that 52 per cent felt that state elections improved the chances of peace and 55 per cent felt that national elections did so, it does not mention that in the Kashmir region over 50 per cent of the people polled felt that neither the national elections nor the state elections made any real difference to chances of a permanent peace; between 5 and 30 per cent people felt elections had some positive impact on the chances of permanent peace. In Udhampur, Kathua and Kargil, a majority of the people felt that elections greatly improved the chances of permanent peace. In Jammu, Rajauri, Punch and Leh, a majority felt that elections had a slightly positive impact on chances for permanent peace. Out of a total of 22 results in the better (much better + slightly better) category, 3 results convey equal hope/expectations from both elections, 10 results convey more hope form the state elections and 9 convey more expectations from national elections.
Finally, the Report selectively ignores some data tables, which, if read in conjunction with other responses, could yield interesting results. These are on:
who should have the right to vote on the future of Kashmir (Table 46 of the Data tables)
whether foreign countries or the United Nations should play a role in resolving the Kashmir dispute (Tables 48 and 50 of the Data tables)
whether people would be better off politically and economically being Indian citizens, Pakistani citizens or citizens of an united Independent Kashmir.
whether the people perceive themselves as Kashmiris or not.
Second, there are also serious issues with the drafting of the questions, which not only misrepresent information to the respondents, but also misrepresent the results. For instance, the section on “Attitudes to the Political Process” refers to the India-Pakistan talks “to resolve the dispute”. In the data tables, however, the questions posed are about the year in which India and Pakistan started talks “over the future of Kashmir”. The questionnaire then offers 1947, 1951, 1978, 1985 and 2003 as alternate dates. It needs to be clarified, therefore, that the composite dialogue between India and Pakistan is “to resolve all bilateral issues”, not “the dispute over Kashmir”. These issues are: Peace and Security, Kashmir, Wullar Barrage, friendly exchanges, Siachen glacier, Sir Creek, terrorism and drug trafficking and economic and commercial cooperation. While the report goes on to state that these talks started in 2003, the Composite Dialogue between the two countries actually began in 1998, got stalled during the Kargil War (1999) and the Parliament attack (2001), and was only resumed in 2004. Given the fractured trajectory of the talks and the manner in which the question was posed, it is not surprising that on an average 40 per cent of the people of Jammu and Kashmir knew of the talks and not the exact dates, another 23 percent did not know when the talks started and 17 per cent felt that the talks had started in 1947/51. Similarly, without explicating on the meaning of the phrase “permanent peace”, it seeks the opinion of people regarding whether the National elections (2009) and the State Assembly elections (2008) have improved chances of “permanent peace” in Kashmir.
What is perhaps the most crucial section of the report on views of people on “Options for a Political Future” is also the most problematic part of the report. Among the options listed are:
independence for Kashmir on both sides of LoC
Kashmir on both sides of LoC to join India
Kashmir on both sides of LoC to join Pakistan
LoC to be made a permanent international border
joint sovereignty of India and Pakistan over the whole of Kashmir for foreign affairs and the whole of Kashmir having autonomy over internal affairs (a)
joint sovereignty of India and Pakistan over the whole of Kashmir for foreign affairs, with local (eg. at state level) control over internal affairs (b)
No change, the status quo to be maintained
According to the author, on an average, 43 per cent of the respondents have voted for an independent future and merely 28 per cent for a future with India. Such a conclusion is too simplistic and therefore not credible. The average percentage of those opting for an independent future is higher because of the very high numbers opting for this option in the four districts of the Valley. The average number therefore errs on the high side, a point the author concedes much later in the conclusion of the Report. Moreover, the option regarding LoC as the permanent border is not an independent option. It implies that everyone who wants this to be the future, also wants to be with India.
Independence
a. Join India
Join Pakistan
b. LoC to be Per
a + b
Srinagar
82
8
6
0
8
Jammu
1
47
*
39
86
Anantnag
74
22
2
0
22
Udhampur
0
73
0
14
87
Baramula
95
2
2
0
2
Kathua
0
63
0
3
69
Leh
30
67
0
2
69
Punch
0
6
0
94
100
Rajuri
0
0
0
100
100
Badgam
75
10
7
1
11
Kargil
20
80
0
0
80
So, the new average of those wanting to join India is approximately 52 per cent, 24 percentage points higher than the earlier average. The error on the higher side is much less in this case because the numbers opting for Independence in all districts except those in the Valley are negligible, whereas there is a cognizable percentage in the Valley districts, except in Baramula, who see a future in joining India.
Another instance of confused drafting of questions is on the status of the Line of Control and cross-border contacts. The questions beg the clarification that trade and travel across the LoC were envisaged as key elements of the Confidence Building Measures (CBMs) between India and Pakistan. The Srinagar-Muzaffarabad bus service commenced on 7 April 2005 and the Punch-Rawalkote route was opened for travel on 20 June 2006. Cross-LoC trade on the same routes commenced from 21 October 2008. In accordance with the permit system currently in place, people crossing the LoC from either side are permitted to travel on an Entry Permit. Only those who have relatives residing on both sides of the LoC are issued permits. People visiting are allowed to stay for 28 days, extendable to 15 more days. During the visit, they are not permitted to travel outside the state of Jammu and Kashmir. It is in this context that one needs to evaluate the response to the questions on the status of the LoC and the need for permits or passports to cross the LoC. The options given are not mutually exclusive, whereas the responses are recorded exclusively for each answer. Thus, those wanting free movement of people and goods cannot be completely separated from those wanting only free movement of goods.
Similarly, questions pertaining to “Attitudes to Security” were not really posed as such. The questions on removal of mines and weapons from both sides of the LoC were posed as - whether or not people support or oppose their removal. The other questions on withdrawal of Indian and Pakistani security forces, end of militant activity and inclusion of All Kashmiri political opinion in talks were posed as - “options for probable solutions to the conflict”. While these questions are security-related, it would be misleading to view them as “Attitudes to Security”, as much as it would be wrong to consider responses to other questions as not being reflective of attitudes to security. In addition, it is surprising to note that the option “include all Kashmiri political opinion in talks” is not part of the data tables on “options for probable solutions to the conflict” (Data tables 34, 36, 38 and 40). In fact, there is no direct question to this effect in the entire data tables. There is a question on “who should have the right to vote on the future of Kashmir” (Data table No. 46), but its results do not tally with the results mentioned in Figure 28 of the report.1
Finally, the report lacks methodological clarity. One does not get a clear sense of the range of questions asked and the alternate options provided until one delves into the data tables provided on the website of Ipsos MORI. For instance, while the top five results regarding the concerns of the people are highlighted in the report, it is not stated that these were among the 32 issues that were included in the poll. It is also not clear whether answers received were enumerated later, or if the alternatives were provided by the polling agents at the very outset. Also, while the report mentions that some districts like Doda, Pulwara (sic) and Kupwara are excluded from the survey, it does not explain why they were excluded, especially since they existed as districts at the time of the census (2001) as well as at the time of the survey (2009). Then, while it is claimed that face-to-face interviews were conducted with 2374 individuals, selected by random procedure based on quotas set by gender, age, district and religion to reflect the population profile (as per the 2001 census), it is not clear why the unweighted size of the sample is not proportional to the size of the population of the district. For instance, for Kargil, which has a population of 119, 307, the sample size is 501, whereas for Jammu, which has a population of 1,588, 772, the sample size is only 457.
The result of all this is a report that lends itself to all kinds of interpretations. It yields results that are predictable in some instances and provides no explanation for the surprise results. It does not even attempt to correlate responses to two questions in the same section, which could provide supporting or contradictory evidence for a claim being made. For instance, a comparative analysis of response to the above questions on elections and utility of militant violence reveals that there is much more faith in elections leading to permanent peace than violence leading to a solution. The people in the Valley are largely neutral on either having any impact, reflecting a sense of hopelessness. Nor does it explain why the people in Rajauri and Punch, who feel they are Kashmiri, do not vote along the lines of the districts in the Kashmir Valley. And it certainly does not provide the “paths to peace”.
1. Correction: This passage has been rephrased subsequently.
“Kashmir: Paths to Peace”: A Misleading Report
More from the author
It is not surprising that the Chatham House Report authored by Robert W. Bradnock, Kashmir: Paths to Peace is receiving serious attention of scholars and policy makers in the western world, given that the report has been propagated as the first ever poll conducted in the two parts of Kashmir. Based on the assumption that “Kashmiri” opinion matters, the poll tries to suggest alternate scenarios for the resolution of the conflict based on attitudes of both sides of the LoC to these scenarios. The results of the poll have been rather cryptically presented in the form of a 31-page report. However, an analysis of the Report and the 86 data tables available on the website of the polling agency Ipsos MORI (http://www.ipsos-mori.com/), reveals serious problems.
First, the findings, as presented in the report, are misleading. Moreover, the report does not mention several significant results that data tables showcase. Thus, in the section related to “Perceptions of Key Problems”, the report identifies unemployment, corruption, poor economic development, human rights abuses and the political conflict over Kashmir as the top five concerns of the sample surveyed (Table 3). The problem with the results as presented is that while these issues get the top scores at the aggregate level for both sides of the LoC as well as at the disaggregated level in AJK and J&K, at the disaggregated level of the districts, among the 32 issues on which people were polled, the ranking of issues is at variance with the aggregate results, and is more reflective of local concerns.
Several interesting findings based on the responses to this question do not find mention in the report. For instance, it fails to bring out that terrorism is not perceived as a major problem in the entire state. In fact, excepting for one respondent in Srinagar, no respondent has even mentioned terrorism as a major problem confronting J&K. Also, on the issue of human rights, in five of the 11 districts of the State, namely Rajauri (5 %), Punch (2 %), Kathua (8 %), Udhampur (0 %), and Jammu (3 %), human rights abuses are not among the top five concerns. The low percentages in Punch (mistakenly marked as 92 % instead of 2 % in Figure 5 of the report) and Rajauri, which were once hotbeds of terrorism, are especially significant, since even till date the area is witness to counterinsurgency operations by the armed forces. What is surprising, however, is that the issue of human rights is perceived as important in Leh (18 %) and Kargil (42%). Another noteworthy fact is that inability to move across the LoC is a serious issue for the people of Leh (21 %) and Kargil (88 %). This is not surprising considering that the cross-LoC routes in Jammu (Punch-Rawalkote) and Kashmir Valley (Srinagar-Muzaffarabad) cater mainly to the local population. More worrying is the fact that despite the opening of the Punch-Rawalkote route, in Punch (43 %) and Rajauri (74%) respondents feel that they cannot move across the LoC. Such a feeling also persists in Srinagar (12 %).
While the report also depicts 55 per cent of people feeling safe as a result of the talks despite the low levels of awareness on details of the talks between India and Pakistan, what the report misses out on is that there is an equally high number of people polled who felt that the talks have made no difference or those who felt less safer as a result of the talks. On an average, approximately 23 per cent people in the Valley feel safer as a result of the talks, whereas in Jammu this percentage is considerably higher at 64 per cent. In the districts of the Valley, where the percentage of those feeling safe is less, a greater proportion feel that talks make no real difference. Also to be noted is that, in Baramulla the number of people feeling less safe is much higher than those feeling safe. Anantnag and Baramulla have nearly a fourth of those polled feeling less safe due to the talks. It is important to note here that excepting in Kathua and Jammu, the percentage feeling less safe is significantly lower than those feeling no real difference. It is also to be noted that whereas in Kargil a greater number feel safe due to talks, in Leh a greater number feel that the talks make no real difference, nearly to the same levels as those in the Valley districts. It also misses out in highlighting the high degree of correspondence between those who personally feel safe as a result of talks and those who feel talks will bring peace.
While the Report concludes that 52 per cent felt that state elections improved the chances of peace and 55 per cent felt that national elections did so, it does not mention that in the Kashmir region over 50 per cent of the people polled felt that neither the national elections nor the state elections made any real difference to chances of a permanent peace; between 5 and 30 per cent people felt elections had some positive impact on the chances of permanent peace. In Udhampur, Kathua and Kargil, a majority of the people felt that elections greatly improved the chances of permanent peace. In Jammu, Rajauri, Punch and Leh, a majority felt that elections had a slightly positive impact on chances for permanent peace. Out of a total of 22 results in the better (much better + slightly better) category, 3 results convey equal hope/expectations from both elections, 10 results convey more hope form the state elections and 9 convey more expectations from national elections.
Finally, the Report selectively ignores some data tables, which, if read in conjunction with other responses, could yield interesting results. These are on:
Second, there are also serious issues with the drafting of the questions, which not only misrepresent information to the respondents, but also misrepresent the results. For instance, the section on “Attitudes to the Political Process” refers to the India-Pakistan talks “to resolve the dispute”. In the data tables, however, the questions posed are about the year in which India and Pakistan started talks “over the future of Kashmir”. The questionnaire then offers 1947, 1951, 1978, 1985 and 2003 as alternate dates. It needs to be clarified, therefore, that the composite dialogue between India and Pakistan is “to resolve all bilateral issues”, not “the dispute over Kashmir”. These issues are: Peace and Security, Kashmir, Wullar Barrage, friendly exchanges, Siachen glacier, Sir Creek, terrorism and drug trafficking and economic and commercial cooperation. While the report goes on to state that these talks started in 2003, the Composite Dialogue between the two countries actually began in 1998, got stalled during the Kargil War (1999) and the Parliament attack (2001), and was only resumed in 2004. Given the fractured trajectory of the talks and the manner in which the question was posed, it is not surprising that on an average 40 per cent of the people of Jammu and Kashmir knew of the talks and not the exact dates, another 23 percent did not know when the talks started and 17 per cent felt that the talks had started in 1947/51. Similarly, without explicating on the meaning of the phrase “permanent peace”, it seeks the opinion of people regarding whether the National elections (2009) and the State Assembly elections (2008) have improved chances of “permanent peace” in Kashmir.
What is perhaps the most crucial section of the report on views of people on “Options for a Political Future” is also the most problematic part of the report. Among the options listed are:
According to the author, on an average, 43 per cent of the respondents have voted for an independent future and merely 28 per cent for a future with India. Such a conclusion is too simplistic and therefore not credible. The average percentage of those opting for an independent future is higher because of the very high numbers opting for this option in the four districts of the Valley. The average number therefore errs on the high side, a point the author concedes much later in the conclusion of the Report. Moreover, the option regarding LoC as the permanent border is not an independent option. It implies that everyone who wants this to be the future, also wants to be with India.
So, the new average of those wanting to join India is approximately 52 per cent, 24 percentage points higher than the earlier average. The error on the higher side is much less in this case because the numbers opting for Independence in all districts except those in the Valley are negligible, whereas there is a cognizable percentage in the Valley districts, except in Baramula, who see a future in joining India.
Another instance of confused drafting of questions is on the status of the Line of Control and cross-border contacts. The questions beg the clarification that trade and travel across the LoC were envisaged as key elements of the Confidence Building Measures (CBMs) between India and Pakistan. The Srinagar-Muzaffarabad bus service commenced on 7 April 2005 and the Punch-Rawalkote route was opened for travel on 20 June 2006. Cross-LoC trade on the same routes commenced from 21 October 2008. In accordance with the permit system currently in place, people crossing the LoC from either side are permitted to travel on an Entry Permit. Only those who have relatives residing on both sides of the LoC are issued permits. People visiting are allowed to stay for 28 days, extendable to 15 more days. During the visit, they are not permitted to travel outside the state of Jammu and Kashmir. It is in this context that one needs to evaluate the response to the questions on the status of the LoC and the need for permits or passports to cross the LoC. The options given are not mutually exclusive, whereas the responses are recorded exclusively for each answer. Thus, those wanting free movement of people and goods cannot be completely separated from those wanting only free movement of goods.
Similarly, questions pertaining to “Attitudes to Security” were not really posed as such. The questions on removal of mines and weapons from both sides of the LoC were posed as - whether or not people support or oppose their removal. The other questions on withdrawal of Indian and Pakistani security forces, end of militant activity and inclusion of All Kashmiri political opinion in talks were posed as - “options for probable solutions to the conflict”. While these questions are security-related, it would be misleading to view them as “Attitudes to Security”, as much as it would be wrong to consider responses to other questions as not being reflective of attitudes to security. In addition, it is surprising to note that the option “include all Kashmiri political opinion in talks” is not part of the data tables on “options for probable solutions to the conflict” (Data tables 34, 36, 38 and 40). In fact, there is no direct question to this effect in the entire data tables. There is a question on “who should have the right to vote on the future of Kashmir” (Data table No. 46), but its results do not tally with the results mentioned in Figure 28 of the report.1
Finally, the report lacks methodological clarity. One does not get a clear sense of the range of questions asked and the alternate options provided until one delves into the data tables provided on the website of Ipsos MORI. For instance, while the top five results regarding the concerns of the people are highlighted in the report, it is not stated that these were among the 32 issues that were included in the poll. It is also not clear whether answers received were enumerated later, or if the alternatives were provided by the polling agents at the very outset. Also, while the report mentions that some districts like Doda, Pulwara (sic) and Kupwara are excluded from the survey, it does not explain why they were excluded, especially since they existed as districts at the time of the census (2001) as well as at the time of the survey (2009). Then, while it is claimed that face-to-face interviews were conducted with 2374 individuals, selected by random procedure based on quotas set by gender, age, district and religion to reflect the population profile (as per the 2001 census), it is not clear why the unweighted size of the sample is not proportional to the size of the population of the district. For instance, for Kargil, which has a population of 119, 307, the sample size is 501, whereas for Jammu, which has a population of 1,588, 772, the sample size is only 457.
The result of all this is a report that lends itself to all kinds of interpretations. It yields results that are predictable in some instances and provides no explanation for the surprise results. It does not even attempt to correlate responses to two questions in the same section, which could provide supporting or contradictory evidence for a claim being made. For instance, a comparative analysis of response to the above questions on elections and utility of militant violence reveals that there is much more faith in elections leading to permanent peace than violence leading to a solution. The people in the Valley are largely neutral on either having any impact, reflecting a sense of hopelessness. Nor does it explain why the people in Rajauri and Punch, who feel they are Kashmiri, do not vote along the lines of the districts in the Kashmir Valley. And it certainly does not provide the “paths to peace”.
Related Publications