When I saw this tweet, I’m ashamed to say that I did not believe it. I clicked through and read the fine print, expecting to find something misinterpreted or at least out of date. But instead I found this fact sheet, which in turn referenced the Association of Universities and Colleges of Canada. The figure is part of a 2007 report on the state of higher education. (There does, at least, seem to be an upward trend.)

It’s funny that I doubted the number, really. There isn’t exactly a surplus of women profs around my department, and thinking back, I wasn’t taught by many during undergrad. While my MA cohort is about 50-50, a classmate recently observed that many of the men and few or none of the women are planning on doing a PhD. She suggested that it is because a PhD would make us unmarriageable. Our classmates, for what it’s worth, disagreed.

Different answers to those questions lead to a different numbers, and that’s one reason we see so many conflicting stats. That’s also why we have some people claiming that the wage gap no longer exists.* So I’d like to ask what I hope will be a clarifying question: What would the end of the wage gap look like?

1. Radical parity

The most popular wage gap statistic is that American women make 77% of men’s earnings. That ratio is calculated based on the median earnings of men and women who work year-round and full-time. This doesn’t account for differences in occupation, education, hours above and beyond the full-time threshold, or any number of other factors. While it corrects for differences in current labour force participation, by excluding the larger proportion of women who work at home, it doesn’t correct for past participation, the reality that among full-time workers more women than men have taken time off in the past.

If this ratio is our indicator, then the wage gap can close in a couple ways: (1) when women working full-time are paid exactly as much as men, even as we continue to work fewer years in total, in occupations that tend to be lower-paid, or (2) when men and women participate in the labour force in exactly the same way, i.e. are equally likely to take time off or cut their hours to raise children, and equally likely to choose any given occupation. I’m not sure this is what most people have in mind when they quote the 77% stat.

2. An end to pure discrimination

Another view holds that the wage gap will be closed when men and women with exactly the same characteristics make the same amount of money. Folks in this camp are only worried about the gap that we can’t attribute to any characteristic other than gender. They are less concerned with the fact that women are more likely to work in lower-paid fields, and take on the labour market penalties of childbearing. This is the gap that tends to be measured by economists, and while it is smaller than 23%, it is non-zero.

3. The right to live like a man

An extreme version of this view is that the wage gap is fine so long as there is some set of choices women can make that allow them to earn as much as men. These people tend to compare the wages of childless men and women, often only young and/or educated, in the same occupations. Don’t like the 77 cent deal you’re being offered? Go to law school, strap on some shoulder pads, golf with the big boys and never, ever marry.

So what is a reasonable finish line? I don’t think #1 is likely, in the short run, though it’s worth pushing back against the restrictive gender roles that got us into this bind. But the weakness of narrower views is that they gloss over the process by which women come to be primary caregivers, the element of coercion in what we tend to characterize as choices.

Plenty of women would spend more time in the workforce if they could find a daycare near work, work more flexible hours, convince their husbands to help out around the house, or be taken seriously by their bosses.** I find #3 especially frustrating, because of the double standard it presents. If men can have kids without making big sacrifices at work, women should be able to do the same. But that’s not really my point. I’m just saying that any measure of the wage gap comes, implicitly or explicitly, with some assumptions about the finish line.

* The other reason, of course, is that people misunderstand or wilfully obscure the truth. Luckily, Echidne is around to set them straight.

** Also, as long as some occupations remain almost entirely segregated by gender we can’t separate the effects of occupation and gender. When we control for segregated occupations, as we do in #2 and #3, we can’t rule out the possibility that nurses for example are paid so little because they are women, not the other way around.

By 1953, the average annual consumption of cigarettes had reached thirty-five hundred per person. On average, an adult American smoked ten cigarettes every day, an average Englishman twelve, and a Scotsman nearly twenty. (241)

It reminds me of a similar passage from another recent nonfiction hit, Daniel Okrent’s Last Call: The Rise and Fall of Prohibition.

By 1830 Americans were guzzling, per capita, a staggering seven gallons of pure alcohol a year. […] In modern terms those seven gallons are equivalent of 1.7 bottles of standard 80-proof liquor per person, per week—nearly 90 bottles a year for every adult in the nation, even with abstainers (and there were millions of them) factored in. Once again figuring per capita, multiply the amount Americans drink today by three and you’ll have an idea of what much of the nineteenth century was like. (8)

So this is what I’m wondering: What current habit will seem as absurd to future readers as these do to us now? Salt or saturated fat consumption? Physical inactivity among office workers? Some food additive or packaging material?

American experiences are seen as general. (A particularly absurd manifestation of this occurs in international relations, where people will write “A superpower like the United States…” There is no superpower like the US. [...])

[...] In medical research, it probably doesn’t matter whether Canadian or American data is used. A medical treatment that works south of the border will generally work north of the border as well. Indeed, drug trials are not-uncommonly carried out in low-income countries to save costs. The generalizability of studies based on US economic data to Canada is less well-established.

I worry that this is especially true in some of the areas I am most interested in, like education. Of course, that’s difficult to verify without more research. It’s kind of a circular problem, really.

If the US is a bad model for other developed economies, what are we likely getting wrong?

I’ve noticed that some media improves when you share it. Actually, I’ve noticed this about television. It doesn’t matter how many times I watch the first episode of the West Wing, say, or Friday Night Lights. If I’m watching it with someone who has never seen it before, I enjoy myself. On my own, I’d be bored.

Also: I am a compulsively social reader. By which I mean, I have to read all the best parts out loud to you. This is one of the things that is great about Twitter, or Facebook, or the “email to a friend” buttons on newspaper websites, right? I can pester you even when you’re out of earshot, and that makes reading more fun.

Speaking of which, via Mathew Ingram, you should read this review of Timothy Ferriss’s new book. Here’s a teaser:

Mr. Ferriss offers advice about so many disparate things — not simply losing weight and building muscle and improving sex and living forever, but learning to hold your breath longer than Houdini (!) and hit baseballs like Babe Ruth (!!) — that paging through “The 4-Hour Body” is like reading the sprawling menu in a dubious diner, quite certain the only thing you’d dare order is the turkey club.

And that’s not even the best part.

And finally, if you’re in Toronto you should know that we now have our very own Hacks and Hackers chapter. I wasn’t able to make either of the first two events because of midterms and then exams, but I will definitely be at the next event, whatever it is.

Lots of smart people have weighed in online. The best follow-up I’ve read so far is from Andrew Gelman, who links to his own piece from American Scientist. (H/t to Chris Blattman.) But as plenty of good has been said about the piece already, I’d like to point out the one thing about it that bugged me.

Early in the piece, Lehrer writes at length about Jonathan Schooler, a psych researcher who has found that he has difficulty reproducing his own results.

…while Schooler was publishing these results in highly reputable journals, a secret worry gnawed at him: it was proving difficult to replicate his earlier findings. “I’d often still see an effect, but the effect just wouldn’t be as strong,” he told me. “It was as if verbal overshadowing, my big new idea, was getting weaker.” At first, he assumed that he’d made an error in experimental design or a statistical miscalculation. But he couldn’t find anything wrong with his research.

The central insight of the piece is that declining statistical significance in a field overall is one of the effects of publication bias. When an idea is new, strong positive results are required for publication. As it becomes accepted knowledge, contradictory results become interesting enough to publish. That’s a brilliant, important observation.

The problem is that Schooler’s experience can’t be explained by publication bias. He is just one researcher. If publication bias was the only force in play, individual researchers wouldn’t see their results change over time, only their chances of getting those results published.

Schooler’s problem is probably better described as regression to the mean following a few anomalous results, as the article acknowledges. (Though it is a pretty weird case, and I’d love to hear a better explanation, especially for his tests of the decline effect itself.) That’s a less important issue than publication bias.It’s not a problem with the scientific method if anomalous results are gradually disproven, but publication bias can twist our perception of the world over the long term.

If the piece is really raising questions with the scientific method, why do we read so much about Schooler? And why do transitions throughout the piece seem to relate his research to the decline effect in fields as a whole? I suspect it’s because even though it isn’t illustrative of the articles’ central point, his story is interesting.

This is the sort of compromise that writers and editors make during revision. Some characters seem too compelling to cut, even as inclusion confuses the point, as I believe it does in this case. But we’re supposed to be servants of the truth, not just great stories, and I expect better of the New Yorker. Maybe it’s an unrealistic standard.

Mr. Lopez, who has designed fitness nutrition plans, recommends options for healthier dining and drinking.

“When Dawn’s at the pub, a turkey burger or sweet potato fries would be better alternatives. On average, one chicken wing has 100 calories, seven grams of fat and 100 grams of sodium.” As for alcohol, “vodka with soda or champagne are her best bets – both are below 100 calories.”

So I’m trying to picture 100 grams of sodium. Sounds like a lot, right? And when you multiply up, the numbers become really silly: There’s a full kilogram of sodium in ten wings? Obviously, it’s supposed to be milligrams. This might look like a typo, and I do think pointing out typos is kind of petty. But this kind of mistake doesn’t surprise me nearly as much as, say, a misspelling of the word “grams” would. And that’s frustrating.

Much more at Regret the Error.

(H/t to a sharp-eyed friend, Mike Markovich.)