Several groups use this, for varied reasons:
Atheists: “The divorce rate among Christians are as high or higher as among unbelievers. Atheism is morally superior.”
Egalitarians: “Complementarian teaching messes marriages up so badly, that Christian divorce rates are equal to, or even worse than, the secular world.”
Gender hierarchists: “If more Christians understood submission and headship, then the divorce rate amongst Christians wouldn’t be equal with or higher than that of the heathen world.”
To show the fallacy of the assumption, here is a little story:
100 professional racing drivers were asked if they crashed any car in the past year. 76 of them did, and several crashed more than once. 100 people who never had a driver’s license was asked the same question. Only one of them did.
Should we conclude that the people without licenses are better drivers? No, of course not. Racing drivers crash more, because they drive much more – on busy tracks where everyone tries to overtake – than the unlicensed.
Similarly, Christians divorce about as much as anyone else, but they marry more. A study done by ARIS in 2001, for example, contained among others this data:
By this study, there are, among the non-religious, 2.11 whose marriage is intact for every one who is divorced. Among Baptists, there are 4.83 whose marriage lasts for every one who is divorced. Christians who did not say what their denomination was (the rest of the statistic page contains many denominations) are 6.22 times as likely to be married as to be divorced.
(Even this data is incomplete, as it don’t differentiate between those who keep their first spouse and the remarried.)
Other surveys by ARIS show less pronounced differences in the marriage rate than this one, but in general, the conclusion holds: The only way atheists could be said to divorce less, is when divorce is taken as a percentage of the general population of atheists and believers (car crashers as a percentage of all drivers and all non-drivers) and not as a percentage of the married (not as a percentage of actual time spent driving).
(The idea that Christians actually divorce significantly more than unbelievers is usually based on a single study by George Barna in 1999, while more recent Barna studies and larger (and more recent) ARIS studies came to the conclusion that the divorce rate do not differ as much. The 1999 Barna study never reported how many were married in each faith group, but the ARIS studies do.)