MIT to offer Degree in Computer Science + Economics

Designing electronic marketplaces will be the focus of a new degree to be offered jointly by MIT’s economics and computer science departments.

“This area is super-hot commercially,” says David Autor, the Ford Professor of Economics and associate head of the Department of Economics. “Hiring economists has become really prominent at tech companies because they’re filling market-design positions.”

Because these companies need analysts who can decide which objectives to maximize, what information and choices to offer, what rules to set, and so on, “companies are really looking for this skill set,” he says.

UBER, Airbnb and Amazon are familiar marketplaces that need computer scientist cum economist designers and online worlds like EveOnline join multiple marketplaces into entire economies.

I’d also note that an increasing number of marketplaces will need to be designed not for people but for non-human traders–this will create entirely new challenges.


Sounds good. I would do it. At a downmarket state school, of course.


As for the news embedded in that post...

When they provide the same degree online, for $99, then I will be impressed. If 10,000 enroll MIT will gross $0.99 million, while perhaps training 1000 new economists who, for a change, actually understand math. Hopefully, they will grasp the implications of non-linearity and sensitivity to initial conditions, but I digress.

Of course, they won't do that, because the "education" behemoth is all about signalling and artificial scarcity. It has more in common with the soup nazi - the longer the line, and thus the suffering, and the greater the abuse by secure tyrants inside, the more attractive and expensive the product. It can only spiral upward in price.

I can't wait for the education bubble to burst. Compulsory K-12 education is a form of slavery and Orwellian indoctrination, only to be followed by a lifetime of debt to serve the interests of the institutions called "higher" education.

Bring it on!

Back in the day, we just studied, like, math.

Up until recently if you wanted to go into data science you usually had to teach yourself a lot, I know I did. My actual masters was in political economy. I'm not sure if I would have chosen a data science masters instead if they existed at the time, it was a lot of fun having to teach myself data science tools to do applied research, rather than simply learning the tools for tools sake. On the other hand, it was a lot of uncertainty and stress trying to teach myself all this stuff (and I continue) with no formal education.

Can you point to a few sources of learning material that you found particularly useful? What would you avoid if you were to have that time over again?

All the paper asset markets are going fully automated otherwise we end up paying monopoly fees to high frequency traders.

A degree cor everything, so nobody can be locked in their own little channel ( and MIT grads are properly rewarded for, well, being associated with MIT).


Double your cuckold by doing both Econ and CS!

Maybe skeptics should explain how they learned NumPy, or R, and why an integrated casework is unnecessary.

.. I took a few weeks of a data science mooc before canceling it for a family ski trip. It was a taste that interested me, even if for me R is hard.

You need to learn the statistics and methods to understand R otherwise it's pointless.

I think so. I should have had better statistics classes sooner. In my precomputer day, it was hard to something like this.

The family ski trip is worth more than an R class - you can do that online anytime.

I really don't see the point of MOOCs for data science (or most things to be honest). You need a good background in stats and other math to be truely useful. The MOOCs just teach you the tools and maybe a bit of theory but it is not comparable to a 4 year degree.

I believe there is an online MS with a series of classes. Mine was more a survey, with just a few lectures on statistics. Not enough, or perhaps too late, for my old brain.

Ah, a few.

Yeah well I really don't see the point you either learn it and use it or you lose it. Unless you have as pressing need from work or some innate passion for stats I see no value in spending too much energy.

This is for an undergraduate degree. It's not a joint degree but a new major. Other colleges have joint degree programs that typically combine a BA/BS with an MA/MS. Speaking of MIT, I really miss Click and Clack, the Tappet Brothers.

Was going to say the same thing: MBA MIS or analytics course equivalent. A number of graduate business school courses, by the way, are taught by Ph.D economists. A number of business schools that don't have MIS or business analytics departments are co-venturing with the comp sci and or stats departments to create similar interdisciplinary programs.

Good that the econ departments are waking up.

The major does indeed seem to be a good idea (although why couldn't the computer science majors simply take a bunch of econ classes, either as a minor or as what MIT calls a HASS concentration -- or vice-versa, econ majors could simply take a bunch of computer science classes).

This sentence though puzzles me:

"In a poll of the introductory economics course 14.01, which all students are required to take, faculty found that a whopping three-quarters of them were interested in the joint major"

What is it that all students are required to take? The poll? Intro econ? Neither makes sense. Or maybe the 14.01 requirement applies to econ majors (but that pretty much goes without saying).

This makes a difference because one interpretation of that sentence is that a very large percentage of students who take intro econ are interested in the new major. But I'm dubious about that because historically most of the students in 14.01 were pretty much not interested in economics at all -- they were only taking it because it was a way of fulfilling their HASS requirement. Econ was for MIT students their version of the "rocks for jocks" or "physics for poets" classes that at other colleges are used by students to fulfill their science distribution requirements.

Maybe this should be the new econ, but econ is too slow to change?

How could anyone not like this guy: The transcripts of his interviews have become as pleasing as his voice in the interviews.

I wonder whether in the future we will program by simply saying:


Take the numeric data from column one and assign that as the y value, and run a regression using the values from column 2, 3 and 4.

Check for heteroskedacity and multicollinearity.

>>>If so,


Maybe we should be teaching today not just R, but also the more abstract features of statistics and econ because the "day to day" grind stuff will be automated.

How many of you still remember Fortran, but really today use only the logic of a flow chart to set up a program.

Voice to code translators exist =)

Thanks. You can always learn from others.

This is just so obvious to anyone who watched the original Star Trek series in thier formative years.

Spock, is that you? Beam me up.

I am lost somewhere in the internets.

I still use Fortran professionally, but I also still tend to think of a computer as a machine that performs arithmetic for the purpose of approximating solutions to partial differential equations.

Sometimes I darkly suspect that before I die there will only be one field of science left, data science. All the others will have been abandoned. No one will know anything except how to extract curve fits that serve well enough to get by.

I did a minor in econ to go with my EECS Major there, very easy way to satisfy your humanities requirements without having to write too much. If they had a version where you could get out of 6.002 it would be heaven.

Greek finance minister Yanis Varoufakis ran Valve's in-game economies

Sorry Bud -- I think you've been in academia far too long.

We already have a degree for people who "understand the marketplace," and it's called Marketing. It's cryptic, I know.

They know how to use a computer, and they about cost tradeoffs. And there are thousands and thousands and thousands of them already. Right here. Today.

I know you're just hyping your field, and you can be forgiven for that, but come on now.

This seems much more useful for MIT to eventually pitch as precursor to product management roles in Sillicon Valley. Most companies will require CS as a baseline to make sure their PMs can interface well with the engineering team, but the economics background providing good grounding in the empirical analysis and strategy required to make good business decisions.

If I had it to do over again, I'd love doing something like this instead of straight C.S.

Increasingly specialized undergraduate degrees should be thought of as a bad thing. There are already plenty of ways a student can show they have multidisciplinary skills such as double majors, minors, or individual courses. Most students that go into computer science already have programming experience before they get to university, so they should simply major in economics.

The brochure mentions von Neumann and Herbert Simon as role models for the offspring of this mixing of genes. Perhaps. But then there is the story of Bernard Shaw who refused to mix his genes with that of a beautiful dancer: He was worried that her genes were dominant for brains, and his for beauty. What if the graduates inherit economic’s memes for realism and practicality, and computer science’s memes for rhetoric and morals?

I am really loving the theme/design of your site. Do you ever run into any internet browser compatibility problems? A number of my blog readers have complained about my site not operating correctly in Explorer but looks great in Opera. Do you have any ideas to help fix this problem?

Comments for this post are closed