CovOps
 Location : Ether-Sphere
Job/hobbies : Irrationality Exterminator
Humor : Über Serious
 |  Subject: 4 Dr. Stats: No, Sex Won't Make You Rich  Thu Aug 22, 2013 1:56 am | |
| Don't believe the study that says otherwise
I have some rather, well, deflating news for you. If your plan for getting a raise at work is getting lucky at home, you need a new plan. More sex won't mean more money -- unless you're getting paid for it -- no matter what an economist (or your partner) tells you. Sorry.
Now, it's a simple -- and tantalizing -- enough idea. Healthy people tend to earn more, and people who have more sex tend to be healthier. So could having more sex make you earn more? Well, as Lydia DePillis of Wonkblog points out, that's the question Nick Drydakis, a senior lecturer in economics at Anglia Ruskin University, tried to answer by looking at Greek survey data from 2008.
But that answer isn't much of an answer. Now, the study did find a positive, and statistically significant, relationship between how much sex people between the ages of 26 and 50 say they have, and how much they say they make. And it control for things like age, health, education, marital status, sexual orientation, ethnicity, industry, and personality type. But even controlling for these things, a correlation between sex and wages doesn't tell us anything. It's just a correlation. We don't know which way -- if at all -- the causation runs. It might be that having more sex really does make people make more. Or it might be that making more makes people less stressed -- and that makes them have more sex. Or it might be that something we aren't observing is causing both.
In other words, we need to figure out if this is just a casual relationship or something more -- a causal one.
This is trickier than it sounds, because it's impossible to tell just by looking at sex and wages. The problem is we can't disentangle sex from wages. There just isn't enough information. We need something that shows us that a change in sexual activity changes earnings. We need what economists call an "instrumental variable."
Here's how instrumental variables work. Say we want to figure out if smoking causes health problems. Now, we know that smokers tend to have more health problems, but that doesn't prove that smoking is why they do. It's just a correlation. But what if we look at cigarette taxes instead of smoking itself. All else equal, higher cigarette taxes should mean fewer cigarettes sold -- which, if smoking really is bad for people, should mean better health. Well, lo and behold, higher cigarette taxes are indeed correlated with both less smoking and fewer health problems. And that tells us what's causing what. See, the magic here is there is nothing to disentangle. Cigarette taxes don't help or hurt people's health, except for how they change how much people smoke -- and they do change how much people smoke. So cigarette taxes show us how a change in smoking causes a change in health outcomes. Cigarette taxes are the perfect instrumental variable.
For sex and wages, we need another instrumental variable. Think of something that only changes how much money people make, because it changes how much sex they have? It's not attractiveness. Good-looking people tend to earn more, because they're more productive. And it's not age. Older people tend to earn more, because they're in leadership positions. So what else could it be? Well, Drydakis looks at whether people believe in God or attend at least two religious services a month instead. The idea is that religious people tend to have less sex -- and that having less sex is the only way that being religious affects what they make. But, of course, wanting religiosity to be an instrumental variable doesn't make it one. We need proof. We need to see that there really is a strong relationship between religiosity and sexual activity, but no direct relationship between religiosity and wages. And, believe it or not, that's what Drydakis found when he ran the numbers.
Don't believe it. And here's why you shouldn't, in two sentences. It's reasonable to say, like we did above, that cigarette taxes make us healthier only because they make people smoke less. But it's absurd to say that being religious makes us poorer only because it makes us have less sex. For one, being religious probably tends to make us richer! And, for another, it probably makes us richer for reasons that have nothing to do with sex. Think about it. Church-goers probably have more social capital than others. That means they have a ready community to help them find work or clients when they need to. And going to church probably socializes them to have more of what psychologists call "conscientiousness" -- that is, to be more motivated and hard-working. Religiosity might affect wages in either or both of these ways -- or entirely different ones. But there is zero chance that it only affects wages through how much sex people have.
So why did the paper say otherwise? Well, there are lies, damn lies, and statistics -- and then there are statistics about Greece in 2008. Maybe its vast political patronage system muted the impact of other kinds of social capital. Or maybe is sclerotic labor markets, which keep young people underemployed (if that) and living at home, decreased conscientiousness. Or, more likely, maybe the data is just bad. The ugly truth is that picking an instrumental variable can involve a fair amount of hand-waving -- and this couldn't even step over that low bar.
It isn't clear how religion changes what we earn. But what is clear is that it doesn't have much, if anything, to do with sex. To revert to stats-speak, religiosity isn't a good instrumental variable for sexual activity and wages. And if it's not a good instrumental variable, then all of these numbers don't tell us anything about whether having more sex is making people make more money. It's still just a correlation.
And that's not sexy.
http://www.theatlantic.com/business/archive/2013/08/no-sex-wont-make-you-rich/278818/?google_editors_picks=true |
|
RR Phantom
Location : Wasted Space
Job/hobbies : Cayman Islands Actuary
| | Subject: Re: 4 Dr. Stats: No, Sex Won't Make You Rich Thu Aug 22, 2013 1:59 am | |
|  |
|
