That said their self-interest in defeating the Republican initiative in California has prompted them to back the plan devised by the Stanford professor, which uses an interstate compact to create a de facto popular election without having to do the practically impossible task of pushing through a constitutional amendment.

I think the public will largely back this but it still won't come into being for a while since it doesn't happen until that majority electoral vote state threshold is reached. States having almost half the electoral votes allow initiatives, so most of it could be done through submitting it directly to the voters of those states. That could create enough momentum in the remaining states to finish the job.

So to spencer's point, selected states doing this could certainly have moved the last couple of elections, but doing it nationally wouldn't have moved 200 or 2004. It would mean that the "battle ground" becomes broader.

While this is, on the surface, true, the main contention that there's a problem with the current system is that the 2000 election brought the wrong candidate (i.e., the candidate that won fewer actual votes) to office. What you've described is that a system based on constitutional districting actually exacerbates that problem. Therein lies one serious flaw in the Republican proposal as it stands (the others being that it that the only true motivation is partisan, and that such a system implemented on a peacemeal basis, as R. Alex points out, will only add layers of complication and error to an already seriously broken system).

Large population areas would be the big winners in a nationwide popularity voting structure and lesser populated areas would lose ground politically.

This is the stated reason for keeping the EC. Two criticisms. First, why is it bad that one person should have one vote? As it stands, a person voting in Wyoming has significantly more impact than a person voting in California, even though California has a much larger aggregate impact due to its obviously higher population. Second, does the system as it exists really benefit smaller states? Sure, they get face time, but how does the president really pay back Idaho (or Rhode Island)?

At any rate, until Texas and Georgia go the way of assigning their electoral votes proportionally, this is doomed to failure, as well it should be. We have to eliminate the EC on a national level, and until the National Popular Vote movement gets enough steam, I don't see it happening.

The reason my vote didn't count wasn't because I am not in the EC, but because there was no contest in my state. In 2004, there were 2 states (NM and Iowa) where the victory difference was less than 1%. In 2000, there were 5 states (FL, IA, NM, OR, WI ) with a difference less than 1%. In Florida, it was 537 votes that decided those EC votes. NM was 366 vote difference.

What I am saying is that states which are 'reliable' for one party or another also mean they aren't courted in the election because swaying 1% of the voters doesn't matter. On the up side, we don't have quite as many election ads.

To say your vote doesn't count because of the EC is to say that your vote doesn't count for congress critter because you don't get to vote directly on the legislation. Suck it up, is called representative democracy.

In 2000 and 2004, my presidential election vote did not count. I voted, it was tallied, but it was all moot as I vote in Texas.

Your vote didn't matter in any other year either. Neither did the votes of 300 million other americans. The popular vote is meaningless since presidents are only elected by the electoral college.

If you want your vote to actually matter you have to do with the EC altogether.

If we are going to keep the EC I do like the idea of reducing the granularity by using the county level, but as above that needs to be done by all states, not just one.

Now looking at this from 2004 perspective. Bush won 250 congressional districts and Kerry won 165 (not sure what is up with the missing 20 congressional districts). Add in the 31 states for Bush, 19 states for Kerry, 3 EV for district of Columbia to Kerry and the end result would be 312 EV for Bush and 206. Even if all the missing congressional districts went to Kerry, the end EV result would have been further for Bush than the actual result of 286 to 252.

Go back to 2000 and we see something similar. 226 congressional districts for Bush, 209 for Gore. 30 states for Bush, 20 states for Gore and DC for Gore. Net result is 286 EV for Bush, 252 EV for Gore, which is again better than the actual result of 271 to 266.

I couldn't find congressional district information for earlier years.

So to spencer's point, selected states doing this could certainly have moved the last couple of elections, but doing it nationally wouldn't have moved 200 or 2004. It would mean that the "battle ground" becomes broader.

James, I think a lot of people looking at this issue (the California/NC one) are overlooking a rather obvious flaw: if it's done by congressional district, gerrymandering battles will become ten times more contentious and will determine not only control of the House but also the presidency. A Democrat could win the vote in Florida but still lose its delegation 16-9 or 14-11 because of how the congressional districts are drawn up. Bush could have lost a majority of Texas's delegation in 2000 as they sent 19 Democratic reps and 13 Repubican reps (or something like that) to the House that year.

In regards to the de facto canning of the electoral college, I'm curious how would recounts work. I'm not saying that a workable system couldn't be developed, but it would add a layer of weirdness if there are some states that participate and others that do not. Getting rid of the EC formally would streamline that (recount everywhere) for better or worse. Getting rid of it this way could make some aspects of the system even more complicated and weird than they are now.

Large population areas would be the big winners in a nationwide popularity voting structure and lesser populated areas would lose ground politically.

But as long as only selected states implement this policy it would function strictly as a method for increasing the electoral vote for one party.

That would certainly be the aggregate national impact. From the state's perspective, though, it could make sense. For one thing, winner-take-all disenfranchises a sizable minority in this case. For another, "California issues" totally get ignored under the present system.

But as long as only selected states implement this policy it would function strictly as a method for increasing the electoral vote for one party.

Would you favor Texas and other southern states implementing this policy while California and the Northeastern states did not?