Are Most Couples Really Meeting Online?
Exposing another meme graph
It has become commonplace to hear that ‘most people meet online now’, confidently asserted as if it’s not even up for debate. If you ask for a source, you’ll get this updated version of a previously famous graph—which I’ll add has been officially noticed by the current richest man on earth.
Personally, viral ‘line go up’ graphs like this immediately set off alarm bells. People might see the ’n=6,591’ and assume that the data must be fairly reliable. This is somewhat misleading, as this includes people who met as far back as 1950. What we’re really interested in are the sample sizes for those who met in more recent years. When we focus on these, the numbers are quite unimpressive, but especially after 2017, as these were follow-up surveys of the same respondents—many of whom were either no longer panel members or didn’t respond. For the years driving the big spike, the sample sizes aren’t even enough to fill a classroom. Additionally, these years were of course marked by a pandemic which limited people’s offline interactions, meaning that of the couples which formed during this period we should probably expect a higher than usual proportion to have met online.
This smoothed line also conceals considerable noise beneath it: after 2017 it dips to 23-24% in the next two surveys before a major surge of almost 200% to 66% in 2020, followed by a slight drop to 58% in 2021.
Reports of online dating’s rise have been greatly exaggerated
The first thing one should do when confronted by a viral chart is, if possible, check if other sources corroborate the data. Sometimes people will point to a recent Pew survey which showed that just 10% of people met their current partner on an online dating site or app (rising to 20% for 18-29s). The problem with stats like this is that they doesn’t distinguish between people who met their partners recently and those who met decades ago, while the How Couples Meet and Stay Together (hereinafter ‘HCMST’) data points reflect how many people met online in each respective year.
Another dataset I was able to get my hands on is an American Perspectives Survey, conducted by The Survey Center on American Life in August of 2022, which like the HCMST used a nationally representative panel. Overall, a similar percentage to that seen in the Pew survey had met their partner through online dating (11%). More importantly, this survey also contains a question about not just how but when people met their partners, allowing me to restrict the sample to those who met in recent years for a more appropriate comparison with the HCMST data.
Among those who met in the past year, 26% had met through an online dating site or app. Although this is far from a majority, it did represent a plurality, with the next highest category being through friends or family at 23%. This sample contained 112 people, which is 3x more than the combined sample size of the people who met in 2020-21 in the HCMST.
Of people who met their partners 1-3 years ago (N = 202), 30% met through online dating; of those who met 3-5 years ago (N = 210), 27% did, and among those 5-10 years ago (N = 433), 21% did. The 2017 HCMST had about 40% of heterosexuals meeting their partners online in 2017 with a sample size of 60. In the APS, the respondents who met their partners 3-5 years ago would have met between 2017-2019. Since the sample size in the APS sample was much higher and since in the subsequent two years the percentage dropped to a similar percentage in the HCMST, the APS figures are most likely closer to the truth. Moreover, there is no evidence here for the meteoric rise shown in the most recent HCMST follow-up survey.
You might expect that meeting online may skew young, but when it comes to people who met their partners within the past 5 years, the youngest adults are actually the least likely to have met their partners through online dating. This is likely because it becomes more challenging to meet new people once you’ve left the social bubble of school and begin to drift away from friends, as well as less of your friends and friends of friends still remaining single.
In another American Perspectives Survey from 2020, of people who met their partners within a year ago (N =175), 21.4% met through online dating, and of those who met 1-3 years ago (N = 201), 24.2% met through online dating. Of those who met within the past 3 years, 18.4% of 18-24s, 20.9% of 25-34s, 31.4% of 35-44s, 18.9% of 45-54s, and 32.8% of 55+s met through online dating.
According to the latest Singles in America survey by Match Group, 17.5% of singles had met their most recent date through online dating, and 7.7% through social media, for a total of 25.2%.
In the 2022 survey, 16.7% had met through online dating, and 8.9% through social media, for a total of 25.6%. I also managed to dig up the codebooks for 2016 and 2017. In 2016, 23.9% met through online dating, and 3.5% through social media (27.6% overall). In 2017, 20% met through online dating, and 6.2% through social media (26.2% overall). It’s worth mentioning that quite a few participants were old and hadn’t dated in a long time, so this may cause there to be an underestimate in the percentage of people meeting online, but once again there doesn’t seem to be any evidence for the exponential rise in internet mediated encounters indicated by the HCMST graph.
Another survey of 978 college students conducted in 2023 found that 15% had met their current or most recent partner through a dating app, while 7% met through other online venues, and half met through school or friends. Only 21% had used a dating app in the past month.
A Swiss survey (Potarca, 2020) found that the proportion of heterosexual couples meeting online rose from 0% in 1995 to about 25% in 2017-18, with 10% having met their partners through dating apps, about 7% met through dating websites, and another 7% through other online venues. The N for 2017-18 was 413—much higher than for 2017 in the HCMST—providing additional evidence that its 2017 data point was also an overestimation.
Some boring caveats
One potential concern is that the HCMST figure refers to people who met ‘online’, which is more of an umbrella term than online dating per se, even if it has often been interpreted to mean online dating. This concern may be alleviated somewhat by the fact that only 6% selected ‘other’ in the APS survey for those who met their partner in the past year. For those who met 3-5 years ago, it was 1%. If people met through another online venue and didn’t feel comfortable selecting online dating, presumably they’d have selected this option. It might be suggested that some may have chosen ‘through friends’ if meeting through another online venue such as social media. The HCMST question was open-ended, and sometimes their answers covered multiple categories. Of those in the HCMST who met online in 2017, none of their answers suggested that their online connection was mediated by friends, family, or others, and 97.2% were complete strangers when meeting, so it’s doubtful that many who met online would’ve selected through friends in the APS. Interestingly, about 25% of those who reported meeting through online dating in the APS also reported being either friends or casual acquaintances with their partners before beginning dating. It’s possible that these people met through online dating and got to know each other online before going on a date, but it’s also possible that some met through social media etc. and selected online dating because it’s the closest one that fit. Unfortunately due to the open-ended nature of the HCMST question, it’s not too informative here, as the online venue was often not able to be identified. Again, a Swiss survey with a larger sample size found a significantly smaller percentage meeting in online in 2017, so unless there are cultural differences driving more Americans than Swiss people to meet online, this suggests that the spikes seen in 2017 and most likely 2020-21 are the product of random error.
Final thoughts
The weight of the evidence points to ‘probably not’ as the answer to the question at hand. Definitely not if we’re just talking dating apps. Regardless, it’s far from justified to making sweeping statements based on such a comically small sample size. This just looks like another 2018 GSS or 2022 Pew situation whereby the statistics which become far and away the most widely known and quoted aren’t those which are the most accurate, but which either have the most viral potential or fit a popular narrative. This seems like an especially egregious example, though; the next viral statistic might as well be based on interviews with four homeless albinos. People are hasty to accept information which feeds into their anxieties around new technology taking over society and leading to one crisis or another, and those who like to blame their dating woes on apps find the notion that most people are meeting online validating—it makes them feel more justified in doing so. ‘Dating apps are real life’ has become a common response to people who advise they try meeting people in the ‘real world’ instead of relying on dating apps. They’re certainly a relevant part of the landscape, but it’s clear that plenty of people are still managing to socialize and meet people outside. Just because you’re permanently online doesn’t mean everyone else is. In fact, most women have never even touched a dating app. In the 2022 Pew survey, 62% of men and 46% of women in the 18-29 age bracket had ever used online dating. Even among singles who reported being interested in dating, 67% of men and 38% of women in this age bracket had used online dating in the past year, and the percentage of women currently using it was only 18%. I know a large part of the appeal of dating apps is that they bypass a lot of the social challenges that come with dating, especially as a male who is still expected to make the first move more often than not, but if there was any doubt, socializing is still important.
It seems dating apps may have already peaked and are struggling to appeal to Gen Z. People already complain about bots and fake profiles, and I doubt AI will make things any better. Of course, this doesn’t necessarily mean we should expect them to reverse the trend and mingle more in real life; perhaps social media will increasingly take its place, or people will just date less in general. I’d be surprised if metaverse dating became anything more than furries and neckbeards RPing as women though. Whatever the implications of more people meeting online may be, as I’ve established, the widespread myth of ‘Chad harems’ don’t appear to among them. Findings on the impact of meeting online on relationship outcomes are somewhat mixed, but overall it doesn’t seem like relationships which begin online fare worse (might do an article on some relevant data from this dataset on this), so either way the whole thing is probably just another dumb moral panic.
Citations
Rosenfeld, Michael J., Reuben J. Thomas, and Sonia Hausen. 2023. How Couples Meet and Stay Together 2017-2020-2022 combined dataset. [Computer files]. Stanford, CA: Stanford University Libraries. https://data.stanford.edu/hcmst2017
Potarca G. (2020). The demography of swiping right. An overview of couples who met through dating apps in Switzerland. PloS one, 15(12), e0243733. https://doi.org/10.1371/journal.pone.0243733