Decades of studies on counterinsurgency have theorized that social support is crucial for securing military victory. Insurgents rely on the support of locals to blend into the population, recruit fighters, and generate resources. Likewise, counterinsurgents rely on locals support to identify and root out insurgents. However, surveys show consistently strong support for the military even in areas dominated by insurgents. This study takes place in Colombia, where polls show high levels of trust in the Colombian military, even in guerilla-controlled and coca-producing regions where the military has fared the worst. What explains this discrepancy? The authors suggest that it may be a result of survey respondents falsifying their reported support for the military. Individuals who rely on insurgents to protect them and/or benefit from illicit activities enabled by insurgent control should be both unsupportive of counterinsurgents and hesitant to reveal that information, suspecting that counterinsurgents may be able to use it to target them. “Preference falsification” is likely to be common in this settings, which would undermine the accuracy of a typical survey intended to assess social support for counterinsurgents.
Using a survey experiment, the authors randomize indirect and direct questions about support for state forces. The indirect measure is a list experiment, a method used in other contexts to reduce pressure on respondents to falsify preferences, due to fear of coercion or social sanctioning. Under a list experiment, one group receives a list of control items, and another group receives the same list with an item of interest added, and respondents in each sample are asked how many items they support. The difference in means between the two groups represents the level of support for the item of interest in the sample. This method hides individual respondent support for a controversial item, as long as it is properly designed so that respondents are not likely to support all or none of the items. Compared to those who receive direct questions, the comparison of those who receive these indirect questions are expected to experience less social pressure and fear, and so should overcome social desirability bias.
The survey sample included 1,897 adult men and women from the noninstitutionalized Colombian population, including a subsample of 1,300 selected to construct a nationally representative sample and an oversample of conflict regions identified by level of violence. Forty-five municipalities in the six regions of the country were selected. The survey experiment was run on 1,423 individuals: for the indirect measure, 474 received the treatment list and 949 received the control list; for the direct measure, 474 respondents from the control group also received this question.
This article bolsters the theory that social support is important for the success of counterinsurgency. The authors reveal plausible evidence of “preference falsification” in response to direct questions about support for the military. The survey was conducted with funding from the U.S. government, and since survey respondents knew that, they likely would have seen it as state-linked, and so feel pressured to express support, perhaps suspecting that their responses would not remain anonymous. Once respondents were provided indirect questions that protect their anonymity using the list experiment, they report lower support for counterinsurgents. These differences were largest in guerrilla-controlled and coca-cultivating municipalities where individuals may have the most to lose if it were revealed to be guerrilla sympathizers or illegal cultivators by stating that they do not support the military. This result suggests that preference falsification is a plausible explanation for why the military receives such consistently strong support in surveys, even in contexts where support is predicted to be lower, given counterinsurgent performance in fighting.
The support of local population likely matters for counterinsurgency, and so understanding how to measure it is critical. Indirect questions can reveal attitudes that might be obscured by direct methods. Thus, list experiments that hide a particular respondent’s view but reveal views of communities may be useful for those seeking to study social support in conflict settings. More accurate measures of social support would allow policymakers to better estimate where and when support is collapsing, when populations are switching sides, and even detect when battlefield conditions might shift. The applications may even extend beyond counterinsurgency settings; more accurate measurement of social support for the government may be able to inform policymakers about “surprising” outcomes, such as sudden uprisings against a state.