Case Study: Brazil
Brazil represents one of the best illustrations
of how national politics intersects with the
increasing salience of the crime issue and
public fickleness regarding policy responses.
IPSOS polls and data from other surveys show that,
when combined with drug trafficking, crime is a close
second to unemployment in Brazilian concerns. Nevertheless, a referendum banning gun ownership in October 2005 failed. This was a surprise (at least to pollsters).
For almost a decade up to October 2005, more than 80
percent of Brazilians were in favor of prohibiting gun
sales. Yet, the referendum lost with a resounding 65
percent negative vote.
At the very least, this suggests that polls could be
misleading. Indeed, voters may state policy positions
contrary to their final intentions as they learn more
about the issue and consider the relative pros and cons.
The Brazilian experience and the referendum show that
nothing is written in stone.
That’s an important lesson—and warning—for policy makers. Public opinion on issues can change quickly
with the slightest push from sources like the media or a
special event like a referendum, and create rough sailing. Politicians must be wary that easy solutions may
be orphaned by the next swing in public opinion.
Here it is important to distinguish two important
concepts concerning public opinion: ( 1) rank ordering
of voter preferences and ( 2) favorability towards a given
issue or policy measure. Voter preferences (or policy
preferences)—the rank ordering of solutions—change
slowly over time, typically as a result of external shocks
to the system such as an economic downturn or a terrorist attack (9/11). On the other hand, favorability towards
a given policy measure can change radically as new
