earlgraytay:

osberend:

earlgraytay:

another thing you need to keep in mind with ‘status’ talk is cui bono

if a woman, to take a very pointed example, technically ‘benefits’ from not being drafted into a war, but is not being drafted because she’s ‘supposed’ to stay home and make babies and does not have the chance to enlist even if she wanted to… she’s not really the one benefiting.  she’s just getting screwed over in a different way.

Is an expectation to “stay home and make babies” actually getting screwed over? (In and of itself, and in general; obviously, there are some particular women for whom it is.) I’m dubious, but not convinced the other way either. The fact of your making this post suggests pretty strongly that you think the answer is “yes.”

Well, even if so, is it getting screwed over at a level comparable to getting drafted? (Taking into account, of course, that one who hastens to volunteer is not going to still be a civilian when conscription starts.) Here, I’d hope that the answer of “no” would be relatively uncontroversial.

If not, what percentage of the male population has to be drafted to make the expectations equal, or at least comparable?

I was raised in a church so cartoonishly patriarchial it makes radfem nightmares sound tame, and I’ve got a draft card sitting in my wallet right now.  And the expectation to be a mother/housewife was much, much worse.  

First: the expectations are, when you look at it, not as dissimilar as you might think. An unwilling mother and a drafted soldier are both expected to spend a large portion of the ‘best’ years of their lives protecting other people, with no real regard for their own desires. They are fed a romanticised, bullshit view of what their new role looks like. But it’s a decision the rest of the wold makes without any regard to what they want- if you’re in a place where you can’t use birth control or get an abortion, the decision to get pregnant is literally as random as a lottery pull.  

Soldiers are more likely to die… but pregnancy and its complications can kill you in all kinds of fun, horrible ways. The death rate for drafted soldiers in Vietnam was about .002%. The current maternal mortality rate is about .0001%, with the full benefit of modern technology. Without that technology, depending on the country, it’s more like 3-9%.   

A good chunk of Vietnam veterans never saw combat at all- somewhere from 50-75%, depending on the source; only 5-10% of soldiers, drafted or otherwise, were ever intended to be in combat at any time. ‘An army marches on its stomach’, and a good chunk of the work in any army is keeping its stomach full. Boredom, loneliness, discouragement, and isolation were bigger problems for most Vietnam vets than being shot at. And interestingly enough, these are the same kinds of complaints young mothers have, especially if they’re also housewives. 

Am I saying that being a mother, willingly or otherwise, is equal to being a combat soldier in a war zone? Hell no. But being the mother of a toddler is probably equal, in both physical and emotional intensity, to peeling potatoes and fixing the wi-fi in an army base somewhere, which is what most drafted soldiers would be doing. 

But there are a couple things that make being shoved into motherhood, IMO, worse than being shoved into the army.

Difference #1: duration. Most soldiers in Vietnam did 1-4 tours of duty, and each tour was a year long. Most mothers have active care of their child for at least 18 years; if you have more than one child, it can be much longer. My mom’s family has nine kids; mom’s youngest sister was ten when I was born. And even after your kid grows up, you’re still expected to play a major role in their life.

Motherhood really is a lifetime commitment. And if it’s one you did not sign up for, do not want, and can’t egally escape… Do you really think that that’s a negligible problem? Do you think a commitment for one year, even if it’s one that carries an increased risk of dying, is worse than a commitment for the rest of your life?

Difference #2: societal expectation. 
I don’t think anyone expects you to be happy about being drafted unless your father is a drill sergeant from an 80s Teen Movie. But imagine they did. Imagine getting drafted is supposed to be the happiest day of your life - if you’re foolish or scared enough not to enlist- and getting a deferment due to poor health is a tragedy.

Imagine that you do have a medical condition that keeps you from being in the army, whether or not it was something you wanted. Imagine your family trying to hide their disappointment from you. Imagine random strangers treating you like you’re evidence of society degrading, because you’re not playing Your Role As A Man- or worse, giving you unwanted advice about dodgy ‘natural’ cures that will totally let you go out to the front. Because that’s definitely what you want, right? 

Imagine that you make it through school and start looking for jobs, and no one will give you one because you’re a young man of military age. After all, you could go off to war at any time, right? Why would they want to invest in someone so unreliable? Imagine that this is illegal, and yet they keep doing it, because it makes financial sense. 

Imagine that once you do get a job, you’re overqualified, underpaid, and overworked. Imagine that your coworkers take a huge interest in your personal life and keep asking you why you’re not in the army, are you in the Reserves, my cousin is in Reserves and he says- and so on, and no matter how much you try to shut them down, they keep at it. 

Imagine having to do the complicated signaling dance of making sure people know you’re not some kind of filthy awful pacifist, but you’re not going to run off and leave your employer high and dry.  Imagine being passed up for promotions in favour of women and older men, because your bosses still think you might go off to war any day now.

Imagine that even as you get older, and you pass out of the ‘military age’ bracket, people start trying to console you for not having served, trying to think up ways you could still take on the ~role of a soldier~, or start trying to pigeonhole you into some political box or another based on your lack of service. 

Imagine how all of that feels. Even if you were pro-draft, it’d make you a fucking pacifist, wouldn’t it? 

Well, that’s the way that our society treats women who can’t be mothers. 

Difference #3, tying into #2: scale.  

Here’s the thing: if you insist that there is one correct role for a specific group of people, that that one correct role is the only thing that will make those people happy, and that if they stray from that role they are either broken and sad or broken and dangerous… a lot of people are going to fall through the gaps. This is true no matter what that group is and no matter what that role is-  and the bigger the group and the more confining the role, the worse it’ll be.

If you insist that anyone with a vagina should be A Wife And Mother, regardless of how fit they are for it, how much they want to do it, or how happy they’d be… even if most people-with-vaginas are happy to be homemaking mothers, they’re still 51% of the human race.  A tiny minority of half the human race is still a huge number of people. 

There are about 323,000,000 people in the USA as of 2016. About half of those people are women, and about half of those women are of childbearing age- between age 15 and age 40. About 8 million women total are the right age to bear children. About 10% of women are infertile; so in total, about 7.9 million women in the USA can get pregnant/have kids. If even just 1% of those women don’t want to have kids, that’s still 79,000 people in the USA who are being forced into a life they don’t want- roughly the population of Santa Fe.  And, while research on this is shaky, it seems like the actual number is more like 20%.

That means there are 1.5 million women who do not want children; a little more than the population of New Orleans. And a sizeable proportion of those women are being pushed, none too gently, towards motherhood anyway, whether by restrictions on abortion, lack of proper sex ed, lack of proper parenting ed, or just plain old social pressure.

As of my research, the various Army branches take men between the ages of 18 and 34. There are about 4.3 million men in the USA in that age range. About 20% of those men have one or another disability that would keep them from serving- so about 4.2 million men could serve in the army to begin with. 

AFAIK during the Vietnam war,  somewhere from 7-9% of eligible men were drafted. That means, if the US reinstituted the draft today, about 400,000 men total would be in danger of getting drafted. 

In terms of sheer scale, “women being pushed towards motherhood that they don’t want” is a bigger problem than “men being forced into army service they don’t want”, even if the draft was active right now. And… well, it isn’t. The draft is so politically toxic that it’d be difficult to get people to consider reinstituting it. As of today, I’m in no danger of having to go to the draft office, and neither are you.  

 I get it. It’s creepy and dehumanising to have to carry a card around that basically says “I am government property”. I think the draft is morally wrong. I’m a pacifist; I do not think the US should be fighting half the wars it’s trying to fight to begin with. If I wasn’t disabled, I’d be a conscientious objector. I’m annoyed and occasionally saddened that a good chunk of the Left seems to care more about who’s directing what blockbuster or who’s speaking on what campus than about what fucked-up shit the government is doing, and the forever war/symbolic draft/military surveillance are all part of that. 

I don’t think any percentage of men (or women) should be drafted at all. But I think ‘women being pushed into motherhood they do not want and are not ready for’ is a great big awful screw-over of equal proportion to the draft. The only thing that differs it is that as of right now there are no legal consequences for not being a mother, but hoo boy, do certain conservative lawmakers want to change that. 

TLDR: you have no idea what the social pressure to be a mother looks like from the inside; in socially conservative areas, it is every bit as bad as the draft and it affects way more people. 

(via earlgraytay)

jakke:

tableclothcape:

nostalgebraist:

We survey 159 empirical economics literatures that draw upon 64,076 estimates of economic parameters reported in more than 6,700 empirical studies. Half of the research areas have nearly 90% of their results under-powered. The median statistical power is 18%, or less. A simple weighted average of those reported results that are adequately powered (power ≥ 80%) reveals that nearly 80% of the reported effects in these empirical economics literatures are exaggerated; typically, by a factor of two and with one-third inflated by a factor of four or more.

wow

(h/t Andrew Gelman)

@jakke, thoughts?

First off, this is an incredibly solid piece of research and the authors here have taken empirical micro incredibly literally and that’s great. One of these authors has (full disclosure) yelled at me at a conference about this kind of issue in the past and it was a really salient point. I’m honestly thrilled that journals are sufficiently transparent that this kind of enormous meta-analysis is feasible.

That being said, I feel like to some extent this isn’t the most important question most readers are asking when they read a paper? Most important questions that have been studied by several different researchers have reached very different conclusions. If you look at different papers on, e.g., the price pass-through of minimum wages, very careful and rigourous estimates vary by an order of magnitude. Even if the authors are claiming somewhat stronger confidence in their point estimates than would be warranted by a strict interpretation of the econometric technique the within-study confidence interval is swamped by the across-study confidence interval. And that’s not a bad thing - different studies are often trying to get at the same underlying question in very different ways, and most readers of empirical micro literature put more weight on a consensus across studies than laser-tight confidence intervals.

The reason different estimates can look so different is primarily that studies are trying to get at causal identification in very different ways. I’d suggest that plausible causal identification is even more important than the correct p-values, especially if the same study can explore multiple different sources of identifying variation and still come up with pretty much the same answer. This is a little similar to astronomers using a whole bunch of very different methods to measure the distance to objects outside our galaxy; any one method is noisy because you’re working with messy real-world data but collectively they all give similar answers which builds our confidence in the consensus. This is a bit tricky for some economists to talk about, because it’s essentially about rhetorical persuasiveness, but even the most rigourously controlled physics experiments rely on persuading a skeptical reader.

Anyway TL;DR this is a solid study but also “is this statistically significant?” is not the most important question in the vast majority of economic inquiry.

Apologies if I’m misunderstanding, but isn’t the issue here the systematic bias in the estimates (“type M errors” in Gelman’s terminology), and wouldn’t that still be a problem when they are aggregated together into a consensus?

The problem with low power is not just that it makes your confidence intervals wide.  When combined with publication bias, it also leads to systematic overestimation: the only estimates of the effect available are in studies that got significance, which (because the power is low) will be the ones where the noise pushed you unusually far away from the null.  If you read off those point estimates as estimates of the true effect, they’ll be too big (since you only get significance when the noise makes the effect look larger, not smaller).  Gelman’s picture:

And the authors of the above paper indeed found these kinds of overestimates:

To understand the practical consequences of accounting for power in economic research better, it might be instructive to consider some specific areas. For example, of the 1,474 reported estimates of the employment elasticity of a US minimum wage increase (Doucouliagos and Stanley, 2009), 96% are underpowered and the median power is 8.5%. The weighted average elasticity of the 60 adequately powered estimates is 0.0113, less than one-tenth (6.5%) of the reported average (0.19) across all of these 1,474 estimates. As a second example, consider the 39 estimates of the value of a statistical life (Doucouliagos et al., 2012), 74% are underpowered. The WAAP estimate of the 10 adequately powered studies is $1.47 million compared to the simple average of $9.5 million across all 39. Of the 110 reported price elasticities of residential water demand (Dalhuisen et al., 2003), 84% are underpowered. The weighted average elasticity of the 10 estimates that have adequate power is 0.1025, while the average across all 110 is 0.378.

There may be ways of doing causal identification that get the right answer even when these kinds of numbers are wrong (I don’t know), but even then, we do care about (some of) these numbers on their own.  Lots of ink has been spilled about the employment elasticity of the minimum wage; clearly people care about what this number is, and even if estimates vary widely, people will still want to extract an aggregate from the literature.  If the literature is systematically biased, such an aggregate will also be.

(via jakke)

Me: “the Power Nine” sounds like a name for the Supreme Court

Me: … okay, this is not very promising even as time-wasting diversions go, but – for each Supreme court justice, which card are they in the Power Nine, and why?

We survey 159 empirical economics literatures that draw upon 64,076 estimates of economic parameters reported in more than 6,700 empirical studies. Half of the research areas have nearly 90% of their results under-powered. The median statistical power is 18%, or less. A simple weighted average of those reported results that are adequately powered (power ≥ 80%) reveals that nearly 80% of the reported effects in these empirical economics literatures are exaggerated; typically, by a factor of two and with one-third inflated by a factor of four or more.

wow

(h/t Andrew Gelman)