A few weeks ago I read an article about why the Electoral College was here to stay and I was inspired to crunch some numbers.
I went to a few .gov sites and collected the statistics I wanted and loaded everything into an Excel spreadsheet which can be seen here (sources included). Unfortunately, I had trouble when I tried to access the .xls from angelfire (owing to bandwidth-theft concerns, I imagine), so I had to use Excel to save it as a .htm.
The math is fairly intuitive, where it says “Persons per vote” I divided the state’s voting age population by the number of electoral votes in that state. Under the column “Votes per person” I have divided that state’s electoral votes by the population and then multiplied that number by 1,000,000 so as to deal away with that pesky, small number; the proportion, of course, stays the same.
In high school, I only took one year of a math course that covered statistics (somewhat briefly, at that), so I want to know if my treatment of the Electoral College is valid or if I’ve made a mathematical snafu…which I am prone to, from time to time.
Funny, I’ve been working out a bunch of different things on the exact same subject. The best comparison I found was comparing the ratio of EC votes the state got to the ratio of population it has. So something like
EC votes/ 535
state pop/USA pop
If an EC vote was exactly as powerful as a popular vote, those should be the same. They aren’t. I didn’t have number for voting age population, though, so I used total population in all my figures. I only really wanted an idea of what was happening, anyway.
What are you looking for with yours, exactly? EC votes per person is certainly EC voted divided by population.
Not necessarily, since the EC makes a multi-step process. So if Michigan republicans won the MI EC representatives, the MI republicans would be representing the entire state of MI, while MI dems would now be “uninvolved” in the election at the second stage. This happens regardless of how the EC votes are distributed among states.
While that may be a downside, there are upsides. I don’t think that the EC is supposed to be perfectly representational. Since every state gets EC representatives for each Senator and Representative, smaller states will have “disproportionate” representation in the EC.
The difference is significant, because the number of congress members from each state is determined by total population (including children and aliens), and not by the number of potential voters. From the same table, the states with the lowest proportions of citizens in their populations are:
California - 81.2%
New York - 87.2%
Nevada - 88.5%
Texas - 88.8%
New Jersey - 89.2%
and the highest is:
West Virginia - 99.4%
Since California has the highest population among the states, you’d expect its votes per person would be the lowest, but this fact puts it more towards the middle.
One more comment: the spreadsheet created by Call me Frank shows Pennsylvania, Michigan, Ohio and Florida as the states with the lowest votes per voter. This is because they are large states with a relatively low proportions of aliens. However, because they are close to evenly balanced between red and blue, three of these (PA, OH and FL) are battleground states, getting a disproprtionate amount of attention from the candidates. Size is not everything.
Exactly right. This is one of the main points of the electoral college, and it was included by the founding fathers specifically to increase the power of small states to guard against the tyranny of the majority.
Regarding the statistics, it might be interesting to compare population/EC votes to registered voters/EC votes. You could make the argument that either general population or registered voters is the appropriate measure. By looking at both, you might find some interesting trends in voter registration. By looking at registered voters instead of just voting age population, you’re choosing to ignore people who have chosen to be ignored. That is, if an eligible voter chooses not to be part of the process, why let them skew your analysis of the process?
If I did that, I would have to adjust how I counted the popular vote. I’m counting all the population votes in one column. It’s called “total”. I have to do the same for the EC. I cannot make other assumptions.
I misread you. You were talking about those ratios being the same, while I thought you were saying that if the EC were perfectly representational, then the process would always be no different from a straight popular-vote election. My bad.
As a side remark, the actual procedure of assigning number of congressional districts per state after each census is more complex than you can imagine.
Every assignment method that can be thought up have remarkably perplexing fairness issues. Once an electoral vote is assigned or not, it effects not only that state but the other 49 as well. Trying to get 438 equally distributed among 50 states and DC, with the additional requirement of a minimum of one, isn’t for the faint hearted.
Civics was covered even more briefly than Statistics…
Here is the same spreadsheet with the calculations based on the state populatoins given here.
I was unsure if the quantities I was comparing were appropriate. As Giles pointed out, they weren’t. Also, it occurred to me while doing some vacuuming after posting that that Colorado is now moving towards a less “winner take all” stance with its EC votes, in which case my numbers regarding Colorado would inaccurate. I wanted to know if there were any other such instances.