If President Obama wins reelection by three or four Electoral College votes next month, the reason may be simple: noncitizens, mostly immigrants, who don't have the right to vote. No, I'm not talking about his immigration policy or his popularity with Latinos. Nor does this have anything to do with voter fraud. Rather, an Obama victory could hinge on a quirk in the Constitution that gives noncitizens, a group that includes illegal immigrants and legal permanent residents, a say in electing the president of the United States.
As required by Article I and the Fourteenth Amendment, the decennial census, which allocates to each state its congressional seats and Electoral College votes, is based on a count of all people who live in the United States, citizens and noncitizens alike -- or as the Constitution phrases it, "the whole number of persons in each state." That means millions of noncitizens who are ineligible to vote are still included in Electoral College calculations, and that benefits some states over others. Most of these noncitizens are here legally; however, the Pew Hispanic Center estimates that about 45 percent of noncitizens are undocumented immigrants.
In 2010 and most previous years, the census did not inquire about citizenship, but the American Community Survey (ACS), which samples our population every month, includes a breakdown of citizens and noncitizens. Plugging the 2010 ACS citizen-only numbers into the Census Bureau's apportionment formula shows that five states benefit electorally from their noncitizen populations: New York, Florida and Washington each gain one congressional seat and thus one Electoral College vote; Texas gains two; and California -- with 5,516,920 noncitizens out of a total population of 37,341,989 -- gains five.
Pennsylvania, Ohio, Michigan, North Carolina, Indiana, Iowa, Missouri, Louisiana, Oklahoma and Montana each lose a seat under the official formula as compared with an apportionment that counts citizens only.
It is ironic that Florida, which is in the news for its heavy-handed effort to purge noncitizens from its voter rolls, actually has a greater voice in the presidential election because of the nearly 2 million noncitizens who live there.
Looking at how the states might vote in November, there is no scenario in which Mitt Romney benefits from the inclusion of noncitizens in the Electoral College calculation, but there are several in which Obama could gain three to five Electoral College votes, thus deciding a close election.
To play out a few of these scenarios, let's first assume that each candidate wins states that are competitive but leaning his way. So Michigan and Pennsylvania go to the president, and North Carolina and Arizona go to Romney.
That leaves eight battleground states up for grabs: Ohio, Florida, Virginia, Wisconsin, Colorado, Nevada, Iowa and New Hampshire. Obama begins with a base of 237 Electoral College votes, and Romney comes in at 206.
If the president won Ohio, Wisconsin and Iowa, and Romney took the rest, Obama would prevail by 271 to 267 Electoral College votes. However, if the Electoral College calculation did not include noncitizens, Mitt Romney would win 270 to 268.
Or say the president won Ohio, Virginia and New Hampshire, with the other battleground states going to Romney. That would result in a 272 to 266 Obama victory -- which could have been a 270 to 268 Romney win if only citizens had been counted.
If the president won New Hampshire, Virginia, Wisconsin and Colorado, that would give him 273 Electoral College votes to Romney's 265. But recalculate it with citizens only, and it would be a Romney victory, 270 to 268.
Various other combinations are possible, but the bottom line is that Obama's Electoral College tally benefits from counting noncitizens, while Romney's does not.
So if including noncitizens distorts the Electoral College and potentially affects the outcome of our nation's most important vote, why do we continue to count "the whole number of persons" -- noncitizens the same as citizens -- in the census?
There are many good reasons.
Communities need accurate population data when considering whether to open or close schools, add police officers and firefighters, and provide health care and social services. Businesses use this information to decide where to locate, what to stock and how much to invest.
Accurate population information also guides how federal dollars are allocated to states and localities, and this money is often critical for communities with large numbers of immigrants, who often need social services. Who lives where also determines infrastructure decisions at the federal, state and local levels.
Nor would we obtain the accurate information we need if we required the Census Bureau to ask about citizenship status. Few undocumented immigrants would willingly cooperate with census-takers or answer the questionnaire if that meant giving the government information that could be used to deport them. Already, the census may be undercounting undocumented immigrants by about 12 percent, according to the Pew Hispanic Center.
It also makes sense to use the full population when allocating congressional districts. The lives of citizens and noncitizens are intertwined, and asking members of Congress to distinguish between them could create unnecessary tension and division.
But it makes no sense to give noncitizens a potentially decisive say in determining who should be president. A core privilege of citizenship is the right to vote and decide the fate of our country.
Before Republicans cry foul that the Electoral College favors Democrats, they should note that they, too, benefit from a distortion the Electoral College creates.
Critics of the Electoral College have long argued that it violates the spirit of one person, one vote, and they are right. Americans may believe that a Wyoming or North Dakota voter should have no more say in electing a president than a California or New York voter. But the Electoral College undermines that principle by giving rural states, which often lean Republican, an outsize role in electing our president. Because the Electoral College allocation for each state is based on the state's number of representatives in Congress as well as its two senators, states with fewer residents have more electors per person than larger states do.
Thus the 15 least-populated states get one elector for every 190,000 to 400,000 residents, whereas California gets one per almost 700,000. (This inequality wouldn't change significantly if Electoral College votes were based solely on the number of citizens in each state.)
These same 15 states combined have less than half the population of California, but together they cast one more Electoral College vote.
Perhaps the two Electoral College distortions -- the small- or rural-state advantage that often helps Republicans and the noncitizen boost that works for Democrats -- neutralize each other. Regardless, this is no way to elect a president.
The Electoral College is an anachronism. We should abandon this 18th-century relic and entrust the choice of our president to all citizens, through a direct, popular vote.
Leonard Steinhorn is a professor of communication at American University specializing in American politics and history. He wrote this article for the Washington Post.