My first AEA conference

Jan 13 JDN 2458497

The last couple of weeks have been a bit of a whirlwind for me. I submitted a grant proposal, I have another, much more complicated proposal due next week, I submitted a paper to a journal, and somewhere in there I went to the AEA conference for the first time.

Going to the conference made it quite clear that the race and gender disparities in economics are quite real: The vast majority of the attendees were middle-aged White males, all wearing one of either two outfits: Sportcoat and khakis, or suit and tie. (And almost all of the suits were grey or black and almost all of the shirts were white or pastel. Had you photographed in greyscale you’d only notice because the hotel carpets looked wrong.) In an upcoming post I’ll go into more detail about this problem, what seems to be causing it, and what might be done to fix it.

But for now I just want to talk about the conference itself, and moreover, the idea of having conferences—is this really the best way to organize ourselves as a profession?

One thing I really do like about the AEA conference is actually something that separates it from other professions: The job market for economics PhDs is a very formalized matching system designed to be efficient and minimize opportunities for bias. It should be a model for other job markets. All the interviews are conducted in rapid succession, at the conference itself, so that candidates can interview for positions all over the country or even abroad.

I wasn’t on the job market yet, but I will be in a few years. I wanted to see what it’s like before I have to run that gauntlet myself.

But then again, why did we need face-to-face interviews at all? What do they actually tell us?

It honestly seems like a face-to-face interview is optimized to maximize opportunities for discrimination. Do you know them personally? Nepotism opportunity. Are they male or female? Sexism opportunity. Are they in good health? Ableism opportunity. Do they seem gay, or mention a same-sex partner? Homophobia opportunity. Is their gender expression normative? Transphobia opportunity. How old are they? Ageism opportunity. Are they White? Racism opportunity. Do they have an accent? Nationalism opportunity. Do they wear fancy clothes? Classism opportunity. There are other forms of bias we don’t even have simple names for: Do they look pregnant? Do they wear a wedding band? Are they physically attractive? Are they tall?

You can construct your resume review system to not include any of this information, by excluding names, pictures, and personal information. But you literally can’t exclude all of this information from a face-to-face interview, and this is the only hiring mechanism that suffers from this fundamental flaw.

If it were really about proving your ability to do the job, they could send you a take-home exam (a lot of tech companies actually do this): Here’s a small sample project similar to what we want you to do, and a reasonable deadline in which to do it. Do it, and we’ll see if it’s good enough.

If they want to offer an opportunity for you to ask or answer specific questions, that could be done via text chat—which could be on the one hand end-to-end encrypted against eavesdropping and on the other hand leave a clear paper trail in case they try to ask you anything they shouldn’t. If they start asking about your sexual interests in the digital interview, you don’t just feel awkward and wonder if you should take the job: You have something to show in court.

Even if they’re interested in things like your social skills and presentation style, those aren’t measured well by interviews anyway. And they probably shouldn’t even be as relevant to hiring as they are.

With that in mind, maybe bringing all the PhD graduates in economics in the entire United States into one hotel for three days isn’t actually necessary. Maybe all these face-to-face interviews aren’t actually all that great, because their small potential benefits are outweighed by their enormous potential biases.

The rest of the conference is more like other academic conferences, which seems even less useful.

The conference format seems like a strange sort of formality, a ritual that we go through. It’s clearly not the optimal way to present ongoing research—though perhaps it’s better than publishing papers in journals, which is our current gold standard. A whole bunch of different people give you brief, superficial presentations of their research, which may be only tangentially related to anything you’re interested in, and you barely even have time to think about it before they go on to the next once. Also, seven of these sessions are going on simultaneously, so unless you have a Time Turner, you have to choose which one to go to. And they are often changed at the last minute, so you may not even end up going to the one you thought you were going to.

I was really struck by how little experimental work was presented. I was under the impression that experimental economics was catching on, but despite specifically trying to go to experiment-related sessions (excluding the 8:00 AM session for migraine reasons), I only counted a handful of experiments, most of them in the field rather than the lab. There was a huge amount of theory and applied econometrics. I guess this isn’t too surprising, as those are the two main kinds of research that only cost a researcher’s time. I guess in some sense this is good news for me: It means I don’t have as much competition as I thought.

Instead of gathering papers into sessions where five different people present vaguely-related papers in far too little time, we could use working papers, or better yet a more sophisticated online forum where research could be discussed in real-time before it even gets written into a paper. We could post results as soon as we get them, and instead of conducting one high-stakes anonymous peer review at the time of publication, conduct dozens of little low-stakes peer reviews as the research is ongoing. Discussants could be turned into collaborators.

The most valuable parts of conferences always seem to be the parts that aren’t official sessions: Luncheons, receptions, mixers. There you get to meet other people in the field. And this can be valuable, to be sure. But I fear that the individual gain is far larger than the social gain: Most of the real benefits of networking get dissipated by the competition to be better-connected than the other candidates. The kind of working relationships that seem to be genuinely valuable are the kind formed by working at the same school for several years, not the kind that can be forged by meeting once at a conference reception.

I guess every relationship has to start somewhere, and perhaps more collaborations have started that way than I realize. But it’s also worth asking: Should we really be putting so much weight on relationships? Is that the best way to organize an academic discipline?

“It’s not what you know, it’s who you know” is an accurate adage in many professions, but it seems like research should be where we would want it least to apply. This is supposed to be about advancing human knowledge, not making friends—and certainly not maintaining the old boys’ club.

Government shutdowns are pure waste

Jan 6 JDN 2458490
At the time of writing, the US federal government is still shut down.

The US government has been shut down in this way 22 times—all of them since 1976. Most countries don’t do this. The US didn’t do it for most of our history. Please keep that in mind: This was an entirely avoidable outcome that most countries never go through.

The consequences of a government shutdown are pure waste on an enormous scale. Most government employees get furloughed without pay, which means they miss their credit card and mortgage payments while they wait for their back pay after the shutdown ends. (And this one happened during Christmas!) Contractors have it even worse: They get their contracts terminated and may never see the money they were promised. This has effects on our whole economy; the 2013 shutdown removed a full $24 billion from the US economy, and the current shutdown is expected to drain $6 billion per week. The government itself is taking losses of about $1 billion per week, mostly in the form of unpaid and unaudited taxes.

I personally don’t know what’s going to happen to an NSF grant proposal I’ve been writing for several weeks: Almost the entire NSF has been furloughed as “non-essential” (most of the military remains operative; almost all basic science gets completely shut down—insert comment about the military-industrial complex here), and in 2013 some of the dissertation grants were outright canceled because of the shutdown.

Why do these shutdowns happen?

A government shutdown occurs when the omnibus appropriations bill fails to pass. This bill is essentially the entire US federal budget in a single bill; like any other bill, it has to be passed by both houses of Congress and signed by the President.

For some reason, our government decided that if this process doesn’t happen on schedule, the correct answer is to shut down all non-essential government services. This is a frankly idiotic answer. The obviously correct solution is that if Congress and the President can’t agree on a new budget, the old budget gets renewed in its entirety with a standard COLA inflation adjustment. This really seems incredibly basic: If the government can’t agree on how to change something, the status quo should remain in effect until they do. And the status quo is an inflation-adjusted version of the existing budget.

This particular shutdown occurred because of Donald Trump’s brinksmanship on the border wall: He demanded at least $5 billion, and the House wouldn’t give it to him.

It won’t be much longer before we’ve already lost more money on the shutdown than that $5 billion; this may tempt you to say that the House should give in. But the wall won’t actually do anything to make our nation safer or better, and building it would displace thousands of people by eminent domain and send an unquestionable signal of xenophobia to the rest of the world. Frankly it sickens me that there were not enough principled Republicans to stand their ground against Trump’s madness; but at least there are now Democrats standing theirs.

Make no mistake: This is Trump’s shutdown, and he said so himself. The House even offered to do what should be done by default, which is renew the old budget while negotiations on the border wall continue—Trump refused this offer. And Trump keeps changing his story with every new tweet.

But the real problem is that this is even something the President is allowed to do. Vetoing the old budget should restore the old budget, not furlough hundreds of thousands of workers and undermine government services. This is a ludicrous way to organize a government, and seems practically designed to make our government as inefficient, wasteful, and hated as possible. This was an absolutely unforced error and we should be enacting policy rules that would prevent it from ever happening again.