Gender Roles; Weak Men, Strong Women And Data Interpretation

You might have read this article from Matthew Moore at The Telegraph; Men more likely to cheat if they earn less than their wives. It's... erm... an odd read that not so much angered as brought on a bout of pity (which is rare).

The study Moore mentions was conducted by Christin L. Munsch as part of her ph.D research at Cornell University. Yes; 'her'. A woman conducted this data collecting research (Moore calls it a 'scientific study'). She believes that men, especially in certain sub-groups of society, still hold on tightly to old traditional ideas of gender roles.

To be fair; she's probably right. A lot of men still do think they should be the 'breadwinners' (I wish I could ban that term), as Matthew Moore proves by being one of them;
The secret to a loyal and lasting relationship is for women to earn 25 per cent less than their husbands, the researchers established.
Yeah. That's totally the secret to a long-term relationship going well; forcing one half of the team to be crap. I can't actually find the report, so I'll have to take his word on it that Munsch 'established' this as a factor of having a lasting relationship. More likely, Munsch simply found that men who earn 25% more than their female partners cheated less than men who earned less... but that does not in anyway say that women should strive to earn less than their husbands to keep them. In fact, if a man cannot stand his partner earning more than him, then she deserves better and he deserves to be dumped, sharpish.

What Moore seems to be saying with this entire article is that if a man has an affair, it's not his fault. He's doing it because his wife earns more than him and is threatening his masculinity. That might be a reason, but it sure as Hell is not an excuse. It's not even a good reason. What kind of pathetic piece of trash has an affair because his wife is good at what she does? Not because he found a prettier woman. Not because she has less time for him. Not because he can. Not even because he has some weirdass fetish she won't help him with. Moore (and possibly Munsch) are saying that these men are having affairs because they're trapped in some warped kind of cave-man mentality that makes them think 'their woman' should be lower than them.

I say warped because, really, any good cave-man wouldn't be caring who brought home dinner, as long as he got dinner.


And how much more likely IS a man to have an affair if his partner earns more than him? 10% more likely? 40%? 90? Moore doesn't say...

The data found is probably a true-ish reflection on society (at least American) in that men who earn less than their wives might very well be having more affairs than their higher earning counterparts. However, it is not as black and white as Munsch SEEMS to be painting it (again, I don't have access to her actual report) and a whole lot more sepia than Moore thinks. The contributing factors to a man (or a woman) having an affair are much more complex than "erg, woman bring home more money... I get new sexual partner to feel man again" (which could be interpreted in a completely different way).

If a person is working long hours in a hard, hectic job then their partner may feel neglected. If a person has just lost their job or recently had to change to a lower paying job they may be unhappy with life in general. If both parties are out all hours, working their butts off then someone might just find comfort in someone else's arms. If a person is spending a lot of time just hanging out with friends rather than working, they might find someone else more interesting than their current partner.

This list goes on and on and on... I'm not saying that any of these are good reasons for an affair, but they are more likely to be the contributing factors in people using someone else's body parts for fun, other than their better half's than the very simple 'breadwinner' issue.

Being (or not being) the breadwinner isn't even the whole story though. Moore brushes over the next important bit;
Men who earn significantly more than their wives are also more likely than average to take a mistress...
Basically; some men are likely to cheat on their partners. Well there's a shocker. And what exactly is this average? Men who earn about the same as their partners? But we keep being told they don't exist!

Eugh and poop!

What people should be taking from this report is that a) America is still a long way from equality (and yes, I know, the UK's not there yet either but we're a bit closer :P ); b) some men are truly pathetic (but not all, not even most of them, before you all get carried away with being offended); c) we need to work on the idea of 'gender roles'.

We (certain chunks of Western society) have worked hard to let women know that they don't have to be defined by 'traditional' ideas of gender. They CAN have high-flying careers. They CAN enjoy their sexuality. They CAN be rough and tumble. They don't HAVE to want children. Now we need to get it through men's heads that they don't have to be all "grr-arg". They CAN look after the family unit. They CAN be sexually quiet (best term I could think for it... got a better one?). They CAN be soft and tender. They don't HAVE to be afraid of changing nappies (or something).

You know what? Let's just get rid of the idea of 'gender roles' all together. You with me?

[EDIT]

Was emailed a response from a person who doesn't want their comments being linked back to their real world persona as they work in a similar field (and once on the internet, nothing is ever removed)... but I liked the comments so asked permission to post as an edit.

I have to emphasise that these are just speculations based on the abstract. This person does not have access to the report and is merely working off the abstract and Moore's article in the Telegraph.

***************
Some things Moore elides from his report (going from the abstract alone): The focus is on Latino men who apparently place higher value in 'traditional masculinity' - Ok so it is for a subgroup of the population, and Moore is generalising that to the wider population - On dodgy ground there Matthew.

I also like the nebulous "although this relationship disappears once individual and institutional mechanisms are controlled". - So it's all about money, until you control for confounds...then it's not...

And:

"Also, the more economically dependent a man’s female partner is on him, the more likely he is to engage in infidelity. This relationship is mediated by relationship satisfaction." - So its all about money until you control for...happiness in the relationship....

And finally:

"Finally, I find that the more times per week an individual attends a religious service the less likely he or she is to cheat, the more education one reports the less likely he or she is to engage in infidelity, and the more satisfied one reports being in his or her relationship, the less likely he or she is to engage in infidelity." - The issue with all of this stuff is that is simple correlation, there is no identifiable causation anywhere her. You can run the stats on each and every one of these factors that she identifies and come up with a story much like the one Moore is running with, but you never actually know which one is driving the effect. Is it religion, or money, education maybe? You just can't tell from a single source of data like this. You need a series of data sets which you can whack in an anova (analysis of variance) or do lme (linear mixed effects modeling) on and pull out the factors which best predict the outcomes. This is why social studies based on corpora are often slated for being reflections of the author's beliefs, rather than actual, objective, analysis.

Oh and by the way: A little googling and it becomes apparent that the 1997 National Longitudinal Survey of Youth (NLSY97) is a self-report survey of 18-28 year olds. So another methodological (self-report is lame for many, many reasons), and subgroup issue (another one that is particularly unlikely to generalise to the population).

Munsch, Christin L. 2010. “The Effect of Unemployment and Relative Income Disparity on Infidelity for Men and Women.” Unpublished Manuscript. Cornell University, Ithaca, NY. To be presented on August 16, 2010 in Atlanta, GA at the American Sociological Association’s 105th Annual Meeting.

This paper investigates the factors related to infidelity for both men and women and uses social identity theory to develop an interactional model of male infidelity. I argue that, for men, making less money than a female partner may threaten men’s gender identity by calling into question the traditional notion of men as breadwinners, and that this relationship may be particularly strong for certain subgroups of the population that highly value traditional masculinity like Latino men. Using panel data from the 1997 National Longitudinal Survey of Youth (NLSY97), I find evidence in favor of the interactional theory proposed. For men, the more economically dependent one is on a female partner, the more likely he is to engage in infidelity, although this relationship disappears once individual and institutional mechanisms are controlled. Also, the more economically dependent a man’s female partner is on him, the more likely he is to engage in infidelity. This relationship is mediated by relationship satisfaction. For women, economic dependency seems to have the opposite effect: the more dependent women are on their male partners, the less likely they are to engage in infidelity, net of the individual and institutional controls. For Hispanic men, being economically dependent on a female partner dramatically increases the likelihood that one will engage in infidelity. Finally, I find that the more times per week an individual attends a religious service the less likely he or she is to cheat, the more education one reports the less likely he or she is to engage in infidelity, and the more satisfied one reports being in his or her relationship, the less likely he or she is to engage in infidelity.


Can't say much more without reading the article, but it certainly feels like an overextension from data to assumption. Maybe part of that is Moore's fault, maybe Munsch's - who knows.

***************
blog comments powered by Disqus