For the Media

Blue staters might take you to a liquor store and, in a blue state rather than a red state, you are slightly more likely to end up at a dog racing track.

The culture war can actually be a lot of fun. For example, bowling alleys per capita, at the state level is correlated with Bush vote in 2004, but miniature golf courses is negatively correlated - meaning the more miniature golf courses in your state, the more your state voted for Kerry in 2004.

Surprised? Comforted to have your world view confirmed? I have a whole list of these correlations between businesses and Presidential vote, but a word of caution first. These are simply correlations. All it means is that two phenomenon tend to vary together. Positive correlation means that as one variable goes up, the other tends to go up, and when one goes down, so does the other. Negative correlation means they vary in opposite directions: when one goes up, the other goes down. We hear about correlations all the time: carbon dioxide positively correlated with global warming, obesity correlated with heart disease, education correlated with earnings, and recently, IQ correlated with Presidential success.

The distinction that must be emphasized is between correlation and causation. Just because two variables are correlated, does not mean that one causes the other. An example of this distinction can be found in the debates about global warming. There is no doubt that carbon dioxide emissions are correlated with global warming. The question is more complicated when it comes to causality - to my knowledge, we cannot prove that carbon dioxide emissions cause global warming (the responsible public policy when faced with overwhelming correlational evidence and grave consequences is a different matter though). In fields like medicine, causation is easier to establish because of the use of controlled experiments. Medical science can establish that a particular drug causes lower blood pressure. In political science, causation is more difficult to establish — although experiments are increasingly employed to establish causation. When causation cannot be established — it is sometimes the case that the correlation is spurious, meaning that one variable really has no direct causal effect on the other variable, but the variables are associated because both variables are also correlated with a third variable. An example would be Diet Coke and obesity. Diet Coke, I don't think, can cause you to be obese. But obesity is correlated with attempts to lose weight, so yes, Diet Coke consumption is correlated with obesity, albeit spuriously.

So, some caution is in order when looking at simple correlations - going back to the original examples, it is hard to imagine how liquor stores could cause a state to vote Democrat (at least in this day and age), so the correlation is probably spurious. But spurious correlations are so much fun, so here we go:

Voting for Kerry in 2004 was correlated per capita by state with:

art schools
dog racing tracks
interior design schools
internet providers
law offices
libraries
liquor stores
miniature golf courses
shoe stores

Voting for Bush in 2004 was correlated per capita by state with:

bars
beautician schools
bowling allies
cellular phone stores
drive-in movie theaters
florists
hardware stores
newspapers
operas
radio stations
RV parks
used car dealers.

While it can certainly provide for a good conversation starter that the more used car salesmen in your state, the more likely it is to vote Republican, the correlation is certainly spurious. However, not all of these correlations are of no practical interest. For example, from a scientific perspective, that there are more cellular phone stores in red states does not prove anything, but it could be a point of departure for political scientists that study political polling. This is because there is a considerable debate about the accuracy of telephone polls that still almost exclusively call land line phones in an increasingly cellular world.

That these correlations do not show causality, also does not mean that they do not have predictive power. If two variables are correlated, one can predict the other. So, assuming 2004 is similar to other years (it might not be), the more shoe stores in your state, the more Democratic your state is and the more likely it is to vote for Obama this November. If you're curious, New Hampshire, as of 2002, had the most shoe stores per capita, while Alaska had the least.

This power of prediction varies depending on the strength of correlation - the shoe store correlation, is strong by social science standards, but not that strong (.36), so the predictive power of shoe stores on Presidential vote is weak. This simply reflects that shoe stores probably don't have that much to do with Presidential vote. The strongest correlations with red states were cellular phone stores and radio stations. For blue states: art schools and libraries.

It should be noted that the types of business establishments for this list were chosen because I thought they looked fun. I did not know the correlations before choosing the list. I did not consciously choose any that I thought would prove a correlation with one party or another. I have to admit that I was most surprised by the correlation of operas and Bush vote. I think this means that there are lot of opera companies out there that I don't know about. I also have to admit that I was hoping that the number of circuses per capita would be correlated with one of the two parties - it didn't matter which one. How great would that be? Unfortunately, from a comedic perspective, it looks like the circus is equal opportunity for red and blue states.

Permalink

On-topic comments that add to the discussion are welcome. Please respect each other and the forum by using your real name and a civil tone. Spam and comments judged by UCLA to be libelous, offensive or abusive may be deleted without notice

Comments

Randy says:

I like wrestling tooo muccchhh... http://www.wrestlingsuperstars.org

pri says:

prikitiuuu says:

This is actually really interesting regarding your fact article here, This article is very informative. http://www.moratmarit.com/2009/08/kenali-dan-kunjungi-objek-wisata-di.html | http://www.moratmarit.com/2009/09/oes-tsetnoc-contestants-from-indonesia.html | http://www.moratmarit.com/ | http://www.cahbagoes.com/ | http://blogger.moratmarit.com/

Randy says:

It's really nice Ryan ---- http://www.wrestlingvalley.org

me says:

http://dirtydigital.sensualwriter.com http://sexbondagesex.blogbugs.org http://lesbianbondagesex.smut.com http://sexbondagesex.sexusblog.com http://sexbondagesex.xlogz.com http://lesbianbondage.sexbiro.com http://bondage.seductiveblogs.com http://bondage.sexblogshost.com http://Bondage.xbloghost.com

Rusli says:

thank you, nice post http://blog.cicurug.com/stop-dreaming-start-action/ http://blog.cicurug.com/rusli-zainal-sang-visioner/

9ay says:

Thanks for this article. <a href="http://www.9ay.org">gebelik</a>

http://www.wwe.com says:

sdsdscdvc cdfdxd

Dennny says:

boses says:

Interested http://autorepairmanual.blogspot.com/

divorce advice says:

I agree with you. Thank you! http://divorceline.org/blog/

ricky says:

Good point Ryan, I totally agree with them. regards. http://itcomputerzone.com

Ryan Enos says:

Alex - I'm glad you think they're cute...you're right that there is something of more substance to the blue/red division, although I think that we can sometimes be too quick to think that we know what it is. Some of the apparent state level divide may hide a good deal of sub-state level variation that makes the geographic and cultural divide not so great. As an example, looking at your comment that the social ills that are prevalent in the red states, it would be interesting to see what happens when you take the South out of this equation. The reason being that I imagine that a considerable portion of the social ills prevalent in South are disproportionally concentrated in the African American population - which some would not consider part of the red state constituency - at least not when constructing the stereotypes that drive part of the fascination with the red/blue divide. My gut feeling has been that you are right about how an illegitimate pregnancy in the Obama family would have received different treatment - but I also have a feeling that there might be a considerable amount of red state republicans out there that would not have as much to say because they would be aware of the hypocrisy involved in such spin.

Alex says:

Cute stats; but I would like to hear the media talking more about what should be a conspicuous red/blue division. We may want to agree w/ Obama's magnanimous declaration that Bristol Palin etc. is out of bounds (even as they relentlessly thrust her just-us-folks family into the spotlight, and passed around her newest kid at the GOP convention) but the truth is that this DOES represent something. On nearly every values issue that the GOP has supposedly used as a wedge to neuter the Dems' economy advantage, the red states do considerably worse than the blue states. Teen pregnancy, divorce, illegitimacy, drug abuse and welfare dependency, etc. Can one imagine how an illegitimate pregnancy in the Obama family would have been spun? Certainly not as, over and above all else, the noble refusal to abort. Unlike the minigolf/bowling divide, this seems to me something bewildered watchers of the debate, and the shameful Palin "small town values" spectacle, should be made aware of.

RSS Alerts

Facebook Twitter Youtube

About

Election analysis and personal commentary by UCLA scholars, unfiltered.

News release

Contact UCLA Newsroom

RSS Feed

For the Media

Current Issues

Useful Links

Red states, blue states, and circuses

Comments

Randy says:

pri says:

prikitiuuu says:

Randy says:

me says:

Rusli says:

9ay says:

http://www.wwe.com says:

Dennny says:

boses says:

divorce advice says:

ricky says:

Ryan Enos says:

Alex says:

Tool Tip

Tool Tip

About

Our Bloggers

Archive