Author Archives: kevin

About kevin

I write the posts

Clusters of identical sellers in a crowded city

Posted on
In and around downtown Udaipur, you run into clusters of merchants selling exactly the same thing. There’s a block with ten pharmacies in a row, all selling roughly the same goods and another block with five banks in a row. In Delhi Gate, there are about twenty women sitting next to each other who all sell flowers. Hotelling’s law says that on a 2-dimensional street with two shops, they will rationally converge at the middle. And I remember reading somewhere that McDonald’s spent a ton of money figuring out the ideal locations for its stores, and Burger King just built a restaurant wherever there was a McDonald’s. But with ten shops all in a row, common sense says a shop in the middle of the row could make more money by moving to the end, or to a different part of town. I can think of a few reasons: 1) The shop-owners all purchase from the same distributor. Perhaps they get a discount by purchasing together. 2) Maybe everyone in Udaipur knows the place to go for flowers or pharmaceutical drugs and so there’s a big market in that particular location. This is similar to the distribution of car dealers in the US. However, car dealers sell differentiable products and you can find pharmacies and flower sellers regularly in other parts of town. 3) Everyone in Udaipur knows everyone else, so buyers spread out their purchases between sellers. The sellers are selling indistinguishable products so it’s hard to stand out as the best. 4) The sellers are colluding and sharing profits. That, or the margins would be low anyway and they enjoy each other’s company. 5) Flowers, drugs and banks are just a front for some kind of illegal business. Not likely as cultural norms against alcohol, gambling and prostitution are quite strong. 6) Peculiar zoning laws mean that pharmacies/banks that set up in a specific spot got a special tax break. I have heard as well that in India the laws favor small businesses, opposed to big business. 7) I hesitate but maybe the sellers have a weak grasp of business concepts, or lack the desire to make higher profits for whatever reason. 5, 6, and 7 are particularly unlikely, and 1 and 4 are the most likely, None of the arguments here are particularly persuasive. Hotelling's law may be more powerful than I thought. I will investigate further.

Liked what you read? I am available for hire.

Interrupting pleasurable activities makes them more pleasurable

Posted on
Eric Barker links to a paper showing that good experiences become more pleasurable (and bad experiences less pleasurable) when they are interrupted. Showering in India means filling a bucket with (thankfully) hot water and scooping it over your head, pausing to rub in shampoo and soap. It isn't that bad, probably in large part because of the effect stated above. Each scoop of hot water is a mini-rush of relief. I am wondering whether user control has an effect. For instance, interrupting the flow of a continuous tap may make the experience more pleasurable but you also may get annoyed with yourself for not leaving the tap on all the time. Constructing an unreliable faucet, that would stay on between ten and twenty seconds and shut off at random intervals of one to five seconds, so that the water is running around 75% of the time, would lead to the optimal showering experience. This suggests we could probably have a better experience surfing the Web if we turned it off for minutes or hours at a time, and also suggests that kicking a drug addiction would be extremely difficult.

Liked what you read? I am available for hire.

Pay zero rupees, end corruption

Posted on

In an attempt to fight corruption, a local NGO in Chennai has begun to print zero-rupee bills. When public officials demand a bribe in exchange for a service, the people hand over a zero-rupee bill as a protest against petty bribery and corruption. The bills have proven so popular that the NGO has had to make a second, much bigger printing run of the bills.

Currently, everyone pays bribes and the people are worse off, overall and relative to the government officials. If I want to protest and choose not to pay a bribe, the government official will laugh in my face. Everyone else is paying bribes, so he doesn’t really need my money. Thousands of individuals face this scenario, and all choose to pay the bribe.

However, the popularity of the zero-rupee bill movement allows the people to coordinate and collectively pay zero bribes to the government officials. Because the movement is popular, and because many bills have been printed, I can confidently choose not to pay the bribe because I know that there are many other people also choosing not to pay.

The success of the movement depends on the incentives of the government officials. At the beginning of the movement, government officials will process no papers, and do no work, without bribes. If the officials can be fired for doing no work, then surely they will eventually begin to perform their job functions without the customary baksheesh, and everyone will be better off. However, if the officials’ job security is not related to the amount of service that they provide, then this movement may lead to the total shutdown of government, which may or may not be a good thing.

Evidence exists for each possible direction. In negotiations, the general principle is that the side that has more patience will win. China, for example, was willing to wait one hundred years to receive control of Hong Kong. The public officials are probably unionized, and as members of government, their jobs are probably hardly dependent on the amount of work that they do. In other words, they will have little incentive to continue doing work in the absence of petty bribes.

In 2005, a group of economists from MIT was approached by the Rajasthan police department, who wanted to improve both their performance and the perception of the police among members of the public. The team implemented a few changes, including a three-day communication and public relation training module, a work-rotation schedule where every police member took turns doing different jobs within the station, and community observation, where community members were invited to sit in the station in three-hour shifts, to facilitate communication and improved performance. These interventions led to a reduction in the number of people who reported fear of the police, and an increased feeling on the part of crime victims (and criminals) that they were being treated fairly (read the project summary here).

One interesting finding from this survey was that the rank-and-file police felt underappreciated, overworked and victims of manipulation by their superiors. These feelings may lead them to seek bribes from the public. This zero-rupee movement may in turn lead them to try and effect change within their institution, rather than trying to pass on their hurt feelings to rank-and-file members of the community.

On the other hand, the Rajasthan reform project was initiated by leaders within the police unit, who cooperated with the proposed reforms. The zero-rupee movement was initiated by members of the public and may lack support from the bureaucratic elite. I am doubtful the movement will have lasting positive effects, but I would love to be proven wrong.

Liked what you read? I am available for hire.

Breakfast at Tiffany’s

Posted on
Continental had an excellent movie selection: I watched The French Connection, The Informant!, Bonnie & Clyde, Breakfast at Tiffany's, All the President's Men, and Gladiator. Here's the final scene from Breakfast at Tiffany's: Notice how she only becomes interested in him after he stops his sycophantic behavior and walks away. Also notice that she's not at all interested in him until she gets turned down by a richer man. Neither of these suggest the relationship is going to last long.

Liked what you read? I am available for hire.

Saying goodbye

Posted on
Exaggerated goodbyes are a way to show that you care. In the old days there was a significant chance that you would never see that person again. But as far as I'm concerned, a far better signal is the one you show me every day, doing any of the things friends do for each other. I know how much you care about me and hopefully you know how much I care about you.

Liked what you read? I am available for hire.

Interviews

Posted on
Are interviews a good way of sorting the wheat from the chaff? Interviews reward those people who are good at telling short vignettes, who know what they want to do 5 years from now, those who are good at looking comfortable, those who can think on their feet, and those who can articulate the qualities about themselves that are desirable to an employer (and those who look comfortable). Maybe these qualities correlate to job success but I am not sure; as Schmidt and Hunter find, work sample tests are the most accurate predictor of job success. Two types of people do well in interviews; the good candidates who will practice interviewing and prepare answers and those who are natural speakers. Should taking some time to think of an answer help or hurt your image in the interviewer's eyes? Responding quickly signals confidence, and intelligence. While I expect the interviewers at most firms to attempt to overcome this quick-response bias they might not do so. Responding quickly means you might not have thought the answer all the way through, especially to a difficult question. On the other hand if the applicant pool is large and the number of acceptances small, the best candidates will be able to respond to the question at hand accurately and quickly. You are not expected to be honest in interviews. "I want this job because it pays very well" or "I want this job because it will give me high status" are not acceptable answers.

Liked what you read? I am available for hire.

Beer pong is dumb

Posted on

Beer pong tableSome people call it beirut, but either way it’s a terrible drinking game, unless there are exactly 2 or 4 people hanging out. It takes about 15 to 20 minutes to finish an average game with people who don’t play very often. The beer gets warm by the time you’re done drinking it. The ball flies into bushes, bounces on the floor, and then into your cup, and you have to drink all of the beer it touches. You might get to shoot only once every minute so it’s hard to become more accurate quickly. Furthermore unless you’re drinking on the side, you are not likely to increase your level of inebriation. I won’t discuss bouncing for 2 cups, or blowing the ball out of a cup, both of which are stupid. The game’s main reward for hitting many cups is “On Fire” which means you get one extra shot – not a great reward when the shooting percentage is in the 15% range.

Amazingly the alternatives to beer pong are either not very exciting games or involve drinking so quickly that people don’t like them very much. There’s got to be a better team game where one side wins, and that can be played in faster than 15 minutes (and involve more people). There should be much more experimentation on the drinking game front.

Liked what you read? I am available for hire.

Theory of racial identification

Posted on

On every internship application, you are given the option to identify your gender and race or to decline identification. What’s the optimal choice?

While it might go unquestioned by the hiring firm, to avoid a lawsuit, lying about yoru race is probably not the smartest idea, so we’ll rule it out.

I would assume that most African-Americans, Latino-Americans and Pacific Islanders, as under-represented races within most tech and finance companies, would mark down their race on the form. This means that pretty much everyone who declines to identify a race is white or Asian-American, like most people who choose not to upload a photo on a dating website don’t think too highly of their looks.

Your type is probably still showing through even though you hit “Decline,” and in addition, now the company knows that you are the sort of person who would choose to try and conceal their race. Companies are explicitly not supposed to take this information into account, but I highly doubt that they are able to stay impartial.

Liked what you read? I am available for hire.

Streamlining dorm visitor check-in

Posted on
One college I visited recently has a great visitor check-in system. At the front desk there's a binder filled with baseball card sleeves, and each dorm resident has their own sleeve. To check in, the visitor hands over their driver's license, which is placed in the sleeve. The visitor gets it back when he/she checks out. So easy! No dumb paperwork, and the college knows who's in the dorm at all times.

Liked what you read? I am available for hire.