Social Desirability Bias and Support for Military Veterans of the Iraq/Afghanistan Wars
Download data and study materials from OSF
University of Maryland
Sample size: 4000
Field period: 4/13/2011-8/1/2011
Although support for military personnel appears high, it remains unclear whether pride in military personnel and gratitude for their service translates to support for their wellbeing after service ends. No research to date has investigated whether support for veterans' benefits may be overstated due to a societal injunction to "support the troops." We evaluate the extent of public support for spending on recent military veterans' healthcare and for their inclusion in social life, and to estimate the extent of socially desirable reporting on these forms of support. We employ a list experiment, a technique that allows investigators to identify the extent of socially desirable reporting on topics, and introduce innovations in the analysis of such data that allow for multivariate modeling. We find the public appears to offer overwhelming support for social engagement with, and spending on healthcare for recent veterans. Support for military veterans is differentially felt, as is the extent to which people overstate their support across subgroups. There are important differences in levels of support by age, race, and political ideology, and variation in the amount of social desirability bias by race, political ideology, and prior military experience. Notably, African Americans express some of the lowest levels of support for returning veterans, and the greatest extent of socially desirable reporting on that support, even as they as a group have served in the military in disproportionate numbers, and have experienced greater returns to that service.
Support the troops hypothesis: Americans will express high levels of support for recent veterans.
Social desirability hypothesis #1: People will exaggerate their "true" support of veterans.
Social desirability hypothesis #2: Groups with relatively lower average levels of support for veterans will be more likely to exaggerate their support than those with higher levels of support.
List experiment in which sensitive item involves recent veteran of the wars in Iraq or Afghanistan, or tax dollars supporting veterans healthcare.
Number of listed items with which respondent agrees when asked about: types of people s/he would be pleased to have as a neighbor and programs to which more tax dollars should be allocated.
Summary of Results
When asked directly, 87.7% of respondents agreed they would be pleased to have a recent veteran move in next door, while approximately 85% would actually be pleased. Average support on the social distance/engagement measure is overstated by nearly three percentage points. On the question of whether more tax dollars should go toward healthcare for recent veterans, roughly 92% supported such a proposition when asked directly, but only 82% would actually support that position. Support for spending on veterans' healthcare is likely overstated by 10 percentage points under direct questioning. These findings are consistent with the "support the troops" hypothesis, as unbiased support is still quite high, with more than three quarters of Americans on the whole expressing support toward veterans on both dimensions. They are also consistent with social desirability hypothesis, which predicted that Americans would overstate their support of veterans, although this overstatement is greater for the question about deservingness. Overall, levels of support are generally high. However, our results suggest that there is a non-trivial amount of socially desirable reporting about support for veterans, especially support for spending on their healthcare, which varies by subgroups. African-Americans in particular appear to fairly consistently overstate their support for veterans by as much as 25 percentage points. Republicans also overstate their support for veterans, an unexpected result given the stereotypes of hawkish supporters of the military.