I recently submitted my thesis to the political science department for honors consideration. I wrote about political competition in autocracies, using the natural or accidental deaths of dictators as natural experiments for succession conditions. My findings weren’t earth shattering – statistical significance is not practical/policy significance – but what I learned through the process was valuable.

1. Start early

I use the verb submitted rather than completed to stress how much of a work in progress it remains, but deadlines are deadlines. A substantial amount of the final text was written in the weeks before the deadline, because I spent most of the nine months of the thesis process in exploratory data analysis and figuring out what the specific question and approach were, and then performing original research, which takes a hell of a lot of time. Early drafts, written in the first half, were completely rewritten and most of the initial text never made it into the final version. The moral of the story is to start early.

You should start thinking about topics by junior fall, take a class or seminar on that topic with a potential advisor in the spring, treat the term paper as a thesis proposal, and use the summer for field research. Most grant applications for field research expenses are due halfway in spring semester, and it can take a while to arrange fieldwork (getting visas, vaccinated, plane tickets etc). Even if you don’t go to the field, it would be a lot better to have a proposal and advisor early on.

2. Pick a contemporary, policy-relevant topic

We were advised to choose a topic of personal interest, since if we didn’t, we would get sick of it quickly. I followed this advice and chose my pet topic, and found that after weeks and weeks of skipped meals and sleepless nights going through archive after archive (protip: pick a topic where there’s good available data in a readily accessible format instead of doing original research!), or running endless permutations of model specifications to get a good fit, I got sick of it anyway.

If you’re going to suffer you might as well suffer strategically, so my advice would be to pick something that has some practical significance, such as a popular contemporary issue, or a specific policy or treatment. This is a good example. Natural experiments are typically not manipulable and hence less policy relevant.

3. Use LaTeX from the beginning

I write up most math and stat assignments in LaTeX, as the math:prose ratio tends to be high, but I write most term papers in Word, since there are typically only one or two equations and tables. Although most theses will also have a low math:prose ratio, I found that I spent a lot of time on formatting, and I could have been more productive if I had invested more time early on automating things. I wrote my draft chapters in Word, and when I had the final draft ready and it was time to format everything, marking up 126 pages and converting all the footnotes into BibTeX was simply impossible. I should have started writing in LaTeX from the start.

LaTeX has a steep learning curve, but the more you use it the easier it becomes, and you can use an editor like LyX to ease the transition from Word. You can also export Stata and R output to LaTeX code, and export from Zotero to BibTeX.

4. Don’t write too much

126 pages is far too long for to comply with most journal submission requirements, which means I am faced with the horrible prospect of rewriting most of it. I’m not sure how it came to be this long – my initial drafts were criticized for terse brevity. To keep the end product short, shoot for something like 50-60 pages including figures, tables, appendices etc. Allocate pages to chapters and sections in an outline, and stick to those allocations when writing out individual sections, or reallocate pages between chapters while holding the total constant.

After having gone through the thesis writing experience, I would not recommend it for everyone, and if you do choose to write one, do it right from the start.

My final semester, in full quant gear. It’s time to man up and math up. If I have to get out of school, I’m going out with a bang, not a whimper.

Required:

POLS C3998x-C3999y Senior Honors Seminar. A two-term seminar for students writing the senior honors thesis.

The thing I worry about the most, all the time. I live in perpetual thesis anxiety. Chris Blattman was right, only commit yourself to writing a thesis if you truly must.

ECON W4921 Political Economy Seminar: The Non-Market Business Environment. The course will apply the lessons of political economy to study the non-market environment within which businesses operate. Topics include lobbying, regulation, media relations, and international trade.

My last required class for the joint major, and it looks like it’ll be fun. Maybe I can get a lobbyist job out of this.

HUMA W1123 Masterpieces of Western music. The course attempts to involve students actively in the process of critical listening, both in the classroom and in concerts that the students attend and write about. The extraordinary richness of musical life in New York is thus an integral part of the course. Although not a history of Western music, the course is taught in a chronological format and includes masterpieces by Josquin des Prez, Monteverdi, Bach, Handel, Mozart, Haydn, Beethoven, Verdi, Wagner, Schoenberg, Stravinsky, Louis Armstrong, and Duke Ellington, among others. Since 2004, the works of jazz composers and improvisers, such as Louis Armstrong, Duke Ellington, and Charlie Parker, have been added to the list of masterpieces to be studied in this class.

My last required class for the Columbia core. One of the fun parts of the class is attending an opera, which I almost always enjoy.

Electives:

POLS W4292 Advanced Topics in Quantitative Research: Models for Panel & Time-Series Cross-Section Data 3 pts. This course covers methods for models for repeated observations data. These kinds of data represent tremendous opportunities as well as formidable challenges for making inferences. The course will focus on how to estimate models for panel and time-series cross-section data. Topics covered include fixed effects, random effects, dynamic panel models, random coefficient models, and models for qualitative dependent variables.

With this, I will have completed the entire polisci graduate methods sequence, well, except the basic classes like the ‘math camp’. The equivalent class in the stat department does not seem to be as focused on social science applications, and I’m not sure if I would ever use Fourier transformations or spectral density estimation in my own research (or any finance quant job), so this one seems more appropriate.

MATH W4061 Introduction To Modern Analysis I 3 pts. Real numbers, metric spaces, elements of general topology. Continuous and differential functions. Implicit functions. Integration; change of variables. Function spaces.

MATH V2500x or y Analysis and Optimization 3 pts. Mathematical methods for economics. Quadratic forms, Hessian, implicit functions. Convex sets, convex functions. Optimization, constrained optimization, Kuhn-Tucker conditions. Elements of the calculus of variations and optimal control. (SC)

Last year’s valedictorian, the god of econ, advised me to take 4061 “and ace it”, if I ever wanted to go to econ grad school. Having given up on ever getting into a good econ program, I’m not sure if I really need to take this, especially since I don’t expect to do much formal modeling, though it seems like some people recommend real/functional analysis anyway, as a kind of mental abstraction jujitsu.

After attending the first class, seeing all the math and physics majors there, and struggling with the first ‘Baby Rudin’ homework, I despaired and dropped down to the easier ‘math for econ’ analysis class, that covers only the subset of real analysis most relevant to econ applications, which will probably help motivate me to stick with it.

There is a graceless human tendency to wish upon others the ills visited upon oneself. Instead of pointing successors towards short cuts, you relish seeing them clambering through identical hoops.

The author of this particularly insightful observation was describing her experience advising a prospective journalist. She was asked for advice about how to write a book about Africa in a thoughtless way that was not conducive to actually receiving the desired information (it provoked the sentiments quoted), but she offered some very different and still valuable wisdom. In other words, the person got what they needed, but not what they wanted to hear. Unfortunately, it’s more likely that her wisdom was ignored by the intended recipient.

Sometimes people ask me for advice, and sometimes I offer my advice unsolicited, and in both cases I am usually glad to help and share what I know. But what people want to hear is usually quite different from what they need to hear. On one hand, telling people what they need to hear will probably mean being ignored or disliked, which is counterproductive to helping them. On the other hand, telling people only what they want to hear is not going to help them either, and may end up making me look like an unreliable source of advice too. I try to strike a balance and offer both types of advice, but it occurs to me that it may be better to perfect the art of disguising one as the other. It’s not easy, but I want to try my best to sweeten the bitter medicine of truth.

Most times I get asked about college admissions and applications. I am happy to share whatever I know, though in all honesty, I don’t have much insight into the internal dynamics of admissions processes (only admissions committee members would), at least no more than any conscientious research would uncover. Nor am I well suited to questions about comparisons between universities, only having attended one, and everyone’s mileage varies. But while I may not know the right answers about college admissions, I do know what the right questions to ask are. Here are two recent ones:

1. An acquaintance from my travels who I had helped with graduate scholarship applications earlier (unsuccessfully) was still pursuing her goal of working in the development/aid industry and continued to apply to graduate school. This time, she wanted advice on whether to attend a particular masters program in the EU, which I was unfamiliar with. To answer her questions, I suggested some topics (placement, funding etc) for her to ask the department about. But I also felt that what she needed to hear was whether or not attending this program, or any graduate program, made sense with her career goals, and advised her to rethink it. She wasn’t too keen on that, perhaps because she didn’t have very clear career goals, and eventually decided to attend, despite the lack of funding and a placement history that may not have been a good fit.

2. Another friend was considering applying to LKYSPP and asked me how competitive it was (I have no idea). I gave him my (uninformed) opinions on the program, the usual questions (placement history etc) to ask the school about, and some general information about living in Singapore. However, I thought what he needed to hear about was whether LKYSPP or any other program was the best way to get where he wanted to be, and advised him to consider other alternatives. He didn’t seem to want to discuss that, so I did not press the issue.

In both cases, they had effectively made up their minds beforehand, and my answer was mainly to confirm what had already been decided upon. I was glad to help them with their specific questions, but sad that I failed to also tell them what they needed to hear: that what is on the table may not be in their real interests.

For a long time I’ve considered getting what I want to hear the main problem. Finding the right adviser, asking the right questions in the right way. But the real problem may be about being open to the right advice when it comes my way. How to obtain the job I ‘want’ presently may be secondary to what I would find meaningful and fulfilling over the long term. The latter is a far more difficult problem to solve, and more in need of good advice.

An upperclassman working at an MBB consulting firm told me that it hadn’t been his ideal choice to take up the offer, but it was a highly practical one compared to what he really wanted to do, which he is putting off for one or two years. When I suggested that he should count his blessings, given that I don’t expect to even have an offer to consider taking, he replied, “It’s easy to just take any job, it’s easy to apply for what’s available. Waiting for something that you really want, that’s hard.”

Sometimes I need to remind myself to listen to what God is telling me, even if it’s not what I want to hear.

The penultimate semester of my undergrad days. Minimum number of classes, maximum attention to each.

POLS C3998x-C3999y Senior Honors Seminar. The Senior Honors Seminar is designed to help students majoring in political science write a full-length, scholarly paper that is of comparable quality to published work in the field. These papers, or theses, may also qualify the author for an Honors degree.

So I’m in the polisci thesis writing seminar. I wish I had planned this way in advance so I could have used the summer to gather field research (and got a grant to do so). Oh well.

POLS W4291x Advanced Topics in Quantitative Research. This course covers methods for empirical models that have dependent variables that are not continuous. These models include dichotomous and polychotomous response models, models for censored and truncated data, sample selection models, duration models, and models for count data.

Was advised that this was a good econometric methods class for my polisci thesis since it’s likely I’ll have lots of dummy DVs. As far as I can tell I am the only undergrad again. At the first class, one of the phd kids asked me why I had finished the 2nd year requirements before even starting. My answer: teching up, so in the likely event that I do not get into a good program, I can find some kind of quant RA job. Truth is, I’m probably still behind the top econ undergrads in technical skills, especially those kids who did the advanced time-series class.

POLS W3952y Seminar in Comparative Politics. Varieties of Capitalism. This course provides a theoretical framework for understanding the variation in economic and social institutions among advanced capitalist economies. Can we meaningfully talk of the German or Swedish model and, if so, what are their distinctive characteristics? In what ways do these economies differ from liberal market economies, such as the United States or the United Kingdom? Do these cross-national differences persist in the face of increased economic integration and globalization? We will explore these questions by examining institutional and policy differences in the following areas: (a) training and skill formation; (b) financial institutions and corporate governance, (c) the welfare state, (d) systems of industrial relations.

Was advised that this was another good class for my thesis. Although its mostly advanced economies, I’ll pick up more cross-country techniques that could be applied.

PHED C1001x Fencing.

Columbia actually requires me to take a PE class to graduate, so I picked something to fulfill my action-adventure fantasies. Even though its unlikely that I’ll become a master swordsman by end of term. At the first class, our instructor told us stories about fencing during the cold war and how it was seen as a ‘bourgeois’ activity until they realized it was a good source of Olympic medals.

Non-class activities: RA job, GRE prep, language study, apps, job hunting

Congratulations! I’m happy to say that we’ve finished reading through all 1500+ applications, and your name has come out on top. You should all be extremely proud of your accomplishment, as this was the most competitive year in Business Today conference history (our acceptance rate was even lower than that of Princeton itself)!

You come from over 20 countries and 100 colleges; You speak more than 50 languages, you’ve started your own businesses, you’ve won international awards, and you are all around world-class students. It didn’t take me long to realize that I wouldn’t have been accepted to my own conference.

This will be my third time at Business Today, and I’m looking forward to seeing old friends again. I wish that more of my friends from GIS had been accepted, though. Some of them are far more accomplished than I. Unfortunately, I suspect the selectivity of the program rises proportionally with the cost of flying the candidate over, and for me that’s a $2.25 subway ride.

In retrospect, it was unrealistic to think that I could do much reading or homework while attending a conference, much less running one, but things turned out fine on Monday despite having to skip lectures to catch up on sleep. I had a great time, even if it was a weekend with an average of four hours sleep a night.

The ‘prestige’ highlight of the weekend was meeting the president of my favorite think tank at its top floor conference room. I didn’t get to say very much, but it was still pretty cool. I also got to listen to great speakers in the movement, and chat with a few of them afterwards.

The ‘fun’ highlight of the weekend was meeting my favorite blogger economist again, who came to pick up his daughter who was a second-time conference attendee. At the social event, most people wanted to talk economics with him, except me – I figured that there was nothing economic I could talk about with any added value to him. Instead, we talked about his other passions, travel and cuisine, and I mentioned the best restaurants and places I went to over the summer. I also asked about DC restaurant week. He advised me never to go anywhere then, and to try out restaurants that refused to participate.

I also got to meet some really interesting kids from all over, some of which were very inspiring with their life stories and achievements. There were also more pretty girls than I expected would come to a conference like this, though of course that isn’t so important: Somehow the topic of Valentine’s day came up (something about taking the missus to Puerto Rico for a weekend), and with that, talk about relationships as transactions. I shouldn’t be taking relationship advice from an economist, but his reminder about competition for mates as a status game was both discouraging and comforting. Discouraging, because I can never be a high status male here. Comforting, because then I can stop trying.

The more I think about it, the more appealing a near-future in DC looks, even if it’s an entry level, pre-grad-school type RA job that pays a bare subsistence wage (including grants!) and necessitates living in sketchy conditions, though some of my friends who are doing that told me that they found it slightly suffocating to live and breathe politics and policy every day and night. I guess it’s hard to tell whether I would feel the same way if I was there full-time, given that we all spent a whole summer interning there and loved it. I’d love to try it out, even if it means being poor for a while.

So I’m off to DC for the weekend to run a conference at George Washington University. The last time I was in DC, I never really got to see the interior of GWU (I saw Gtown, SAIS, and GMU), so this will be a new experience, and I get to see all my friends in DC again too. If only I didn’t have midterms next week and problem sets due on Monday, or I would be able to fully enjoy the conference. Conferences are best enjoyed during the summer… but that’s also when its hardest to take time off work to go.

This has to be my toughest semester, with many long days in class and late nights reading or finishing problem sets. This isn’t because I’m taking the maximum courseload, like my peers who accelerated their graduation by a year or more, but because I’m taking hard classes. This is what my days are like.

The 2 required classes can’t be helped. I have to take intro IR and political economy for my major, and I have to take them now. For electives, I had a range of quantitative classes to choose from: the applied regression class, the polisci department’s regression theory class, the econ department’s regression theory and time-series/forecasting class, and game theory. My life is one big constrained optimization problem. To ensure that my GPA improves, my courseload has to be at a minimum, so I can only choose 2 electives from the above. But I have to choose well: Too easy and I won’t look competitive. Too hard and my GPA will not improve. Balance what I need to learn with what I need on my transcript.

4412 Advanced Econometrics
Having already taken the econ department’s basic regression class and matrix algebra last semester, I thought I would try my hand at the advanced class, which uses matrix algebra. I dropped out three weeks in. The class is technically undergraduate level, but it is about 20-25% applied masters students (and one visiting scholar, I have no idea why he’s there). It is also 50% Asian, 80% male, and 100% hypercompetitive. I recognized a lot of faces on the first day, because they had been my TAs in previous econ core classes, which means that they are among the top undergrads in the department (I am not). Most of them are joint math or stat majors. When the instructor explained that day that the grading curve would be steep, I knew I was in the wrong class. The learning curve is also very steep; the instructor sped through the derivations in matrix form and I left class each time feeling distinctly confused. Although I’m no longer registered, I still attend from time to time to get more familiar with R and Matlab.

4912 Multivariate Political Analysis
I switched to the equivalent class in polisci, which is also technically an undergraduate class, but is 95% graduate students. The remaining 5% is me, it’s a small class. About half are PhD students on the quant track. The learning curve is a lot less steep, since the class population doesn’t have as much math background, so I’m not at as much of a disadvantage. Instead of R, which is the preferred statistical package of the econ and stats departments, the class uses Stata. The assignments are part math and part applied, and usually involve reading journal articles and replicating their results.

4911 Analysis of Political Data
Right after regression theory is the applied class, which is about five times larger. The course is 100% applied, all the assignments are practical (collecting data, running regressions, writing a paper). No sit-down exam, no calculus, no matrix algebra, which is why the class is so popular. I attend the lectures from time to time to listen to the practical aspects. The class is also across campus and up five floors, so it’s good morning exercise.

4209 Game Theory & Political Theory
And then there is the polisci department’s game theory class. Another nominally undergraduate class (4000-level) which is actually graduate (6000-level), it had sixty people registered at the beginning of the semester. Half of them have dropped out. There are only three or four undergrads remaining, and everyone else is PhD or applied masters. I expect the non-econ undergrads to drop out shortly. The class is hard, the learning curve is steep, and the assignments are proof-based. I spend hours and hours on the proofs late into the night. Recently I’ve been having nightmares about them.

Although staying in the class is a big risk to my sanity and GPA, there are some good reasons to hang in there. First, the instructor is great, and so are the TAs. They make the class fun, and since the class size has halved, there’s plenty of access to them. Second, since very few of the non-econ students in the class have backgrounds in formal modeling or modern analysis, it’s a level playing field. Third, a lot of the course material overlaps with my political economy class, so my semester is well-complemented. Fourth, I will have mastered R and LaTeX through the coursework.

6640 State & Illegality in Southeast Asia
My fun class of the semester is a SIPA course on political corruption taught by the Southeast Asian Studies fellow. SIPA classes aren’t open to undergrads but I go to lectures anyway. The class is mostly area studies instead of thematic, but I get a lot out of the case studies too, and it’s always fun to meet SIPA students.

Congratulations! This e-mail is to inform you of your official acceptance to the Ivy Leadership Summit IX, which will be held at Brown University this year, February 6-7. This year’s theme is “Leading in a Global Community,” and will feature panel discussions ranging from sustainable energy, to civic engagement, to cultural leadership. This will be a wonderful leadership development opportunity as well as a chance to meet and form connections with your peers from other Ivy Institutions.

This will be a good time to try out the ‘conference commando’ chapter from Never Eat Alone.

When I was in Washington DC last summer, I really wanted to join a White House tour, but it never materialized – our internship coordinator put one together before I had joined the think-tank. Instead I got to see the Capitol and various congressional buildings through the course of the summer. The closest I got to the White House was with all the other tourists on the South side of the park, peering through the fence, trying to identify which little speck in the distance was the oval office.

Zhongnanhai is China’s equivalent of the White House, part of the Forbidden City that was occupied by various warlords after the last emperor’s abdication and now serves as the headquarters of the Communist Party – it is where Mao and Deng lived. Now that tourists can flow in and out of the palace museum, this is effectively the new Forbidden City of modern China.

My academic and somewhat personal interest in the nature of autocratic regimes makes me curious about what lies beyond those high walls. In the two school terms before this summer, I turned my curriculum to the study of modern Chinese history, politics, and foreign relations – which all revolve around decisions made there. I always imagine it as a quiet palace courtyard, with mandarins sipping tea while playing chess, while aides scurry back and forth with reports of the state of the middle kingdom written in calligraphic poetry. Others imagine it to be a more interesting place.

(Zhongnanhai, Beijing) The interior sign is hard to read but it says “Serve the People”… which people exactly?

Since I am in Beijing this summer, I hoped to catch a glimpse of this mysterious world myself, and planned it into my itinerary. The southern entrance, the Gate of New China (新华门), is just one block to the left from Tiananmen Square, but unlike the White House, there are no crowds of tourists gathered outside snapping photographs. I was the only one who wandered that far from the tour track around the Forbidden City entrance. I waited for a bit hoping to see some central committee cadre arrive in a motorcade of black limousines, escorted into the compound by waiting attendants, but no one came. Later a schoolmate informed me that Hu Jintao and other leaders don’t actually use this entrance, which is purely decorative. (They use the one on Fuyou street)

I walked around the side streets to look for more insights into the top leadership, but it was very quiet. All I saw were uniformed guards standing behind iron gates. The only place of interest was a foreign language school, presumably for the children of top cadres. Several RMB1 popsicles later, I decided this was far too much walking for me, and decided to meet my friends. I sent out a few texts to my schoolmates here over the summer, and I got one back from a Mandarin class buddy about a birthday dinner in Sanlitun, an expat playground known for… excess. I didn’t have any other dinner plans and opted in.

I thought it would be a guys night out with other schoolmates (the text didn’t say very much), so I arrived in shorts, sandals and a funny t-shirt from a previous summer camp. I ended up sitting next to the ambassador’s son at a posh dinner party full of socialites and power brokers. Oops. I tried as best I could to make up for my gross disregard for dress codes with my witty conversation, and got invited along to the afterparty, which was at a chill hangout on the rooftop of some building out in 3rd ring, where I heard more English than Mandarin.

(Club in Beijing) Count the number of expats!

Amidst the house beat and dancing, I realized that this was the center of gravity I had been searching for all day. This was where a new ruling class of nouveau riche elites and their foreign friends sealed friendships and deals. And I wondered what I was doing here: middle-class public-school/transport me barely getting by on grant funding, next to people to whom money is not an issue. They spent more in that evening that I did in my entire time in Beijing.

It occurred to me then that my schoolmates from the mainland (at least among the undergraduates) all come from particularly privileged backgrounds. Not those closely related to the top party leadership, as one such friend later told me while sipping tieguanyin at a Houhai teahouse, because those would supposedly consider studying in the US (or be considered as) a security risk. Rather, they tend to be the scions of business empires, and go to the same few preppy international schools that feed into the Ivy+ machine.

Another friend, one who went to one such preppy international school in Shanghai, noted that the concentration of extreme wealth among the undergraduate mainland students is due to a confluence of factors in Ivy+ admissions (need-sensitivity, country/diversity quotas, number of competitive applicants etc) that effectively exclude applicants from China who require financial aid – which in purchasing power parity terms means that someone is very privileged. This phenomenon is less noticeable at the masters and PhD level. Granted, all playing fields everywhere are tilted to the rich, and the difference is in degree.