Detour

I can tell by the spike in hits to my home-site that the application process for wanna-be economists hoping to get into a great graduate economics program (with funding…of course) is in full swing. Winslet in Washington DC has a site called Sophomore in the Capitol which is pretty good. Check it out if you want to reminisce about the pain and turmoil of the process (or if you aspire to one day put yourself through the punishment that is an economics PhD).

Winslet has collected some quotes about Econ Math Camp…

At Berkeley:
“Math camp was a horrible experience, but I got through it and have to say that Advanced Calc at AU really did help. I didn’t at all know as much analysis as I should have, but at least I had some idea what a proof was. (I think I would have cried if I had gone in with no analysis background at all.) The macro, micro, and metrics theory courses aren’t incredibly interesting all the time, but they’re probably not as dry as they could be. ”

- my former AU classmate

At Boston U:
“Guys, Math Camp is the BEST. Seriously! It’s so good to be around people who weren’t embarrassed to respond to the question, “What are you doing this summer?” with “Going to math camp.” Or, at least it is good to be around people who were also mocked for it.”

- fomer IIE RA

At Northwestern:
“Math camp, on the other hand, is very hard. The 3hr lectures are followable, but the (daily) problem sets are pretty brutal, and just about nobody is getting the answers before they give them to us for the discussion section (fortunately, there are no grades). The material they are reviewing is supposed to be the math you will need to know for the first year. We have Real Analysis & Topology, Linear Algebra, MV Calc, Optimization, and Probability sections, all 2-3 lectures each. The students are on very different levels math-wise. The ones who seem to be doing the best appear to be those who just graduated undergrad w/ a math major. Those of us who have been out of classes during the past couple years are having to re-learn quite a bit. We have a textbook (Simon & Blume, Mathematics for Economists) but the book is very simple compared to what they are teaching in class. I’m hoping that knowing the stuff in the textbook will get me through the year. ”

- former IIE RA

The University of Virginia has a cheating scandal brewing in the economics graduate department this summer which has implications for every student I have ever met in an economics PhD program. According to Inside Higher Education magazine…

An “alarmingly large fraction” of the first-year class of economics graduate students at the University of Virginia were involved in a cheating incident that came to light this month, according to the department chair.

Department officials said that some problem sets from textbooks used in introductory graduate economics courses have answer keys online. At least one student found answers for a course taken by all first-year students, and apparently shared the information with classmates. Though the solutions were apparently available, David Mills, chair of the economics department, said students should have “known it was off-limits,” but that they instead “used it without the professor being aware.”

I find it incredibly hard to believe that the faculty of any economics graduate department does not know that within a month of arrival most students have access to solution manuals for the books and up to ten years worth of old problem sets and homework assignments.

As my friend, John Morrow, at the University of Wisconsin, Madison puts it…

The current status of graduate programs seems to be that some students, notably those from countries where solution manuals and entire texts are copied as a matter of habit (generally out of necessity) have access to solution manuals. Incentives are clearly not to share them, since grades are often based upon homework covered by manuals. Given that this seems to be the case most places, failing to help distribute manuals known to be in circulation penalizes the just and rewards the wicked….It is also my impression that most students that visit this site, or at least the ones I have corresponded with, are interested in self-study for department examinations, etc. If one is trying to cover 100+ involved problems in a month to prepare for a test, a student without access to a solution manual is at a huge disadvantage to one who has access to a manual.

I remember the frustration of spending hours on a problem, going down the wrong path when a simple check of the first step of a solution would have shown that my simple calculus mistake was the reason I had an illogical answer. I learned fairly quickly that many of the other students did not have the same inefficient time wasting study habits; if they were stumped for more then a half an hour on a problem they would just look at the trusty solutions manual for a hint and continue solving the problem afterwards. Also, some exam questions were straight from the book (unassigned problems) so the student with a solutions manual could do every problem in every chapter and be better prepared for exams in which the mean hovered around 40%. Those students were able to grade their work and spot errors that gave them a deeper understanding of the voluminous amounts of materiel you are expected to assimilate in a first year economics PhD program. It is a huge advantage to do as many problems as possible before a test to hone the problem solving acumen that got you into the program to begin with (you and everyone else in an economics PhD program).

Some friends of mine have speculated that maybe UV is trying to shake out a large first year class and have used this as an excuse to cut down the number of students. It is just unbelievable to think that the professors are so out of touch with reality that they don’t know solutions are photocopied in volumes every semester. Hell we had a professor give everybody full credit for our homework because everyone had the solutions manual anyways. I guess it could be possible that a naïve group of professors don’t know this goes on. A friend at Cornell pointed out that he couldn’t remember the last time a professor in a first year PhD program even graded an assignment? I know my TA’s graded damn near everything…I still have the unintelligible handwriting on my problem sets to prove it.

I don’t know anybody at the UV graduate economics department (I was accepted with a fellowship to attend the program and would have been in this years class had I had the inclination), but I would love to know the real reason for this controversy. Is it internal faculty politics? Naivety? Too many people passed the qualifying exam and the department wants to save money? If anyone knows, please email me.

What I am certain of is that, based on the information available so far, nearly every student in every graduate level economics department is guilty of cheating. Could it really be that having a copy of the solutions manual really cheating? Damn!

Update: It has been brought to my attention that the University of Virginia is usually referred to as UVA (not UV as I posted). Ooooops!

Update II: Ian writes in comments a post that deserves to see the light of day…

Glad to see you post something on this Chris. As you know, I think this is a very bizarre move on the part of the UVA econ department. I would think that a program like UVA, which I understand to be rather competitive in the first year, it would be preferable to ensure that students not gain an edge over each other by the accident of access to information. I suppose kicking students out for “cheating” is one way to do that, but of course all it means is that the people who have the materials will just do a better job of hiding them from one another. If solution manuals are outlawed, only outlaws will have solution manuals. The other way to remove the advantage conferred by having the solutions is to make them available to all students.
This is why it is incumbent upon graduate students to make old class materials as widely available as possible. First, it is the ‘fair’ thing to do, since it restores the balance of competition within departments (not to open a debate on the notion of ‘equality of opportunity’). From a practical standpoint, if graduate students are ‘out’ about sharing old course materials and solutions it forces administrators to acknowledge the issue early on. At the very least it makes it harder for them to pretend ignorance of the practice at the end of the year.
John Morrow deserves a great deal of credit for getting this information out there. I hope to do something similar for the incoming class at Cornell, and would like to open a dialogue with our DGS on this topic. Of course, there are always copyright issues to worry about. Mr. Morrow could speak more to these than I, since his website seems to have drawn some fire from publishers. Maybe it’s time for econ grad students to begin a grand project of ‘open source’ solutions. We wouldn’t have to worry about copyright violations, and they would probably be better than some of the terrible (and often wrong) solutions in the official publications (not that I’ve ever seen them, of course; I’ve just heard).

Update III: Crowebar has more here.

This just doesn’t add up.

Wait…lemme check the answers to make sure….

nope…this just doesn’t add up.

Update IV: Steven Levitt writes…

A few things surprise me about this article.

First, when I was in grad school, no one cared how you did on problem sets. They had nothing to do with your final grade. I got a “check minus” on every problem set in the most important fall quarter class the first year (Frank Fisher’s micro class), but I got an A in the class. Of all the things to cheat on, problem sets would be at the bottom of my list.

Second, I wonder how the professor didn’t notice. Either the students were creative, or the problems must have been pretty easy, so a lot of right answers weren’t surprising. At MIT and Chicago, I feel like hardly anyone ever gets
a completely right answer to anything.

Does the honor code apply to professors? If, say, a professor pretended to grade some undergraduate papers, but really a grad student did it, would the only choice be that the professor gets expelled?

an annonymous commenter writes that…

The problem sets are what the papers are focusing on but they are not what is being focused on here at UVa. Somewhere between 3 and 9 students are suspected of cheating on the prelims which is what started the whole thing.

This would make more sense to me.

The Q results were announced today. The early word is, yet again, the students were slaughtered. I feel for you guys. I’ll write more when I know more. How did you do Govind?

Update: 33% pass rate for Applied Econ at Cornell this year…I can’t get anyone to talk about what happened at the pure econ department…the lack of willingness to talk about it tells me that the numbers are not very good in econ either. Damn!

Update II: The study group I abondened multiple times passed with flying colors. Congratulations to Jessica, Ian, and Amanda. Good job.

Chris Silvey

The trade-off between speed of hiring and expected quality of the worker is a common concern for hiring managers and HR staffs. When the decision to hire a new employee is finalized the desire to get the right candidate in a short amount of time for a fair wage is a pipe dream that often will not be realized. Generally there is a trade-off between the speed of a hire and the quality of the candidate. Companies such as Google have hiring processes that seemingly have no end. When I came back to California to look for a job, Google was hiring Financial Analyst…if you check monster.com you will find that they are still hiring for that position, almost five months later. Google is looking for the absolute best candidate, there seems to be no time pressure for the hire. Most companies do not have the luxury of patience that Google does. Many companies aren’t as lucky and need to fill a slot within a finite time frame. If at the end of a month a company doesn’t have the right candidate for the job they have to make a hard decision. Either raise the compensation offered or accept a less then optimal candidate.

Jed DeVaro at Cornell University recently wrote about Recruitment Strategies in the economics journal Economic Inquiry. Dr. DeVaro assumes that the preference between speed and quality remains unchanged by a company during their recruiting search. The only changeable variable is wage; therefore a bad job campaign will lead to a higher wage offered. However, a long campaign does not necessarily lead to a higher wage. A long campaign either indicates a bad campaign or high quality expectations (ala Google). So poor campaigns that go long will lead to higher wages offered, but long campaigns don’t necessarily lead to higher wages due to the high expectation variable. Dr. DeVaro begins to speculate about differing search methods and how they can affect the speed of hiring, and therefore the wage that is offered.

Dr. DeVaro’s paper has left me wondering about other issues. First, I wonder if high quality expectations lead to higher wages. Does Google, or other companies with exacting hiring standards, end up paying more for like candidates then other companies, ceteris paribus? Although the small supply of qualified candidate will certainly raise wage levels there are other variables such as the opportunity to work with other thoughtful and highly talented people that could lead to people willing to work for a depressed wage.

An additional question Dr. DeVaro has left me with is whether a good job campaign will lead to lower wages. As an example let’s assume company X lists a job for one vacancy at $20 an hour and gets hundreds of qualified applications for the position. Would company X ever call the applicants and tell them that due to the enormous response they are changing the compensation down to $10 n hour and hiring two applicants instead? Although this is an exaggerated example can you think of a real world situation where huge response will lead to a reduction in wage offered in the same period. DeVaro thinks it doesn’t happen…does it?

I got a note from a professor at Cornell who sold me on coming to the school a few days ago. It read…

I am spending my day going through Econ admissions folders for this year,
and that made me remember that I actually haven’t yet met you & some other
1st yr students I emailed with, so just wanted to write a quick email to
say hi and see how things have been going with your first semester.
Hope all is well!

err, I guess she hasn’t heard the news yet.

Ian (a married PhD first year at Cornell) responds to the bottom two quotes…

I feel like I have to comment on both of these. I just attended a presentation by one of my (married) fellow students here at Cornell who has some very good data on PhD programs in the humanities. He found that married men are more likely to finish (rather than quit) a PhD and to finish early. Unfortunately, his data doesn’t include experience in the job market, so maybe married students really do write crappier dissertations, as the poster suggests. And, of course, this is for students in the humanities, where the average completion rate is like 8 years, or something. The point is that there is reason to think that married students fare as well as their single counterparts. Of course, you have to be sure that your significant other is ready to suffer with you through the privations of graduate school life; something that is all the more difficult, I think, in Ithaca.

Which brings me to my second point, in response to Doug, who is worried about what to do between now and math camp. My advice would be not to get too freaked out about the work at this point. There’ll be time for that when you get there. I am only in my first year (at Cornell), so I haven’t seen a lot, and I don’t know what it’s like at the higher-ranked schools, but plenty of people make it through the first year of an econ PhD, even if they aren’t math gods. Being a math god helps, but the reality is that if you work hard, and stay happy, you can learn what you need to learn to do what you want to do. I think Chris’ experience is instructive, in that he left grad school for the reason I think most people do: unhappiness. Sure, some people fail out, but most people who quit do so because they decide they don’t like what they are doing. Studying dynamic programming or constrained optimization for the next six months isn’t going to make you like grad school more, so don’t do it. Spend the next six months enjoying life. If you want to do something related to economics, start reading the paper, watching TV, and talking to friends to find out what makes things tick. Write down every idea you have about economics, because once you start a program, you won’t have time to think about it. Then the next thing you know, you have to start writing papers, and if you don’t have a good idea, it’s going to be haaard.

I will no longer be pursuing a PhD in economics.

The withdrawal papers were signed today and I have resigned to the thought that I will never be called Prof., Dr., or have the letters PhD in any way associated with my name. There are many reasons for this decision and I make it with a heavy heart. The main reason for my decision is the fact that my wife is miserable in Ithaca, New York and didn’t want to come back after the winter break. Upon a thorough discussion we decided that five years of poverty in Upstate New York just isn’t worth the cost of postponing family and happiness. I will write more about the tertiary reasons for my decision. For now, I thought I would let you know why I haven’t been very prolific lately.

Don’t underestimate the hardship that pursuing a PhD inflicts on spouses and other immediate family. Although it is surely possible to achieve your goals, make sure your partner is happy and in some way connected to a life of sporadic solitude and abandonment as Kuhn-Tucker, Euler, and whoever invented Log-Linearization steals your loved one.

Update: There is a good discussion going on over at testmagic about whether a Phd is worthwhile or just for dreamers. I think the answer varies and is dependent on many extremely personal criteria. Check out the discussion and give your own input.

If the link does not take you directly to the discussion, ou can find it at…

TestMagic Forum > University Admissions Forum > University Admissions: Graduate > Why do you guys do a PhD?

Update: David Tufte has some words about graduate school success…

Condolences. I have three contributions. Also, see my post on what determines success in graduate school.
Something like this can happen at any time in your career. It happened to me the year I went up for tenure: fixer-upper house, surprise pregnancy, administrative position, family health problems. The upshot is that a research program that was really starting to take off is just about dead. All the work on my serious publications was finished by summer of 1997. I’m not making a drastic life-change like you are, but the career I loved is mostly collecting dust in file cabinets while I struggle to find time to teach classes, go to meetings, and occasionally make a post. You may be better off going through this now.
When we eloped to Lake Tahoe (in my first tenure-track year) the casino had a wedding photographer on retainer. He offered one of the best lines about life I have ever heard: “In another lifetime I was a (botany) professor.” He was 50-ish, and had no regrets about either having been a professor, or having walked away from it all. So, chalk it up as another entry on your CV.
I grew up in upstate New York (suburban Buffalo). I both love it, and recognize that it is an appalling place for 7 months out of the year. I didn’t give up my career as the botany professor did (although I’m sorely tempted sometimes), but I did make a conscious choice to move to where I wanted to live (the rural intermountain southwest) instead of where I wanted to work (a research school). Having done that, I feel like I where a teflon shield about job problems: disappointing students (no problem, go for a hike), nutty colleagues (no problem, go for a drive), useless administrative tasks (no problem, go up in the mountains). I think you’re making the right decision (but beware if your wife is someone who just has trouble being happy).

Did you know this existed? It turns out the President of Cornell has been actively involved during his life.

It seems even University Presidents can have too much free time on their hands…I wish graduate students had such a luxurious amount of time in the year.

Mas-Colell, Whinston, Green Exercises, Chapter 3, Microeconomic Theory, Page 100, 3.G.6

After the first two commodities demand functions the problem states
“…where greek letters are non-zero constants.”

The answer then proceeds to declare…

“We must thus have δ = 0”….no errata correction in the solutions manual….ahhhh!

δ still is a greek letter, isn’t it?

I am reviewing Chapter 3 in Mas-Colell and came across this problem from one of our problem sets that always struck me as having an unacceptable answer. Here is the problem…

The TA in her solution sheet just ‘guessed’ (read magiclly came up with the answer with no work shown) that the utility function takes the form u(x,y) = min [x, 2y] and worked the problem from there. I am trying a more documentable route in which I show that the demand fn’s are homogenous of degree zero (HOD 0) and that the substitution matrix is negative semi-definite (NSD) and symmetric (by definition it is because it is a 2 x 2 sub-matrix). Any thoughts on the correctness of this route?