Ranting to /dev/null

Divorce Statistics

UPDATE: I realized overnight that the rate I computed for my results might best be expressed as a ratio, not a difference. I'm going to evaluate this and get back later. Might take a few days, this being right before a holiday and all.

I was reading another New York Times Magazine article, this one about how states' rights might be a new liberal talking point. He's talking about how liberals' sudden realization that they're in the minority in the nation as a whole should not discount the fact that they make up a majority in certain areas and states. I didn't get very far into it, because the following paragraph jumped out at me:

"Marriage affords a vivid example. In some states it is evidently more imperiled than in others. The Bible Belt states, in particular, have a shockingly high divorce rate, around 50 percent above the national average. Given such marital instability, these states are anxious to defend the institution of heterosexual matrimony, which may explain their hostility to gay marriage. The state of Massachusetts, by contrast, has the lowest divorce rate in the nation. So its people -- or at least its liberal judges -- perhaps feel more comfortable allowing some progressive experimentation. It will be interesting to see how this experiment plays out, assuming the Bush administration does not succeed in choking off the right of a state to recognize same-sex marriages by getting the Federal Marriage Amendment enacted."

Okay, that sounded fishy to me. Red states really have a higher divorce rate than Blue states? Hmm. So I did some fishing. I figured that the census might be a good place to start, and I was right. The relevant info can be found here (pdf) on the last page of the document.

Hmm. Looks like Massachusetts, the most liberal state in the nation, has one of the lowest divorce rates per 1000 people. And hey, Texas is a lot higher than that. But don't let those numbers fool you. It turns out they're actually really misleading.

First, take a look at Nevada. I have one word for you: "Vegas". So we can completely ignore that one. I'm tempted to do the same for Hawaii, but it's at least in the same order of magnitude as the rest of the country, so we'll leave it be.

But it's important to notice that the marriage rate in Massachusetts is also really low. So it would make sense that not many people are getting divorced if not many people are getting married. Granted, the people who got divorced in a given year are probably mostly not the people who got married that year, but throwing in average marriage-duration would make this really complex, and I don't want to go there, so we'll just take it as a wild rule of thumb.

I went and put the data in a nice little Excel spreadsheet that you can look at if you like. What I did was subtract the divorce rate from the marriage rate to try and get a rough idea of how many people are getting divorced compared to how many people are getting married. I then set that data into a graph sorted by this new rate. Turns out that Massachusetts only does so-so. Texas is actually doing pretty well in the nation when you look at that stat.

It also turns out that there is no discernable Red/Blue pattern here. The top five most "faithful" states are, in order, Hawaii, Tennessee, Arkansas, Utah, and Vermont. That's split pretty evenly. But as you go farther down the list, most of the Blue states are in the bottom half, at least in the 2001 chart.

Notes about the chart. I've not included Nevada, as it makes no statistical sense. The number of people per capita who get married is at least 9-10 times as large as anywhere else. Again: Vegas. Also, some states don't include divorce stats, so I can't do anything with that. Furthermore, it seems that getting married period is a Red-state type thing to do. Which says to me that fooling around and hooking up is more common in liberal/poor areas. Which is intuitively correct, to me at least.

This, it would seem, would be a far better way of supporting the thesis of the NYT magazine article. People in Blue states are far less likely to be married, and thus probably don't care as much about the institution of marriage.

And if you're wondering why I took the time to do this, let's just say that learning to think of varying reaction times in terms of graphs makes you want to see graphs for other data sets too.

Posted by ryan at November 21, 2004 10:22 PM | TrackBack