Call Us: US - +1 845 478 5244 | UK - +44 20 7193 7850 | AUS - +61 2 8005 4826

stewardship of modern day challenges

One explanation for the discrepancy between measures is confirmation bias. If respondents in the treatment group learned that civic engagement is normatively desirable, they
are more likely to dishonestly report engaging in civic activity. The positive effect on the
survey index may thus be a result of a treatment effect on norms rather than actual behavior.
Indeed, treatment makes people more likely to report civic participation on the survey than
on the register. I test this by creating a dishonesty index in which I subtract the total number
of events reported on the register from the Survey activity index calculated from the survey.
The greater this difference, the more likely they say they engaged in civic activity on the
survey without backing it up with reported events on the register. Treatment is positively
correlated with dishonesty. For this reason, I henceforth use only the civic activity register.
In all cases, the relationship between the control variables and civic participation go in
the expected direction. Schooling, leadership, relationship to the chief, and economic assets
are all positively and significantly correlated with civic engagement. Conversely, membership
in a marginalized group such as women, youth, and minority ethnicities is negatively and
significantly associated with civic engagement. Age has the expected quadratic relationship
where middle aged individuals are more likely to participate and elderly individuals are less
likely to participate.