A preprint on aRxiv, The Strength of Absent Ties: Social Integration via Online Dating, purports to explain the increased rate of interracial/ethnic marriage in the United States as a consequence of online dating. They have ways to control for the fact that the proportion of non-whites in the United States has been increasing over the same time period. It diminishes the effect based on their model that online dating results in more interracial marriage, but it does not abolish the effect. They observe that interracial marriages seem to have increased around 1995, when the internet began, 2006, when services like OKCupid became very popular, and in 2014 when Tindr became a phenomenon.
I don’t know about these dates and the impacts of these services on the census data. Rather, the key and more interesting findings are that many more Americans now marry people who are total strangers outside of their social networks.* Previously more people tended to marry people who they were loosely connected to. Not close friends, but perhaps acquaintances, or friends of friends. With ~30% of marriages being attributable to online dating, people who are totally unconnected are now marrying, and so binding together two very distinct networks (in theory). This is an important dynamic to observe and note.
On the whole, I’m pessimistic about the United States. These results make me optimistic.
* Though to be fair, among a certain set it has always been common to marry people you meet at university. In this case, you are likely from two different social networks entirely. This is the case for me, and most of my friends from what I can tell. So the effect must be more impactful lower down the socioeconomic index.