Leave Me Alone: Depression, Item Response, and Political Representation

• In Event: Unequal Access? Voting and Participation of Disabled People

Abstract

Response to polls and surveys is an important means by which citizens' policy preferences are communicated to and represented by elites (Brehm, 1993). The representation process is thus skewed if certain groups are less likely to respond -- either to the survey as a whole or to individual items within it. Mental health, and particularly depression, is an obvious source of skew when it comes to item non-response. Sufferers from depression are prone to avoid exactly the kind of cognitive and emotional effort that is involved in tasks like recalling considerations or reconciling internal conflicts, and so will be attracted by non-response or other satisficing options such as acquiescence. Drawing on a wide range of health and social attitudes surveys, including the European Social Survey and Understanding Society, we test this hypothesis by modelling the likelihood of non-response (and other satisficing devices) based on depression and various other individual characteristics. Then, shifting to a meta-analytic approach in which the unit of analysis is the survey item, we test various hypotheses about the types of questions on which depressives are particularly prone to non-response. Together, these analyses provide a detailed picture of how - and how far - depression affects this channel of political participation and representation.

Authors

• Luca Bernardi, University of Liverpool

• Robert Johns, University of Essex