At an event in Cleveland on Wednesday, President Obama endorsed a system of mandatory voting, a policy that would penalize eligible voters who don't make it to the polls on Election Day. The idea is not universally embraced, though it has significantly boosted turnout in Australia, where it's in place.

It's very easy to see why Obama likes the idea. He spoke longingly of increased turnout during the 50th anniversary ceremony in Selma, Ala., earlier this month. There's value in getting more people involved in the democratic process that doesn't tie directly to the success of his party, but it's clear that that's a motivation. During last year's election — an election in which non-white voters comprised only 25 percent of the vote, according to exit poll data — Obama's party was routed. Increasing turnout would almost certainly have helped.

We can actually run a thought experiment that looks at an alternate 2012 universe in which voting was mandatory. If every citizen of voting age had come out to vote during Obama's reelection race, the results would have been the same (that is, he would have won), but the electoral map would have looked much different.

[ Update: We revisited this idea with a look at state-by-state data in the 2004 election. ]

The Census Bureau has historic data broken down by state, race and gender, which we can compare to exit poll data. We don't have access to every state's data, which is a very big caveat, but we can still see what the election might have looked like.

Here's the actual result.


Here's how the electoral map would have looked if every voting-age person voted and we apply exit poll data by race.


And the same, if we apply exit poll data by gender.


Why the huge difference? A few reasons. First of all, notice that Indiana is dark red on that gender-based map. That's because we have data from Indiana specifically, which we don't for other, more Republican states. It's also more heavily white than many other places. This thought experiment doesn't differentiate between men and women within racial groups, which can vary widely; it's a rough calculation.

But black and Hispanic voters often turn out less heavily than white voters, and more heavily favored Democrats in 2012. Adjusting those numbers upward has a big effect. Obama clearly understands that such a boost could pay benefits in future elections, too.


State-by-state results

This table shows the net vote margin for Obama and Romney based on 100 percent turnout. States for which exit poll data was used are marked with an asterisk.

State Result based on
gender data
Result based on
race data
Alabama Obama 85,980 (2.5%) Obama 306,610 (8.8%)
Alaska Obama 8,730 (1.8%) Romney 35,300 (7.1%)
Arizona* Romney 210,060 (4.9%) Obama 9,080 (0.2%)
Arkansas Obama 50,910 (2.4%) Romney 19,340 (0.9%)
California* Obama 4,864,340 (20.8%) Obama 5,722,270 (24.4%)
Colorado* Obama 140,790 (4%) Obama 113,300 (3.2%)
Connecticut* Obama 401,490 (16.1%) Obama 428,690 (17.2%)
Delaware Obama 16,150 (2.5%) Obama 40,860 (6.4%)
District Of Columbia Obama 12,550 (2.7%) Obama 176,220 (38.2%)
Florida* Obama 107,840 (0.8%) Obama 774,350 (5.8%)
Georgia Obama 164,100 (2.4%) Obama 1,158,590 (17.2%)
Hawaii Obama 18,960 (2%) Obama 185,290 (19.9%)
Idaho Obama 23,620 (2.2%) Romney 150,510 (14.1%)
Illinois Obama 207,600 (2.4%) Obama 348,800 (3.9%)
Indiana* Romney 481,110 (10.2%) Romney 535,570 (11.3%)
Iowa* Obama 120,560 (5.4%) Obama 175,420 (7.9%)
Kansas Obama 45,490 (2.3%) Romney 180,550 (9.2%)
Kentucky Obama 75,760 (2.4%) Romney 355,300 (11.1%)
Louisiana Obama 87,370 (2.7%) Obama 483,940 (14.9%)
Maine Obama 22,020 (2.2%) Romney 183,300 (18%)
Maryland Obama 96,250 (2.4%) Obama 653,450 (16.3%)
Massachusetts* Obama 1,055,030 (22.1%) Obama 1,102,100 (23.1%)
Michigan Obama 172,530 (2.4%) Romney 133,240 (1.8%)
Minnesota Obama 83,440 (2.1%) Romney 477,940 (12.2%)
Mississippi Obama 55,200 (2.6%) Obama 395,470 (18.6%)
Missouri* Romney 416,140 (9.4%) Romney 760,280 (17.2%)
Montana Obama 15,550 (2.1%) Romney 129,890 (17.2%)
Nebraska Obama 27,540 (2.1%) Romney 152,850 (11.8%)
Nevada* Obama 134,580 (7.4%) Obama 52,440 (2.9%)
New Hampshire* Obama 62,160 (6.3%) Obama 60,870 (6.1%)
New Jersey* Obama 1,087,560 (18.3%) Obama 729,410 (12.3%)
New Mexico Obama 33,740 (2.4%) Obama 152,290 (10.7%)
New York* Obama 3,415,350 (26.1%) Obama 3,260,670 (24.9%)
North Carolina* Romney 209,510 (3.1%) Romney 301,960 (4.5%)
North Dakota Obama 10,460 (2%) Romney 80,300 (15.6%)
Ohio* Obama 207,180 (2.4%) Romney 166,660 (1.9%)
Oklahoma Obama 63,210 (2.3%) Romney 144,580 (5.3%)
Oregon Obama 60,800 (2.2%) Romney 358,090 (12.8%)
Pennsylvania* Obama 503,510 (5.3%) Romney 107,860 (1.1%)
Rhode Island Obama 19,250 (2.6%) Romney 55,160 (7.3%)
South Carolina Obama 87,510 (2.6%) Obama 350,560 (10.4%)
South Dakota Obama 12,770 (2.1%) Romney 94,300 (15.5%)
Tennessee Obama 118,040 (2.5%) Romney 49,280 (1.1%)
Texas Obama 377,510 (2.4%) Obama 2,703,410 (16.8%)
Utah Obama 40,810 (2.3%) Romney 206,290 (11.5%)
Vermont Obama 10,800 (2.2%) Romney 83,740 (17.2%)
Virginia* Obama 163,900 (2.9%) Obama 58,940 (1%)
Washington Obama 110,790 (2.3%) Romney 383,910 (7.9%)
West Virginia Obama 32,620 (2.3%) Romney 234,230 (16.2%)
Wisconsin* Obama 202,250 (4.8%) Obama 171,530 (4%)
Wyoming Obama 8,580 (2%) Romney 59,100 (14.1%)