If you're in a top-10-ish US undergrad, there is a playbook which still gets you a good shot at top-10 econ PhD programs straight out of undergrad
(I think it's extremely unfair that this essentially only works for top US programs, but, info is info)
1. Major in math, or any major that lets you take hard math classes. Definitely take real analysis, and if possible a couple higher level classes (e.g. measure theory, stochastics, functional analysis, etc.)
2. Skip most of undergrad econ. Take a few classes in the PhD first year and get A's
3. RA for econ faculty, starting from around 2nd or 3rd year
4. Aim to have all this done by end of 3rd year/start of 4th year
At the end of this, you have good math grades, econ PhD first year grades which are a strong signal, and a combination of recs from the profs you RA'd for and from your classes. Which is basically checking most of the boxes
I'm not 100% sure this is still enough, but it was back when I applied. I'm putting it out there because it's a bit counterintuitive and I think not well known even at these schools, from students I've talked to
In particular, it's weird that the optimal strat involves basically skipping undergrad econ, but in my experience the signal value of undergrad econ classes is low enough that, if the goal is purely maximize admissions %, there's not much point in taking them
Is this hard/impossible/requires very very strong background coming from high school? Yes, ideally you know multivariable calc going into college to do this. Point is though, if you do, it is I think still possible to skip the 2-year predoc, which is a massive time sink
Is it risky? Yes, if you do it and don't get good grades, it can make it harder to get in even after a pre-doc. Caveat emptor.
(i.e. a B or lower on real analysis, econ 1st year classes, for top programs is arguably as bad/worse than just not taking the class, IMO)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
There is a huge demand to non-fungible-ize things. Own a mass-produced watch, pass it down a few generations, and it becomes a family heirloom - different from every other identical watch coming off the same factory line. A physical NFT
Bottle up some grape juice, hire some famous artists to slap a label on it and it's worth thousands (wine folks, don't cancel me)
Quote. Applies for giving feedback on papers also. I find the most useful feedback comes from, put yourself in the author's shoes. What is something they _could realistically do_, which _they might also be interested in doing_, that they may have missed?
I think here are some "mistakes":
- Wishful thinking: wouldn't it be great if you used microdata instead of aggregate data (yes but I don't have it), wouldn't it be great if you had a perfect instrument (don't have it)
Fun fact: there is a financing trick called a "seller carryback", or just "seller financing". Where you buy a house/business/etc., with money borrowed from the house seller. Pretty popular in real estate and small business world, apparently. Here's how it works.
Suppose you see a small business, say, this car wash. It makes $80k a year. You think you could run it more efficiently, to produce say $120k a year. You want to buy it from the owner for say $2mil. However you don't have $2mil, you need financing
Suppose you can't find a bank to lend you money to buy the car wash. You can instead buy the car wash with money borrowed from the car wash owner. That is, you pay the owner $2mil, but you immediately borrow $1.5mil from them, that you pay off at, say, 4% annual interest or so
Building on @ben_golub's about a field coffee-chat Slack group: what if we had Slack (or other chat) groups organized around commonly used datasets? e.g. a demographic data Slack (ACS, CPS, Census, PSID, etc.), or banking data (call reports, HMDA), etc
Goal being, essentially, to build a slightly more modern version of statalist. Or basically an econ data stackexchange, in chat form. It seems to me that Q&A-style social platforms are relatively easy to get going, compared to pure social platforms like Clubhouse
cc @arpitrage from the other thread. To be honest, a lot of the documentation issues might be better solved with a stackexchange than a big manual. Nobody reads big manuals, but everyone googles questions
Random thought: there seems to be "hidden curriculum" stuff in many lines of work. Academia, of course, but also: how to schmooze your way into a tech internship/offer, how to get onto good projects, be visible, and get promoted quickly in tech,
What specific combination of extracurriculars gets you into Harvard without buying the fencing coach a house or winning the IMO, etc
My parents think this is a fairly "hidden curriculum-heavy" era. Perhaps they are biased by their experience, growing up in post-cultural-revolution China. Exams were newly re-established, basically meritocratic and without many loopholes. Maybe most eras look like our era