Tuesday, March 10, 2009

What McCain was up against: The Economy

I just cracked open the initial release of the 2008 ANES data and thought I'd begin by focusing on economic evaluations. What the data show is something I think we all had a sense of at the time, but the pattern of economic evaluations is pretty dramatic nonetheless. Here's a look at the distribution of retrospective economic evaluations (national economy over the past twelve months) from the pre-election survey:
As you can see, economic retrospections were decidedly negative, with fully 89% rating the economy as "somewhat" or "much" worse than a year ago. Moreover, virtually no one (3%) rated the economy as doing better, and only 9% said it was the same as a year before. The best way to fully appreciated the negativity of these economic attitudes is to compare the distribution of responses in 2008 to other years:

Here we see that 2008 stands out in two important ways: the overall level of negativity and the near uniform agreement about economic conditions. Economic evaluations were generally negative in a couple of the other years (1980, 1992), but even in those cases there was more variation around the mean outcome. McCain was up against a very negative economic environment and there was widespread agreement about that environment.

In fact, given the lack of variation in economic attitudes, it occurred to me that even Republicans must have expressed a high level of negativity. Here are are the data broken down by party affiliation:

Sure, there is a tendency for Democrats to be more negative than Republicans, but "much worse" is the modal response in every partisan category, and fully 83% strong Republicans rated the economy as either "somewhat" (35%) or "much" (48%) worse.

There's a lot more to look at here and I'll try to follow up with other angles on the economy and the 2008 election.


Josh Putnam said...

Thanks for posting this, Tom.

One additional layer I'd like to see added is when the actual interview took place. The ANES questionnaire documentation indicates that the pre-election interviews began on September 2 (first delayed day of the Republican convention) and ended on the day before the election. What happened about two weeks into the process was the Lehman Brothers collapse.

How do the pre- and post-Lehman evaluations differ?

Granted, we're talking about a period of just under two weeks out of the entire nine week interview period, but I'd still like to see how those economic evaluations differ. I'd suspect quite a bit (and that things went from bad to worse for McCain in the process).

Tom said...


Good question, and one that I am very interested in exploring. Unfortunately (at least based on my perusal of the early release code book), this version of the data set does not have the date-of-interview variable (let me know if you find it!). I can say, based on the June and October waves of the NES 2008-2009 panel study, that there was very little movement in retrospective evaluations between June and October--they were decidedly negative in both waves, with October just a bit more negative. I'll try to post some data on this in the next couple of days.


Josh Putnam said...

Yeah, I glanced through the codebook after I left that comment and saw that variable wasn't included. That's too bad and somewhat odd to me.

But you're right, the panel data does get at this and it's very interesting that there wasn't too much movement in retrospective evaluations.

I look forward to the posts. Thanks again. Good stuff.

Anonymous said...

hello... hapi blogging... have a nice day! just visiting here....

Wow Gold said...

Very Nice Blog

Wow Gold said...

very nice blog

jaycutler said...

inventory journal

operation software

director software

driver software

factory software

pub software

baker software

baker software

county software

dancing log