Why Climatetologists Shouldn't be Economists (or Political Scientists)

This image was taken from one of the previous issues of The Economist. Basically the graph compares California and other "blue states" (as defined as supporting John Kerry in the previous election), compared to the "red states" as defined by those that voted for George Bush. Now, the Economist, as well as the source for this graph (http://sciencepolicy.colorado.edu/prometheus/archives/climate_change/001093understanding_us_cli.html, believe that this is some sort of indicator for something political. Of course, this graph leads alot to be desired. For example, using votes for the President as an indicator of the political leanings of the state is probably a sloppy way of determining the two. But it's quick and easily identifiable.

This doesn't even concern the title of the graph itself. For example, states that are "red" are more likely to be in the South or Midwest. These are also states that have hot summers. About 90% of all air conditioning is run on electric. So there is absolutely no adjustment for climate. The same can be said for just using per capita kWh usage without adjusting for industrial usage. Red States are much more likely to have energy intensive industries there than Blue States. For example, I happen to know that there are alot more Alcoa plants in the Southeast than in the Northeast. So obviously, per person kWh usage is going to be higher in states that have these electric-intensive industries than those without. But this graph doesn't make any adjustment for that. A much more telling graph would have been to compare RESIDENTIAL per capita kWh usage compared across red and blue states. But this doesn't happen.

The problem is that these sources believe that this is some sort of indicator of future politics. Case in point is this article by the same source, telling titled: Why Al Gore Will be the Next President of the United States

Ignoring for the moment that the former Vice President hasn't even announced that he's even exploring a Presidential candidate, the author of this post seems to make his position based on a couple of ideas:

1. 2008 will be different than 2004. Elites have decided that global climate change is an issue worth politicizing, that is to say, worth making an issue in politics. Therefore, carbon dioxide emissions will be an issue in the 2008 election.
2. But it seems reasonable to think that the smaller the cost (or perceived cost) of policies on carbon dioxide, the more likely that such policies will be accepted. Similarly, the higher the costs, the greater the likelihood of opposition.

I'm not going to disagree with him on point number two. That makes sense. Number one, I'm going to disagree with. Just because the "elites" have deemed this to be an issue, that doesn't mean that it will be a "make or break" issue. Most polls have shown that global warming/environmental issues rank rather low on voters lists of concerns. Here's some data on voter's priorities. Notice how the environment (unless it gets lumped in with the "energy" category) doesn't show up.

But the bigger issue (and where the author is completely misled) is that he thinks that states with higher CO2 emissions are MORE likely to switch sides. The problem is that this is the opposite. A candidate that would favor higher CO2 restrictions would more likely lose voters than gain them.

Let's take a look at an example. Use a generic high CO2 state, say Pennsylvania. Now, this state has a baseline economic performance concerning production, costs, etc. This baseline, for whatever reason, includes producing more CO2 than other states. Now, say you have a choice in two situations. One keeps the status quo for CO2 and the other creates some kind of cost to production of CO2. If you institute CO2 costs, then states with higher CO2 production will have a higher impact on their economic performance. Thus, either these costs will be passed onto voters (ie higher electricity costs, higher prices for goods) and/or the company will shutter facilities in Pennsylvania and move them to a state with lower costs. Either way, voters in Pennsylvania will see CO2 charges as a NEGATIVE. Thus, if they are rational (a tenuous assumption at best), they will chose the candidate that will impose less of an impact on their economic well-being. Unless you think that voters would want someone to basically force lower CO2 consumption on them. In which case, states with higher CO2 production would support an Al Gore over a Republican. That's what the author believes.

Of course, if that's the case, then all the low CO2 states would support Republicans since they already have low CO2 consumption and thus would care less about global warming issues. But the author wants to have his cake and eat it too. He thinks that low CO2 states will continue their support for CO2 restrictions even though it won't affect them while states with higher CO2 production will want to have these restrictions put in place, in spite of their disproportionate consequences. It's one or the other, but not both.

In the end, the author makes two huge mistakes. The first is that he believes that global warming will be the make or break issue in 2008, when in fact there is considerable evidence to the contrary. The second is that he thinks that he confuses correlation with causation. Basically, he's writing that since states with lower CO2 emissions vote Democratic, that if your state has lower CO2 emissions, you will vote for Al Gore. However, there's no causation between the two.

Then again, I learned that in my statistics class back in 2002. Perhaps they didn't teach that when this professor took his statistics class.