The Biggest Predictors of Infidelity
Infidelity usually says more about the relationship than the person.
Posted November 8, 2021 | Reviewed by Devon Frye
- A new study used a machine-learning algorithm to determine what does (and doesn't) predict infidelity.
- Whereas demographic and personality factors were weak and inconsistent predictors, relationship factors were much stronger.
- Low sexual and relationship satisfaction, high sexual desire, and lack of love are the most robust predictors of infidelity.
- The gender gap in infidelity appears to be shrinking.
Infidelity is common. Saying exactly how many people have ever cheated is a little dicey, though, because the numbers vary a lot depending on how you ask the question (e.g., sexual vs. emotional infidelity, online vs. in-person infidelity, etc.). However, if you look specifically at sexual infidelity in marital relationships, the number is pretty reliably around 1 in 4 or 1 in 5; by contrast, if you look at young adults in dating relationships, it’s more likely 1 in 2 or 1 in 3.
Infidelity is one of the most common reasons for divorce, and it’s also one of the most common things that prompts people to seek sex and relationship therapy. As a result, there’s a lot of interest in understanding whether infidelity is predictable—because if you can identify the key predictors, you can potentially develop strategies for reducing infidelity and the devastating effects that often accompany it.
A recent set of studies published in the Journal of Sex Research used a machine-learning algorithm to determine whether infidelity is predictable and also what the biggest predictors are.
One study involved a survey of 891 adults who were asked questions about their experiences with in-person sexual infidelity as well as online sexual infidelity. The second study involved a survey of 202 mixed-sex couples who were given the same infidelity questions. In both studies, the researchers collected extensive information about people’s sex lives, relationships, demographic backgrounds, and personalities.
In answering the question of whether infidelity is predictable, the researchers’ conclusion was: “somewhat.” They found some factors that were weak predictors and others that were strong predictors, and that it was only by adding a large number of variables together to the algorithm that they could predict infidelity moderately well.
The Biggest Predictors of Infidelity
In terms of which variables were the most predictive, the researchers found that “the most robust predictors of infidelity lie within the relationship.” In other words, the demographic (e.g., education level) and personality factors (e.g., attachment style) weren’t really explaining much.
Even gender was, at best, an inconsistent predictor in this research. Being male was a strong predictor of online infidelity in the couple’s study only. The fact that gender wasn’t among the main predictors overall suggests that the historical gender gap in infidelity may be decreasing (although it’s not entirely clear whether women are cheating more now, or if they’re just more likely to report it than they were in the past).
The most consistent predictors across samples and across types of infidelity (in-person vs. online) tended to be characteristics of people’s sex lives and relationships. Those who were more likely to cheat tended to:
- Be less satisfied with their relationships overall,
- Have lower levels of sexual satisfaction specifically,
- Have higher levels of sexual desire in general,
- Report less love for one’s partner.
People’s sexual attitudes and behaviors were also predictive, pointing to an increased propensity for cheating among those with more liberal sexual attitudes, as well as those who had engaged in a wider range of sexual behaviors before (such as having had anal sex or having used sex toys). In other words, those who view sex through a more restrictive lens appear less inclined to cheat, but that may be because they hold more negative views toward infidelity in the first place.
Infidelity is Complex and Does Not Have Just One Cause
An interesting finding was that, while being less satisfied with one’s relationship predicted greater odds of having cheated, there was a subsample of highly satisfied folks who had cheated. As I’ve written before, infidelity isn’t always driven by being in an unhappy relationship or having bad (or no) sex—sometimes it’s about something else entirely.
What all of this tells us is that infidelity is complex—and it doesn’t have just one possible cause. So when we’re trying to predict infidelity, we can’t just point to one thing and assume that cheating is a foregone conclusion. It’s likely the result of multiple things working together in tandem.
For example, if you’re unhappy with your relationship and this is coupled with high sexual desire and a permissive view of sex, the odds of infidelity will be quite a bit higher than, say, if you’re unhappy with your relationship but have low sexual desire and a restrictive view of sex. For this reason, we can’t just look at these factors in isolation.
These results also suggest that when it comes to preventing infidelity, addressing sexual needs/desires and relationship issues early on is likely to be the most fruitful strategy because infidelity usually says more about the relationship than anything else.
Vowels, L. M., Vowels, M. J., & Mark, K. P. (2021). Is Infidelity Predictable? Using Explainable Machine Learning to Identify the Most Important Predictors of Infidelity. The Journal of Sex Research.