There is a theory that the increased use of online dating, and the accompanying ability to filter mates by characteristics like political affiliation, will lead to increased polarization. This is a silly theory.

The idea stems in part from this study conducted by researchers from Brown, the University of Miami, and Penn State. While we haven't seen the full study (it is expensive) the abstract offers a peek. "Using a sample of Internet dating profiles we find that both liberals and conservatives seek to date individuals who are like themselves. This result suggests a pathway by which long-term couples come to share political preferences, which in turn could be fueling the widening ideological gap in the United States."

Pacific Standard extrapolates.

People who use Internet dating sites are choosing who to date based on criteria that are highly correlated with political preferences … . As a result, the study suggests, there may be long-term consequences for political polarization: not only are such couples more likely to move to the ideological extremes because they lack access to contradictory opinions, they also are likely to produce children who hold ideologically extreme positions. The end result is a more polarized America where more and more people cannot understand how others could possibly think differently from themselves.

"The next time you see a bumper sticker that says, 'He’s not my President,'" it suggests, "you may want to ask the person if they met their spouse online." (We do not recommend this.) Here are the problems with this theory.

1. American politics are already on a long trend of polarization.

Here, via Polarized America, is a graph of the increased polarization of the House and Senate over time.

Big increase, right? Now, click the image to see the timeframe we're talking about. In other words, polarization has been on a sharp incline since the end of World War II. Pacific Standard points to a study suggesting that one-third of marriages now begin online. It's not clear what that means specifically — over email? Craigslist? — but it is clear that the figure has grown sharply of late. Particularly since the advent of the commercial internet, which happened right about where it says "House" on that second graph.

2. Geography is already a powerful political sorter.

Geography is almost certainly a stronger indicator of the likelihood two people will get together than having matching political views on OKCupid. A resident of Boston may find the perfect match possible on a website, but if that match likes in Honolulu, he's much more likely to end up with someone from Cambridge.

Geographical regions already tend to have a lot of political homogeneity. Take New York City. An April 2012 survey found that 82 percent of the city was registered Democratic, as at right. So there's a four-to-one chance that the person you meet randomly on the street will share your registration if you're a Democrat. If you try to find someone who disagrees with you, it will be difficult.

Oh, and then there's this study from the University of Iowa which shows that "partisans relocate based on destination characteristics such as racial composition, income, and population density, but additionally prefer to relocate in areas populated with copartisans." In other words, people move to areas that share their politics. As that study's author notes, "it has the potential to make important imprints on the political landscape of the United States."

3. People have sought ideologically compatible partners for a long, long time.

An informal survey of the offices of The Atlantic Wire found one (1) person whose parents didn't already share the same political ideology. One. None of our parents met online, for perhaps obvious reasons. (We are not all 13 years old.) Asked how his parents met, if not online, the answer was simple: geography. "They attended neighboring high schools," that staffer told us, "and were introduced by mutual friends after college."

4. People's politics change.

And then he continued. "It's likely that my father was more liberal at the time, though I wasn't around to witness it."

Voter registration patterns change. Older people are far more likely to be Republican at this point, but that's the same population of people that in the 1960s was mostly liberal. Issues change and people change and politics change. And if we're assuming that marriage lasts for decades, the idea that the people who marry are exactly the same 40 years later seems at best optimistic.

America's polarization problem is not a function of our online matchmaking. Not now, and probably not in the future. So please don't hesitate to filter your PlentyOfFish search by politics (if that's an option on that site). It's far more important that you can stand one another over the first three dates than that Congress be slightly less divisive by the time you have grandkids.