Liberals have been touting a New York Times/CBS poll
released last weekend suggesting broad support for government
health care and higher taxes. Bill O'Reilly has written a
column
criticizing the poll because 48 percent of respondents said they
voted for Obama, compared to just 25 percent who said they voted
for McCain -- obviously a huge discrepancy. But I noticed this on
Monday, contacted a few people, including the New York
Times, and decided not to write anything, because there was
an explanation that sounded plausible to me -- that after an
election, there's a tendency for people to want to say they voted
for the winner rather than the loser. If you look at the
party identification and
ideological breakdown of the respondents in the poll, it's
actually pretty fair: 29 percent Republican, 35 percent Democrat,
and 31 percent independent; 19 percent liberal, 41 percent
moderate, and 34 percent conservative.
Here's the way Marjorie Connelly of the New York Times
described it to me in an email:
When we ask respondents to our surveys who they voted for in
2008, we do so in order to look at voters who say they voted
for Barack Obama as compared to those who say they voted for
John McCain. We're not looking to recreate the 2008 election,
because recall of past vote is notoriously unreliable as a
guide to what actually happened.
Traditionally, candidates who have won the election get a boost
after the fact. Voting is socially desirable, and people who
didn't vote for whatever reason think they had and say they
voted for the winning candidate. In addition, the gap between
the candidates often expands and contracts as the
president's popularity goes up and down.
In general, we use this question to provide crosstabs. For
example, we might have looked at people who said they voted for
Mr. McCain to see how what they said about health care reform.
But, as you can see from our article, we used party
identification for analysis instead.
Does this mean I think that most Americans are ready to embrace
government health care? Not at all. The Washington
Post/ABC
poll that came out on Wednesday asked some helpful followups
that showed how susceptible voters' attitudes are to messaging.
For instance, when asked whether they would still support the
creation of a new government plan if it would drive private
insurers out of business, support nosedived to 37 percent. We
still have a long debate ahead of us, and if the
lackluster ratings for ABC's Obama health care special are
any indication, Americans are not very engaged right now.
