Saturday, July 5, 2008

Reverse-Engineering Rasmussen's Ideological Breakdown of the Electorate

Today, Rasmussen put up this awesome ideological breakdown of the electorate, based on 15,000 likely voter interviews conducted throughout June. This is the kind of thing that really makes me wish Rasmussen would allow you to buy access to cross-tabs for individual polls, rather than have to pay $20 a month to subscribe. Not wanting to shell out $20, I've instead decided to reverse-engineer the numbers they've publicized to try to come up with the cross-tabs myself. As a caveat, I should say that the published numbers are rounded and therefore my numbers will not be exact.

Here's what I've found, remember these are, at best, estimates of the actual numbers.

On Fiscal issues, voters breakdown asi:

Conservative 38.5%
Moderate 45.5%
Liberal 11%
Not Sure/Other 5%

On Social issues, asi:

Conservative 35.8%
Moderate 32.7%
Liberal 29.6%

But the real genius of this effort by Rasmussen is that they pair ideological positions in each category to more closely show the ideological contours of the electorate. For example, they tell us (aka, not my estimates) that 24% of the electorate is both fiscally and socially conservative.

Here's that breakdown:

Fiscal and Social Conservatives (24%)
Fiscal and Social Moderates (20%)
Fiscal Moderates and Social Liberals (15%)
Fiscal Conservatives and Social Moderates (10%)
Fiscal Moderates and Social Conservatives (10%)
Fiscal and Social Liberals (9%)
Fiscal Conservatives and Social Liberals (4%)

Here's another way to look at it (also based on published data). Of social conservatives (35.8% of voters), 67% are also fiscally conservative, while 28% are fiscally moderate. Among social moderates (32.7% of voters), 62% are also fiscally moderate, while 30% are fiscally conservative -considering myself a social moderate (uncomfortable with abortion) and a fiscal liberal, I'm apparently in a tiny, largely non-existent group. Of social liberals (29.6% of voters), 51% are fiscal moderates, while 30% are fiscal liberals and about 13% are fiscal conservatives.

From yet another angle we see that fiscal conservatives (38.5%) are 62% social conservatives and 26% social moderates. Fiscal moderates (45.5%) are 43% social moderates, 33% social liberals, and 22% social conservatives. Fiscal liberals are 79% socially liberal.

How do all of these categories translate into votes?

Fiscal and Social Liberals (9%) support Obama 91% to 6%.
Fiscal Moderates and Social Liberals (15%) support Obama 80% to 13%.
Fiscal and Social Moderates (20%) support Obama 59% to 30%.
Fiscal Conservatives and Social Liberals (4%) support Obama 53% to 38%.

On the other hand,

Fiscal and Social Conservatives (24%) support McCain 82% to 13%.
Fiscal Conservatives and Social Moderates (10%) support McCain 67% to 25%.
Fiscal Moderates and Social Conservatives (10%) support McCain 51% to 40%.

Or, as Rasmussen says,

"Looked at from a different perspective, 25% of Obama’s support comes from voters who are fiscally moderate and socially liberal. Twenty-four percent (24%) are both fiscally and socially moderate while 17% are fiscally and socially liberal. No other group provides more than 8% of Obama’s support.

Forty-five percent (45%) of McCain supporters are both fiscally and socially conservative, 15% are fiscally conservative and socially moderate, 14% are both fiscally and socially moderate, and 12% are fiscally moderate and socially conservative."

Maybe I'll post on what I think all of this means after I mull it over a bit.

1 comment:

clark said...

I really want to see the cross-tabs. My first reaction was shock that social liberal/conservativeness had such an overwhelming effect on presidential preference, but I think we're really seeing the generation gap rehashed. Older people are significantly more likely to see themselves as socially conservative, and even though our generation tends to be more conservative on the issue of abortion than that of our parents, I don't think that that is the definitive issue when it comes to self-classification of social leanings. In fact, I would argue that glbt rights is the defining social issue of the day, which is likely why I see the age issue cut through.