The ambivalent effects of bonding social capital for local
representation
re-evaluating the effect of civic engagement on mass-elite policy
agenda agreement
Contributing to the debate which forms of social capital are the
cornerstone of functioning democratic governance and which are
harmful, Hill and Matsubayashi (2005) find that membership in
'bonding organizations' is negatively associated with mass-elite
policy agenda agreement in a community while bridging social capital
has no measurable effect. Our paper shows that the effects are much
more ambiguous than the authors claim. Hill and Matsubayashi (2005)
constructs the critical indices for social capital and participation
by aggregating from a small sample (~25 per community) of individual-
level data. Nevertheless the article treats these measures as
?precise?. We demonstrate that when taking into account the sampling
variability of this information, the uncertainty of the estimates
does not allow conclusive findings. Furthermore when we exclude
variables that pose post-treatment bias we find that bridging social
capital might actually have a positive effect on policy concurrence.
However a dataset with a substantially larger sample size per
community would be necessary to reliably assess these relationships.
---
Thomas Soehl
MPA candidate 2007 / McCloy Scholar
John F. Kennedy School of Government
Harvard University