If you're curious why the proposed changes to the electoral college in Virginia are a big deal that would take America several steps further away from true democracy, I'll make it as simple as I can by using three states to illustrate the problem.
I'll begin with North Carolina, mostly because that's where I live, but it also was conveniently close in 2012. North Carolina has 9.7 million people living in it, making it the 10th largest state by population. Each state gets two Senators and a number of Representatives to be determined by its population according to the most recent census. For NC, that's 13. A state's electoral vote count is then determined by its combined representation in the Senate and House, which is how North Carolina ends up with 15 electoral votes.
North Carolina had been a reliably red state until 2008, voting for the Republican Candidate from 1980 until 2004, spanning six elections. A growing minority population combined with a decline in manufacturing and farming jobs and an increase in bio-tech, information technology, and banking jobs in the Research Triangle Park area -- home to some of the best medical facilities and colleges in the country -- has been pushing North Carolina towards the center for decades. Although Mitt Romney clawed it back last year, NC is going to be a battleground state for many years to come, and may eventually become more blue than than purple.
[Note: Originally, I was going to take a map of county wins for the 2012 election (standard red/blue) and overlay district lines for North Carolina, and spent half an hour drawing the district borders myself in Photoshop before I realized that I was using district lines from just after the civil war. Oops. Then I spent another 15 minutes trying to find the borders for 2012, and found out that what I want to do is impossible. Districts in North Carolina have been so horribly gerrymandered by the state's Republicans that only a handful of districts mostly follow county borders, and none of them follow only county borders. I would need expensive mapping software and precise GPS coordinates for every single district border in order to properly overlay them, because they are physically impossible to draw by hand with accuracy. Think about that for a moment; gerrymandering in North Carolina was so bad after 2010 that you can't draw a district map without satellite mapping technology.]
Barack Obama won North Carolina in 2008 with 2.14 million votes and three districts to John McCain's 2.12 million votes and 10 districts. Mitt Romney won it last year with 2.27 million votes and 10 districts to Obama's 2.17 million votes and 3 districts.
That really shows how awful gerrymandering has become today, and how unrepresentative the U.S. House of Representatives is. (Democrats won more votes in 2012, but Republicans won a majority of the seats.) Obama only won three districts, but won them by more than 40 points: +43 (1st), +43.2 (4th), +58.5 (12th). Those districts are the only districts that elected a Democrat to the House of Representatives in 2012 other than NC-07, which elected Democrat Mike McIntyre while voting for Mitt Romney 59-40. Mitt Romney won all of his districts but one (NC-11) by less than 26 points, but didn't win any with less than a 10 point margin, making all of them safe districts for Republicans.
Democrats have been concentrated in three districts to make them overwhelmingly Democratic as a trade off for taking Democrats out of the overwhelming majority of the rest of the districts to make all of them much safer for Republicans. This is compounded by Democrats being naturally concentrated in cities and suburbs. That is how the game is now played. Three Democratic districts that Obama can win 70-30 is a good trade for 10 districts that Romney could win 58-42 or 60-40. The goal of gerrymandering is to make all districts uncompetitive for House races, but the result under the Virginia plan would also make the White House uncompetitive as well.
Under the Virginia plan to award electoral votes by total number of district victories, John McCain would have won North Carolina in 11.5 to 3.5 instead of losing all of its votes. The net change would have been 11.5 more electoral votes for John McCain and 3.5 fewer for Mitt Romney. Romney's case is why you're only seeing proposals for changing the electoral college this way in states that voted for Barack Obama, that also have legislatures controlled by the Republican Party, but not in any reliably red state that. State Republicans don't want to reform the electoral college, they want to rig the election to favor themselves. Otherwise they'd be pushing this plan in every state in the union, like Texas and California, the other two states serving as examples.
That is not a political claim, it is a statement of fact.
Mitt Romney won 12 of California's 53 districts to Barack Obama's 41, giving them something like 14 (Romney) electoral votes and 42 (Obama) if you round them off, whereas Obama won all 55 last year. The Republican candidate would gain more electoral votes from "reforming" California than they could win under the current system from the entire states of Virginia, Washington, Tennessee, Massachusetts, Indiana, Arizona, Wisconsin, Missouri, Minnesota, Maryland, South Carolina, Colorado, Alabama, Louisiana, Kentucky, Oregon, Oklahoma, Connecticut, Utah, Nevada, Mississippi, Kansas, Iowa, Arkansas, West Virginia, New Mexico, Nebraska, Rhode Island, New Hampshire, Maine, Idaho, Hawaii, Wyoming, Vermont, South Dakota, North Dakota, Montana, Delaware, and Alaska.
In retrospect, I probably should have listed only the states that have more electoral votes in total than a Republican could gain by "reforming" California: North Carolina, Georgia, Michigan, Ohio, Illinois, Pennsylvania, Florida, New York, and Texas. (New Jersey has exactly 14).
Of the 11 states that have 14 or more electoral votes, eight of them voted for Barack Obama in 2012. That's 72%, which isn't unexpected. With the exception of Texas (which will probably change within a decade) the more populous a state is, the more liberal it votes at the national level.
Speaking of Texas, I don't have complete information for 2012, but I do have it for 2008, so I'll project where necessary. Mitt Romney won 72% of the districts last year and all 38 of the state's electoral votes. Under the Virginia plan, Romney would instead only win 26 electoral votes, to Obama's 10 electoral votes from winning 27% of the districts in Texas. Obama actually did better in Texas than Romney did in California, but Romney would come out ahead because Texas has a smaller population.
To summarize, here's what the electoral vote distribution would look like under the Virginia play for my three states:
North Carolina: Obama 11.5, Romney 3.5 (net change from 2012: Romney +3.5)
Texas: Romney 26, Obama 10 (net: Romney -12)
California: Obama 41, Romney 12 (net: Romney +14)
The net result from those three states would be -1.5 electoral votes for Mitt Romney and +1.5 for Barack Obama. That's why you are hearing about these kinds of plans in states like Virginia, and Pennsylvania. (You won't hear about it for California because Republicans are utterly powerless there; Democrats control every state office and now hold super majorities in both houses of the state legislature.)
Now look at the popular vote in those states:
North Carolina: Obama 50.4%, Romney 48.4%
Texas: Romney 57.2%, Obama 41.4%
California: Obama 60.2%, Romney 37.1%
Nate Silver ran national numbers (thankfully) and found a net gain of 71 electoral votes for Mitt Romney last year if every state enacted something like the Virginia plan, enough to win the election by 7 electoral votes. That's despite Barack Obama winning more states and the popular vote by nearly 5 million votes.
Although districts are supposed to be equal in representation (each of North Carolina's 13 districts has roughly 619,000 people in them) in the total number of people, gerrymandering has upset the balance such that districts will vote to elect whoever you want based on where you put the borders. Mitt Romney only had 92,000 more votes that Barack Obama in 2012, but the districts as drawn would have awarded Romney 76% of the state's electoral votes despite winning the state popular vote by just 2.06%.
Now put the two things together. Here are those three states under the Virginia plan:
Obama's popular vote share: 48.4%
Obama's electoral vote share: 23%
Obama's actual 2012 electoral vote share: 0%
Romney's popular vote share: 50.4%
Romney's electoral vote share: 77%
Romney's actual 2012 electoral vote share: 100%
Romney's popular vote share: 57.2%
Romney's electoral vote share: 72%
Romney's actual 2012 electoral vote share: 100%
Obama's popular vote share: 41.4%
Obama's electoral vote share: 27%
Obama's actual 2012 electoral vote share: 0%
Obama's popular vote share: 60.2%
Obama's electoral vote share: 77.3%
Obama's actual 2012 electoral vote share: 100%
Romney's popular vote share: 37.1%
Romney's electoral vote share: 22.6%
Romney's actual 2012 electoral vote share: 0%
While it's clear that the popular vote winner taking all of the electoral votes is unfair, this new system wouldn't be any better. It's also worth noting that Republicans use a winner-takes-all system in most of their primary elections and have no problem with the disenfranchisement of voters when all of them are conservatives.
The better reform system would be to stop using the electoral college and move to the direct election of the President by popular vote instead. The only change in the last 50 years would have been Al Gore winning the election in 2000, while the new system would probably have resulted in far more Republican Presidents under a system that is no more fair to voters than the one we currently have.
* * *
Looking at Nate's numbers more closely, one thing has become very obvious to me all of the sudden. If all states adopted the Virginia plan to rig the Presidential election via gerrymandered districts, (not maximized), it would only take not getting about eight electoral votes to make it fall apart on the GOP side. Getting all the states on the red side except Michigan would leave the GOP candidate (based on 2012 results) two electoral votes short. Failing to rig anything larger than Michigan, like Florida (a 16 EV swing) would only make the gap larger.
Thus Republicans could rig every single state in the nation, and all Democrats have to do to keep it from being an end-game scenario is block "reform" in any of these states: Michigan, Virginia, California, Ohio, Pennsylvania, or Florida. Any one of them would make it pointless.
Things are a bit different on the Democratic side, and that's the reason I haven't bothered to acknowledge the obvious, which is that while Democrats can play this game too, there is both less upside for them (as Nate's data shows) and also less need. Democrats already start in the 250-260 EV territory with their safe states based on the current system which allocates all of a state's EVs to the winner of the state popular vote, except Nebraska and one or two other small and (no offense) inconsequential states.
The fun thing to do is look at the electoral college result if each side got every state they wanted, and the other side got none.
In that case, Mitt Romney would have won 309-229, and Barack Obama 368-170. Given that Obama won 332-206, it's not the case that Democratic shenanigans would be stealing the election, so much as it would be improving the Democratic candidate's position from default favorite to win, to almost untouchable -- while still legitimately winning the popular vote. On the flip side, it would take a Republican candidate and move them from default favorite to lose, to narrowly winning by almost the smallest margins possible, while losing the popular vote by a wide margin.
* * *
While it's true that Democrats have engaged in gerrymandering, they haven't been anywhere near as aggressive about it as Republicans have. Gerrymandering is largely why we have a Republican-controlled House of Representatives in a country that has more Democrats than Republicans, that once elected Democrats to a majority for 40 straight years. With surging minority populations coupled with party affiliation polls, the data tells us that the House should be controlled by Democrats right now, with Democrats gaining larger and larger majorities every election for the foreseeable future.
Finally, I have the necessary data for most states to tell you how the 2012 results would have changed with the Virginia plan. Feel free to ask in comments if you're curious.