In the post below, Nate comments on the elimination of the electoral college. Its one of those ideas which has been around for quite a while, but never really gets any attention till you have an election like the one we had in 2000. Unfortunately, it is almost a mathematical impossiblility to change the system. Here is some data to show why:
| State | Electors | % of EC | Population | % of Pop | Diff | Var |
| California | 55 | 10.22% | 33,930,798 | 12.03% | -1.81% | 1.81% |
| Texas | 34 | 6.32% | 20,903,994 | 7.41% | -1.09% | 1.09% |
| New York | 31 | 5.76% | 19,004,973 | 6.74% | -0.98% | 0.98% |
| Florida | 27 | 5.02% | 16,028,890 | 5.68% | -0.66% | 0.66% |
| Illinois | 21 | 3.90% | 12,439,042 | 4.41% | -0.51% | 0.51% |
| Pennsylvania | 21 | 3.90% | 12,300,670 | 4.36% | -0.46% | 0.46% |
| Michigan | 17 | 3.16% | 9,955,829 | 3.53% | -0.37% | 0.37% |
| Ohio | 20 | 3.72% | 11,374,540 | 4.03% | -0.31% | 0.31% |
| New Jersey | 15 | 2.79% | 8,424,354 | 2.99% | -0.20% | 0.20% |
| Indiana | 11 | 2.04% | 6,090,782 | 2.16% | -0.12% | 0.12% |
| Georgia | 15 | 2.79% | 8,206,975 | 2.91% | -0.12% | 0.12% |
| Virginia | 13 | 2.42% | 7,100,702 | 2.52% | -0.10% | 0.10% |
| North Carolina | 15 | 2.79% | 8,067,673 | 2.86% | -0.07% | 0.07% |
| Washington | 11 | 2.04% | 5,908,684 | 2.10% | -0.06% | 0.06% |
| Wisconsin | 10 | 1.86% | 5,371,210 | 1.90% | -0.04% | 0.04% |
| Maryland | 10 | 1.86% | 5,307,886 | 1.88% | -0.02% | 0.02% |
| Massachusetts | 12 | 2.23% | 6,355,568 | 2.25% | -0.02% | 0.02% |
| Tennessee | 11 | 2.04% | 5,700,037 | 2.02% | 0.02% | 0.02% |
| Arizona | 10 | 1.86% | 5,140,683 | 1.82% | 0.04% | 0.04% |
| Kentucky | 8 | 1.49% | 4,049,431 | 1.44% | 0.05% | 0.05% |
| Missouri | 11 | 2.04% | 5,606,260 | 1.99% | 0.05% | 0.05% |
| South Carolina | 8 | 1.49% | 4,025,061 | 1.43% | 0.06% | 0.06% |
| Oklahoma | 7 | 1.30% | 3,458,819 | 1.23% | 0.07% | 0.07% |
| Oregon | 7 | 1.30% | 3,428,543 | 1.22% | 0.08% | 0.08% |
| Louisiana | 9 | 1.67% | 4,480,271 | 1.59% | 0.08% | 0.08% |
| Alabama | 9 | 1.67% | 4,461,130 | 1.58% | 0.09% | 0.09% |
| Connecticut | 7 | 1.30% | 3,409,535 | 1.21% | 0.09% | 0.09% |
| Minnesota | 10 | 1.86% | 4,925,670 | 1.75% | 0.11% | 0.11% |
| Mississippi | 6 | 1.12% | 2,852,927 | 1.01% | 0.11% | 0.11% |
| Colorado | 9 | 1.67% | 4,311,882 | 1.53% | 0.14% | 0.14% |
| Utah | 5 | 0.93% | 2,236,714 | 0.79% | 0.14% | 0.14% |
| Kansas | 6 | 1.12% | 2,693,824 | 0.96% | 0.16% | 0.16% |
| Arkansas | 6 | 1.12% | 2,679,733 | 0.95% | 0.17% | 0.17% |
| Nevada | 5 | 0.93% | 2,002,032 | 0.71% | 0.22% | 0.22% |
| Montana | 3 | 0.56% | 905,316 | 0.32% | 0.24% | 0.24% |
| Iowa | 7 | 1.30% | 2,931,923 | 1.04% | 0.26% | 0.26% |
| Delaware | 3 | 0.56% | 785,068 | 0.28% | 0.28% | 0.28% |
| Idaho | 4 | 0.74% | 1,297,274 | 0.46% | 0.28% | 0.28% |
| New Mexico | 5 | 0.93% | 1,823,821 | 0.65% | 0.28% | 0.28% |
| Maine | 4 | 0.74% | 1,277,731 | 0.45% | 0.29% | 0.29% |
| South Dakota | 3 | 0.56% | 756,874 | 0.27% | 0.29% | 0.29% |
| West Virginia | 5 | 0.93% | 1,813,077 | 0.64% | 0.29% | 0.29% |
| New Hampshire | 4 | 0.74% | 1,238,415 | 0.44% | 0.30% | 0.30% |
| Hawaii | 4 | 0.74% | 1,216,642 | 0.43% | 0.31% | 0.31% |
| Nebraska | 5 | 0.93% | 1,715,369 | 0.61% | 0.32% | 0.32% |
| North Dakota | 3 | 0.56% | 643,756 | 0.23% | 0.33% | 0.33% |
| Vermont | 3 | 0.56% | 609,890 | 0.22% | 0.34% | 0.34% |
| Alaska | 3 | 0.56% | 628,933 | 0.22% | 0.34% | 0.34% |
| District of Columbia | 3 | 0.56% | 574,096 | 0.20% | 0.36% | 0.36% |
| Rhode Island | 4 | 0.74% | 1,049,662 | 0.37% | 0.37% | 0.37% |
| Wyoming | 3 | 0.56% | 495,304 | 0.18% | 0.38% | 0.38% |
| Totals | 538 | 100.00% | 281,998,273 | 100.00% | 0.00% | 13.88% |
| 0.27% |
The table shows the difference between a states percentage of the national population (based on the 2000 Census) and their percentage of the Electoral College (based on the 2004 election). As you can see, California gets screwed the most with a 1.81% discrepency between their population and their electoral vote.
Also, there are only 17 states which suffer any sort of disadvantage and only 5 states which suffer any sort of big disadvantage (greater than 0.5%). The important number is that 34 states (we are counting DC as a state for this purpose) have some sort of electoral advantage. It requires 3/4 of the states (not including DC) to pass a constitutional amendment. To get to the magic 38 states required, you’d need the 17 states with a disadvantage to band together, and then get 21 states to vote against their self interest to benefit the 17. That is something I just don’t ever see happening.
Also, it should be noted that in the US, there are technically no nation wide elections for anything. The 2000 election shows why that is probably a good idea. Because we use an electoral system, all the focus was on Florida, because other close states (Oregon, New Mexico, Wisconsin) didn’t matter. The estimated number of flawed ballots in the US in the 200 election was estimated to be as high as 2m…….4x the margin of victory Gore got in the popular vote. The process of recounting votes in a nation this big could take longer than the transition period. It almost took that long in one state alone.
The problem is, in any close election, you can expect the margin of error in bad ballots, errors, miscounting, etc to be greater than the margin of victory. This usually isn’t an issue in most elections because in any one place, the odds of it happening are low so it doesn’t effect the outcome.
In the last column, I took the variance for each state (just the absolute value of the difference) and averaged it. The average variance for each state is only 0.27%. Given the low variance, you’d think that a mismatch between electoral results and popular results would be less common than it is. If you include the 1960 election (see below) its happened four times in history now. (2000, 1960, 1876, 1824) That’s 7% of all presidential races.
Also, few countries have direct election for the Head of Government (as opposed to a Presidential Head of State). All parlimentary systems use indirect voting to elect a Prime Minister. The largest countries I can think of that have a President with power are Russia and South Korea, and I don’t know exactly how their elections work. In the case of South Korea, the population is small enough that it would be like the California governor election.
All this aside, 7% of the elections is still too much. There are three solutions outside of a constitutional amendment that will reduce the odds of a 2000-type outcomes. 1) Split up the big states. California and Texas and probably Florida, should each become 2-3 seperate states. 2) Do proportional representation like they do in Nebraska and Maine. 3) Increase the size of the House of Representatives. The house is the same size as it was in the 19th Century for a much greater population.
***************************
Postscript: I realized something after I posted it. The fact that the 2000 and 1960 elections had skewed results might not be a coincidence. Before I put the table up I used electoral college data from 2000, not 2004. In that, California had an even bigger disadvantage, as did other growing states. In fact, every 20 years, the census is the same year as a presidential election. However, the results of the census won’t apply to new house seats till after the election.
In both 1960 and 2000, Nixon and Gore won California. While California wasn’t the largest state in 1960, it gained 8 electors between 1960 and 1964, and the basis of the 1964 electors was the population in 1960. Likewise the largest state, New York, which Kennedy carried lost 2 electors. These two states wouldn’t have changed the outcome, but there is a discrepency between when the head count is taken and when the electors are picked. I haven’t checked the numbers, but I’d guess that you’d find the least amount of variance between the popular vote and the electoral college in elections that took place in a year ending in ‘2’, and the most in elections ending in a ‘0’………..assuming the census effect is real.