Aptitude for Econ content versus expertise in tools

[Econtobe]

I've been reading the criteria for admissions for quite sometime now and I can't help but wonder (rather, muse!) on these lines..

I view mathematics and statistics as just tools to crunch economic data.. If top PhD programs only admit people who are expected to be math/stats specialists it means they are only intent on creating economics craftsmen.. ones who are experts in using the tools; not necessarily the ones who have the potential to become thought-leaders in economics content, isn't it? Please pardon my naivete in thinking this way.

Or, is there an implicit assumption that everybody who are intent on pursuing an economics PhD should already have the 'aptitude' and 'potential' for econ thought leadership and their math/stat prowess only increases the probability of making an Adam Smith or Ricardo out of them?

Also, can't an economics researcher hire a lot of quantitative RAs from math and statistics departments to get his tooling done, while he focuses on analysis?

Historically, political scientists like Chanakya or economists like Smith may not have been as much mathematical jocks as contemporary PhDs in these fields.. Maybe the adcoms think that all thought-leadership required in economics has already been done.. but this is hard to believe when so much disparities exist in the world.

I am sure there is a big hole in my logic & thoughts.. and the forum would quickly point it out..

[econphilomath]

Its a lot easier to spot someone who will not be able to pass prelims because of a lack in math skills than it is to spot a brilliant future "thought-leader". Thus adcoms look well on students who show potential in economics AND have solid quantitative skills.

Another point you raised about hiring quantitative RAs....where do you think professors get their RAs from? Yup from their students.... and they are on the adcoms picking which students they want to have...

That being said, I do think math is over emphasized here on TM but then again math is not bad per se.

[Econtobe]

Hmm.. The basic premise seems to be that adcoms are very risk averse. For some reason this leads me to believe that the next path-breaking super-star economist is going to come from a mid-range school..

Although I understand the risk present in picking someone who might not pass prelims, I find it tough to digest the fact that we've to take so much math just to make adcoms decision easier. Life doesn't seem to be easy.

I was talking about using math/statistics RAs from 'other' respective departments not econ PhD/MA students. Of course I have zero-knowledge about how this entire RA system works, hence pardon all the ignorance.

[YoungEconomist]

I find it tough to digest the fact that we've to take so much math just to make adcoms decision easier.

Well, it's not just about that. Modern economic theory is highly mathematical and statistical, so in order to even pass the core classes one needs to have a certain level of profeciency in math/stat. Don't forget that they are interested in training people who can do top notch economics research which currently takes pretty advanced math/stat skills (at least in comparison to other social sciences). Maybe this will change in the future, and maybe economics will become less mathematical again, who knows? But the point is that currently it is heavily mathematical, and if you want to be an economist then you need to learn this stuff.

Besides, the more philosophical aspects of economics that you're talking about may still be advanced by someone who's highly technical. Some economists have both aspects down really well. Or maybe it will come from someone who is in a heterodox branch of econ. Or maybe they won't be an economist by training. F.A. Hayek was a great economic thinker, yet he was more of a philosopher than he was an economist.

[polkaparty]

Also, can't an economics researcher hire a lot of quantitative RAs from math and statistics departments to get his tooling done, while he focuses on analysis?

No they can’t. Analysis in economics is doing a lot of quantitative work. If you’re a theorist then you’re going to be doing a lot of mathematical theory. If you’re an empiricist then you’re going to do a lot of statistics. The only tasks which are reasonable to outsource are data preparation. When it comes to graduate RAs, they might do more meaningful work (I don’t know given that I’m not in a grad program yet), but RAs are like apprentices. Hence you give them more work to help them learn, so it makes sense.

Now, if you’re doing some hardcore theory in an obscure field of math, you might want to co-author with a mathematician who’s a specialist in that field. Math gets so specific that two people from different fields probably can't understand what each other are doing.

I view mathematics and statistics as just tools to crunch economic data.. If top PhD programs only admit people who are expected to be math/stats specialists it means they are only intent on creating economics craftsmen.. ones who are experts in using the tools; not necessarily the ones who have the potential to become thought-leaders in economics content, isn't it? Please pardon my naivete in thinking this way.

Or, is there an implicit assumption that everybody who are intent on pursuing an economics PhD should already have the 'aptitude' and 'potential' for econ thought leadership and their math/stat prowess only increases the probability of making an Adam Smith or Ricardo out of them?

This is what LORs are for: demonstrating that you have the exceptional ability to do independent research. Top programs do not admit people just because they can integrate everything under the sun.

Hmm.. The basic premise seems to be that adcoms are very risk averse. For some reason this leads me to believe that the next path-breaking super-star economist is going to come from a mid-range school.

If you define risk averse by not accepting people who haven’t had anything more than calc 1 and haven’t demonstrated an ability to do research, then yes they are risk averse. I think it’s obvious that top schools look over some amazing candidates, and hence those people will emerge from lower ranked schools and perhaps become superstars, but…everyone they do admit is still an amazing candidate by most measures! Hence the odds are stacked in the top schools’ favor.


Finally, one note about math in econ:
I have somewhat come full circle regarding this. Consider the concept of marginal quantities--a core economic idea. It is clearly a quantitative concept, but at its simplest level it is nothing more than measurement. The big leap comes when we begin to model marginal quantities using derivatives. When I first learned this I was astonished. Why, what I used to know is actually just this object in math. I thought that a marginal quantity is aderivative. It's not. The derivative is simply used to model marginal quantities. By introducing derivatives we need a function and hence some sort of space and the derivative has to be defined, etc. All of this is extra baggage on top of the original concept.

Anyway, I didn't intend on starting a huge meta rambling, but I just wanted to note as YoungEconomist said, math is just a way of thinking about the world, a way of thinking about economics. There are those who think we can study economics without it (the Austrians in particular), but right now it's the mainstream approach, so if you want to get a mainstream education you need to know some math.

[econphilomath]

No they can’t. Analysis in economics is doing a lot of quantitative work. If you’re a theorist then you’re going to be doing a lot of mathematical theory. If you’re an empiricist then you’re going to do a lot of statistics. The only tasks which are reasonable to outsource are data preparation. When it comes to graduate RAs, they might do more meaningful work (I don’t know given that I’m not in a grad program yet), but RAs are like apprentices. Hence you give them more work to help them learn, so it makes sense.

Now, if you’re doing some hardcore theory in an obscure field of math, you might want to co-author with a mathematician who’s a specialist in that field. Math gets so specific that two people from different fields probably can't understand what each other are doing.




This is what LORs are for: demonstrating that you have the exceptional ability to do independent research. Top programs do not admit people just because they can integrate everything under the sun.




If you define risk averse by not accepting people who haven’t had anything more than calc 1 and haven’t demonstrated an ability to do research, then yes they are risk averse. I think it’s obvious that top schools look over some amazing candidates, and hence those people will emerge from lower ranked schools and perhaps become superstars, but…everyone they do admit is still an amazing candidate by most measures! Hence the odds are stacked in the top schools’ favor.


Finally, one note about math in econ:
I have somewhat come full circle regarding this. Consider the concept of marginal quantities--a core economic idea. It is clearly a quantitative concept, but at its simplest level it is nothing more than measurement. The big leap comes when we begin to model marginal quantities using derivatives. When I first learned this I was astonished. Why, what I used to know is actually just this object in math. I thought that a marginal quantity is aderivative. It's not. The derivative is simply used to model marginal quantities. By introducing derivatives we need a function and hence some sort of space and the derivative has to be defined, etc. All of this is extra baggage on top of the original concept.

Anyway, I didn't intend on starting a huge meta rambling, but I just wanted to note as YoungEconomist said, math is just a way of thinking about the world, a way of thinking about economics. There are those who think we can study economics without it (the Austrians in particular), but right now it's the mainstream approach, so if you want to get a mainstream education you need to know some math.

Amen!

[Mr.Keen]

I agree with my fellow TM-ers here. Mainstream economics, for better or worse (I'd like to think for better), has become highly mathematical. That such a transformation may preclude top departments from discovering the next Adam Smith or the next Ricardo is possible. However, you must bear in mind that with high probability the new Smiths and Ricardos have a much higher mathematical training, or have at least the ability to acquire it. Mathematics is not a tool to crunch numbers, it is a tool to concisely communicate ideas and develop logical arguments. It is like a technological advance that has detractors and followers, and those who refuse to adopt it, or simply can't, are doomed to stay behind. You can read the Journal of Economic Perspectives and find great thinkers writing about economics without using a single equation. But behind those articles there is a vast body of literature built around mathematically rigurous models. Call us, average econophiles, bad writers if you will.

[Econtobe]

Wow, I'm amazed and impressed by the reasoning behind using so much math. Continuing musings..

Mathematics is not a tool to crunch numbers, it is a tool to concisely communicate ideas and develop logical arguments.

Math is indeed a fantastic language for this purpose.. If its a medium then I would imagine other social sciences having difficulty in communicating complicated logic should use it as well. For instance philosophy, does anyone know if math has penetrated mainstream philosophy too? I've read works of Bertrand Russell but not sure of contemporary research there..

Moreover, so much math really makes it difficult for a layman like me to understand econ research. I remember I started reading Welfare Economics by Amartya Sen and was amazed to see it completely math oriented and inscrutable.. unfortunately had to put it down.

[Mr.Keen]

I remember I started reading Welfare Economics by Amartya Sen and was amazed to see it completely math oriented and inscrutable.. unfortunately had to put it down.

Then again, Sen is one of the best thinkers in economics that can also communicate complex ideas without resorting to math. The linkages between math and philosophy can be traced back to Aristotle if I am not mistaken. Bertrand Russell, undoubtedly, provides an example of very ambitious and complex thinking where it is very hard to distinguish between math and philosophy.

[buckykatt]

However, you must bear in mind that with high probability the new Smiths and Ricardos have a much higher mathematical training, or have at least the ability to acquire it.

Indeed. It's easy to forget how recent a development the calculus is, especially as a mature subject with rigorous foundations. To put things in perspective, consider that Smith was born while Newton was still alive, and that Ricardo was born before both Bolzano and Cauchy. Probability theory is an even younger subject...