Textbook suggestions

[phdeconomics]

I know there has been a couple of past posts about this, but I thought I would add another. What advanced undergraduate textbooks would you guys suggest for econometrics and linear algebra for independent study?

 

My econometrics class used Stock and Watson, and my linear algebra class used Linear Algebra with Applications, by Leon.

[cjyNel]

If you are going to do PhD economics after graduation, you might want to read the beginning chapters of Hayashi. Though it is a graduate-level book, it is readable for undergraduates with good understanding of the material in a typical prob-stat sequence. Alternatively, Greene might also be a good choice.

 

For linear algebra, I once read Leon's book several years ago and felt that it was a good one. You may also want to try Friedberg's classic text on linear algebra. That one contains a bit more "theory" than Leon's and is very rigorous. I heard that Linear Algebra Done Right by Axler is also very good, but I have not read that.

[phdeconomics]

Thanks for the suggestions--someone else suggested Linear Algebra Done Right as well, so I might try to take a look at that. I am doing an RA position before grad school, and it's been a couple semesters since I took econometrics, so I was hoping to get an econometrics book to refresh my memory before I start the RA job. I would just reread Stock and Watson, but I really didn't enjoy that textbook very much so I was thinking of trying something else. I'd like something at more of an advanced undergrad level, just because I'm taking a full class load my last semester and I won't have a ton of time to dedicate to the book.

[tm_member]

If it's an empirically-oriented RA position buy and read Mastering Metrics and Mostly Harmless Econometrics by Angrist and Pischke. Formal textbooks are good for grad classes in theoretical econometrics. Doing empirical work people will appreciate is so much more than knowing the theory.

 

 

Thanks for the suggestions--someone else suggested Linear Algebra Done Right as well, so I might try to take a look at that. I am doing an RA position before grad school, and it's been a couple semesters since I took econometrics, so I was hoping to get an econometrics book to refresh my memory before I start the RA job. I would just reread Stock and Watson, but I really didn't enjoy that textbook very much so I was thinking of trying something else. I'd like something at more of an advanced undergrad level, just because I'm taking a full class load my last semester and I won't have a ton of time to dedicate to the book.

[phdeconomics]

It is an empiricially-oriented RA position, so thanks for those suggestions!

[chateauheart]

There's probably a corollary to what TM was suggesting: if you're going to spend time to prepare for this job, then 50--75% of it should be spent on programming.

 

Or let me re-iterate: your value-added as an RA comes from being able to code. Econometric theory and research design are both marginally relevant. These other skills might be useful if you really want to shine as an RA, but as a first priority, you should always make sure you can code really quickly.

 

If you don't know as of now, it may be a good idea to ask for what language(s) you'd be coding in, then practice a lot with online resources.

[tm_member]

There's probably a corollary to what TM was suggesting: if you're going to spend time to prepare for this job, then 50--75% of it should be spent on programming.

 

Or let me re-iterate: your value-added as an RA comes from being able to code. Econometric theory and research design are both marginally relevant. These other skills might be useful if you really want to shine as an RA, but as a first priority, you should always make sure you can code really quickly.

 

If you don't know as of now, it may be a good idea to ask for what language(s) you'd be coding in, then practice a lot with online resources.

 

That's definitely good advice but I assumed the ability to write code was a given for the position and if not, they intend to teach it or give the student time. The hardest thing about having a new RA for me (I've been one and now have had two) was that they don't speak the language of empirical micro. Being familiar with common research designs and their associated/common robustness tests is crucial (difference-in-difference, regression discontinuity, event study, IV, spatial). For the majority of empirical research papers, someone else's approach fits like a glove. That means that being able to say "take a look at the data from those two states, see if using Card and Krueger's minimum wage DinD approach gives us anything interesting" or "that's a triple diff, like Gruber (1993)" and have the person know what is you are talking about is valuable for both the researcher and the RA.

[Spectrum]

That's definitely good advice but I assumed the ability to write code was a given for the position and if not, they intend to teach it or give the student time. The hardest thing about having a new RA for me (I've been one and now have had two) was that they don't speak the language of empirical micro. Being familiar with common research designs and their associated/common robustness tests is crucial (difference-in-difference, regression discontinuity, event study, IV, spatial). For the majority of empirical research papers, someone else's approach fits like a glove. That means that being able to say "take a look at the data from those two states, see if using Card and Krueger's minimum wage DinD approach gives us anything interesting" or "that's a triple diff, like Gruber (1993)" and have the person know what is you are talking about is valuable for both the researcher and the RA.

So the most important thing for landing a good RA position is to have a letter of recommendation that says... what? That this student is well read in the most influential modern empirical work?

[Kaysa]

So the most important thing for landing a good RA position is to have a letter of recommendation that says... what? That this student is well read in the most influential modern empirical work?

That they can hit the ground running on day one.

[tm_member]

That they can hit the ground running on day one.

 

Exactly.

[chateauheart]

Work ethic is also hugely important for any kind of RA work. If I have to guess, "this student had a B- midterm but received an A- by the end of the course" is probably more valuable than "This student received an A on both exams."

[mathenomics]

most RA job postings also explicitly state the qualities they are looking for in their candidates. Research experience, strong programming skills, rigorous coursework, able to work independently, motivated, aiming to become a professional economist, etc. It seems like these would all be important for rec letters to RA positions

[phdeconomics]

Spectrum -- for me, the places I applied to had some sort of programming data task. If you did well enough on that they'd request an interview. For me, it seemed like the most important component of the interview was being able to talk about my previous research experience; why I did something a certain way, why I used the data I did, what problems I ran into and how I figured it out--stuff like that. This was just my experience, but being able to talk clearly about research (like your thesis or some other RA experience), strong programming skills, and obviously good grades in things like econometrics and statistics should put you in a good position for landing a good RA job.