Friday, August 28, 2009

types of taxes

Such a list as I will formalize below would be mind-numblingly boring to someone who understands economics, but unfortunately only 14% of legislators understand the law of supply and demand (People buy less when it costs more! That's not even the law of supply of supply and demand, just demand, but it's already too much for these people!), and I'm sure far fewer in the general population (god bless representative democracy), so I'm going to write it anyway. (At least so my roommate can read it and hopefully have one more source of economic logic in her life than personal anecdotes on This American Life.)

There are some types of taxes, categorized by intent, that are easily confused and have very different pragmatic requirements/implications. Let's get them straight for once.
  1. Taxes to price in negative externalities: Also known as Pigou taxes. If driving your car to work costs you five dollars, and costs society 3 dollars in pollution and road congestion, you should be taxed by 3 dollars in order to correct the market failure that will result. This shifts the equilibrium from one in which too many people drive to work back to one in which the socially optimal number of people drive to work. Practical issues of calculation and enforcement can be complicated but I'm all in favor of these types of taxes when they can be done well.
  2. Taxes to punish "bad" behavior. For example, a tax on soda or fast food or SUVs. If you just want to slap people on the wrist for doing something, whether or not they keep doing it, it doesn't matter what the target is. But if you really want to curb consumption, you need demand to be very elastic. A tax on soda may cause a lot of people to switch to juice or water, but a tax on heroine is not going to get anyone to break the habit and stop shooting up. These types of taxes are paternalistic bullcrap.
  3. Taxes to raise revenue. This can be levied on anything you want, but it needs to be something with very inelastic demand, so you won't just kill the market and fail to collect any tax revenues. These types of taxes are a necessarily evil but can be done in very good or very bad ways. Bad, mostly by failing to remember the "inelastic demand" qualification, but also by having undesirable implications on wealth distribution. This leads to:
  4. Taxes to redistribute wealth. This would be a secondary goal along with #3. If you can raise as much money by taxing private jets as by taxing bread, clearly the former is much better at transferring money from the rich to the poor. That's sometimes a good goal and sometimes not, but if it is the goal, that's what you should do. Note that this is very different from taxing "conspicuous consumption" or something like that. That falls under #2. Inelastic demand is again important for these taxes. If you want equality, you want rich people to spend their money so it gets transferred back to poorer people, and it doesn't much matter what they spend it on (and a tax on luxury goods with very elastic demand could just make them hoard it in their bank account or stick it in a trust fund to create more rich people, exactly the opposite of spreading the wealth.)
Let's consider an example... Tolls on the Bay Bridge. If you want the toll to compensate for the negative externalities of using the Bay bridge, charge the per user cost of bridge maintenance, plus the cost of increasing road congestion by one car (this cost varies based on current congestion. Tolls should automatically increase during rush hour.) Additionally, the toll should be much higher for cash payers than Fastlane payers, since the latter hold up traffic less, lower for motorcycles, higher for semis, etc.

Say, on the other hand, you want to keep people off the bridge, polluting the air and clogging the highway, and put them on the BART. You should just raise the toll until few enough people want to use the bridge, however many you decide that is. Cars should be taxed MORE than semis, since the drivers of cars are making a "bad" decision to drive, whereas truck drivers have no choice. Hybrid cars should be taxed less than regular cars, since they're being more "responsible" about gas usage, and motorcycles taxed even less, since they're more responsible about both gas usage and traffic congestion.

Now how about the case where you just want to use the bridge as a holdup to raise as much money as possible. Then it doesn't matter what kind of vehicle it is, you just adjust the toll upwards until your revenue is maximized (for awhile you can increase the toll without many cars stopping using the bridge, but eventually people will switch to BART and revenue will decrease. Just before that is where you set the toll.)

Now say you just want to redistribute wealth. Then you put higher tolls on more expensive and emptier cars and make sure you're not just killing your wealthy consumer base, so you actually raise money to give to poor people. City busses don't have to pay (remember that taxes on businesses are just passed on to customers - higher bus tolls translate to higher fares for their poor riders.)

Luckily, it looks like the toll structure most closely follows an effort to correct externalities. Taxes are higher for larger vehicles, and carpools are free during rush hours. Not perfect, but in the right direction.

So, next time someone tells you they want to tax SUVs to offset vehicle pollution, call their bluff. If they wanted to do that, they would tax gasoline by the gallon (which, lo and behold, already charges SUVs drivers more since they use more gas.) All an SUV tax does is spitefully punish "bad" behavior, and redistribute wealth.

Wednesday, August 26, 2009

The placebo effect

Very interesting article (via Matt Yglesias) about the placebo effect, and particularly how it is increasing, culturally dependent, and recently studied more directly.

I've always wondered why it was not the focus of study, but exclusively the baseline for study of other drugs, and why there is not a protocol (the ethical dilemmas should not be impossible to address) in place for directly prescribing placebos in cases where it is is likely to be effective and when, for example, side effects of "real" drugs are severe in general or for a particular patient. Doctors self-admittedly already do this by prescribing ineffective treatments - why can they not simply prescribe a placebo and call it something else (and I mean really brazenly: prescribe Valium and give them sugar pills, don't prescribe drug RG8789 which is secretly a sugar pill). This should at least be allowed in situations where the patient has previously signed on to such a program (maybe as an option through your insurance plan that any doctor can see - or as an option with each individual doctor or hospital, although I suspect that the more often you say "yes you are allowed to lie to me", the more you expect to be lied to, and the less effective is the placebo effect, so ideally it would be through insurance, which is both rarely changed and psychologically removed from the actions of particular doctors during particular office visits.)

Anyway there are lots of really interesting points in the article and this question of making placebos a mainstream remedy was not really one of them, that's just my own interest. More relevantly, placebo effects are highly culturally sensitive (drug companies now like to go offshore to run trials) and are increasing over time (possibly due to the success of pharmaceutical advertising raising our expectations of the effects of pills in the last 20 years). They have certain known chemical mechanisms (opioids released in the brain under stress, which reduce pain and moderate heart rate and respiration), and many un-understood mechanisms (blue pills typically reduce anxiety except in Italy where it is the national soccer team's color...). The placebo effect is more potent when fake drugs are administered with higher frequency, but can be a long-term benefit, contrary to popular wisdom. And, most intriguingly, it can be very very strong.

Tuesday, August 25, 2009

Sign me up!

Oh thank you marginal revolution for absolutely making my day with this. Atheist pet lovers paid to take care of animals when they're all left behind at Rapture! I need to look into this further - do volunteer pet care providers get paid a fraction of the $110 immediately or only in case of Rapture? Frankly either way I'll take the money. What better way to celebrate the end of evangelism than new kittens?

I also am very amused by the flip of asymmetric belief. More frequently, skeptical buyers purchase from sellers who claim to believe in their product (psychics, banking in the afterlife, etc). But here's a business run by people who are required to not believe in the value of its product, counting on customers who do.

You've committed your life to Jesus. You know you're saved. But when the Rapture comes what's to become of your loving pets who are left behind? Eternal Earth-Bound Pets takes that burden off your mind.
We are a group of dedicated animal lovers, and atheists. Each Eternal Earth-Bound Pet representative is a confirmed atheist, and as such will still be here on Earth after you've received your reward. Our network of animal activists are committed to step in when you step up to Jesus.

We are currently active in 20 states and growing. Our representatives have been screened to ensure that they are atheists, animal lovers, are moral / ethical with no criminal background, have the ability and desire to rescue your pet and the means to retrieve them and ensure their care for your pet's natural life.

Our service is plain and simple; our fee structure is reasonable. For $110.00 we will guarantee that should the Rapture occur within ten (10) years of receipt of payment, one pet per residence will be saved. Each additional pet at your residence will be saved for an additional $15.00 fee. A small price to pay for your peace of mind and the health and safety of your four legged friends.

Unfortunately at this time we are not equipped to accommodate all species and must limit our services to dogs, cats, birds, rabbits, and small caged mammals.

Thank you for your interest in Eternal Earth-Bound Pets. We hope we can help provide you with peace of mind.

Sunday, August 23, 2009

academic humor

Greg Mankiw strikes again, the decidedly unhumorous economist always leading me the way to hilarity. You must read this.

And, on a similar subject of the journal submission and referee process, you must read this, by Preston McAfee (who, as I have previously mentioned, is one of the most hilarious and interesting economists I've met.)


Friday, August 21, 2009

high-speed rail

I have to ask, what is with all the hype about high-speed rails lately? As far as I can tell, it's originated and propagated solely by Europhile liberals with wistful idealistic daydreams of zipping silently through the Swiss countryside. That may be a practical solution where major hubs of business and government are clustered within a couple hundred kilometers of each other across an entire continent and where a long tradition of rail travel has ingrained in the population the idea that rail travel is the default and easiest mode of transportation and that the noise associated with railways is just something society has to put up with, but these are two massive cultural differences that the U.S. cannot overcome to make such a system worthwhile here, even in the few isolated areas where major cities are close enough to warrant rail construction (and where a rail system does not already exist. This eliminates ... everywhere.)

Amtrak is an unmitigated disaster. A trip I can make in 15 hours by car from Los Angeles to Seattle for about $120 in gas money costs more to do by train and takes about 40 hours (despite schedules that persistently insist trains will arrive 16 hours before they do.) I can make the same trip for about the same price as it costs to drive by plane, and that only takes about 5 hours, including airport waits. I've never travelled this exact route by Greyhound, but I'd bet anything that is also cheaper and faster than the train, and has the added advantage that every few hours you can actually get off and buy some food that isn't sold at a 1200% markup. The only people who ever travel via Amtrak are rich vacationing retirees. And more of those Europhile idealistic liberals.

Not to mention Amtrak is a bankrupt company that only survives on government aid (guess which government figures push for that aid...) This is absolutely insane. An inefficient and outdated private company with a disproportionately wealthy clientele propped up by the US government? Please.

Rail travel is so much better for the environment than car or plane, and are a net benefit when you take into consideration those hidden costs, you say? Well then, please take a look at this (and the linked-to reports for details.) I will quote (emphasis mine):
Edward Glaeser (over at the Economix blog) and I have been writing about high-speed rail (HSR) over the past couple of weeks; he just finished his cost-benefit analysis of a hypothetical Dallas-Houston line with a look at land-use impacts. His overall conclusion, even making some very generous assumptions in favor of rail, is that the line would be a net cost to society of at least $375 million per year. This includes HSR’s potential environmental benefits as well as the direct gains to riders.
Just because one thing is better environmentally doesn't mean that it will automatically be economically worthwhile when you include those hidden costs to society. I am very sympathetic to these arguments when they are valid, having been an ardent environmentalist making exactly this argument for renewable energy and conservation since I was about 10, but it has become the knee-jerk explanation sans followup recently and we should impose some discipline on the makers of these claims.

Why is there not a push for innovation and reforms to make air travel cleaner and faster? The only reason people would consider even a high-speed rail journey, 2 or 3 hours, from LA to San Francisco, over a half hour flight, is the 2 hours of bureaucracy and security we have to endure to get on the plan. I suspect that if rail travel became more dominant, not only would security waits for trains drastically increase as security concerns mounted, thus eliminating much of that advantage, but that there is quite a bit of room for improvement in airport efficiency, especially for tiny regional hops that would be the substitute for rail travel.

Also, I suspect that similarly to the environmental gains of replacing a 9 mpg hummer with a 16 mph SUV dwarfing the gains of replacing a Toyota Camry with a Toyota Prius, there are enormous potential gains in environmental effects in airline travel.

Why are these things not being explored thoroughly and aggressively before to committing the government to billions and billions of both fixed and ongoing costs for infrastructure for an outdated mode of transportation in an ill-suited environment and skeptical potential consumers?

If the government hadn't been looking for every possible avenue to force money into the economy earlier this year, there is no way such a ridiculous plan would now be taken so seriously.

Monday, August 17, 2009


A Heartbreaking Work of Staggering Genius, by David Eggers - A thoroughly enjoyable and well-written novel (including the introduction - don't skip it) that perfectly captures the zeitgeist of 1990's youth and young adults, that generation of upper-middle-class children who, with a lack of traditional life obstacles to focus on, grew up with the luxury of diverting attention to self-analysis and finding angst and trauma in family turmoil and tragedy. Of course losing both of ones parents to cancer within a month at age 21 would be truly traumatic for anyone, and for the first 350 pages I was convinced that an otherwise normal guy had his life turned upside down by these events and he had a story exploding inside him that he had to put to paper semi-autobiographically in order to cope and move on. But those last 87 pages killed that character and left me with a lasting bad aftertaste, as it became clear that this was a guy who would always intentionally look for turmoil and exploit any situation for the emotional wreckage that leads to the ability to write. The rest of the book lost a lot of its integrity with that truth emerging more and more strongly throughout the denouement.

Economic Gangsters, by Edward Miguel and Ray Fisman - (Working for Ted Miguel this summer, of course I have to read his books, and am biased in my take on them. But consider that my praise for this book is not so much of a biased review of a book by someone I enjoy working for but a more innocuous result of selection bias, since I choose to work for people in the first place that are probably very good at writing books.) I tend to mentally use Freakonomics as the gold standard for pop-econ books, and while Economic Gangsters wasn't quite as whimsical and fun, it was definitely more interesting as a whole as every story focused on parts of a larger problem of economic development. Through diplomatic parking tickets at the UN, rainfall-based predictions of violence in Africa, thinking like a smuggler in Hong Kong and like a businessman in Suharto's Indonesia, and tales of war recovery from Sierra Leone to Vietnam, certain themes emerge as clearly important (culture of corruption) and others become frustratingly vague (war damages), and, just as with Freakonomics, the case is made extremely strongly that to understand problems on this scale, it's imperative to think like an economist.

Also, in this book I discovered my new hero, Antanas Mockus, philosopher-mathematician and mayor of Bogota who used the city for experimental new tactics of law enforcement such as hiring un-armed, unable-to-make-arrests mimes to mock violators of traffic laws. (Violations plummeted.)

Self-made Man, by Norah Vincent - A woman dresses up like a man for a year, joins a man's bowling league, stays in a monastery, gets a sales job, dates, and writes a book about her revelations about what it's really like to be a man. Bizarrely, I both really enjoyed reading it, and really didn't think it was very good, and it's hard to pinpoint why either of those are true. I think its just a bit too self-indulgent and personal to be a "good" examination of gender stereotypes, but it's certainly a fun collection of stories that press lots of buttons in the reader from a fresh perspective (well, not entirely fresh, Camille Paglia is similar and much more credible in her opinions somehow, maybe just because I really like her and not Norah Vincent, but obviously even the two of them together are very rare in their views on gender roles and feminism.)

Thursday, August 13, 2009

mixed signalling

Tyler Cowen (fittingly, in a post about signalling) says "in Sicilian terms I'm relatively well off".

This is a world famous tenured economist at a very good University. He's very well of by anyone's standards except maybe Michael Bloomberg. Clearly the he's trying to avoid sounding conceited. But instead he sounds oblivious to how the world really is.

How do you concisely convey "I'm factually on the upper end of the spectrum but it hasn't gotten to my head"? (And while avoiding the trap of saying it hasn't gotten to your head, because then the ability to not let it get to your head has gotten to your head...)

Wednesday, August 12, 2009

Hubble Ultra Deep Field in 3-D

An astounding astronomy video (via Starts with a Bang):

Tuesday, August 11, 2009

knowledge and metaknowledge

By the first year or two of college I had already forgotten all of the details of things I learned in high school and was left with only a shell of relationships between the details and concepts and methods of reasoning (which conveniently is most of what math, my major, is). Yet despite the fact I would have done just as badly on pre-tests on those subjects I had already learned as the less prepared of my peers, I felt that I had a huge advantage just by having a ghostlike shell of logic buried in my memory.

When I was working on Wall Street doing research with zero guidance for the first time I couldn't remember any of the statistical techniques I'd learned in five semesters of statistics and econometrics in college (Side note: frankly, all the talk about Wall Street hiring the "best and brightest" is total BS. Only rarely do you come across a mathematical genius on wall street who just happens to hate academia and love money. I was thoroughly on my own.) But simply having a vague recollection that some technique does in fact exist was enough to find it and learn it and use it on my own.

Especially in the age of google, knowing that something exists is much more important than knowing the details of that thing. One of the most useful classes I took in college was an introduction to experiments in social science. It was just an informal lecture, no theory involved, nothing in depth at all, but as a result I feel much more fluent in the field of experimental economics than, say, microeconomic theory, which I've taken MANY theory-based classes in. I could easily come up with an experimental/behavioral research question and know where to look to make progress on it, yet I have no earthly idea what modern general equilibrium theorists are working on.

Before information was so accessible, it was very important to have all relevant information on instant recall. Now that scientific careers are so much more specialized, not only is it impossible to remember everything you need to know, it's completely unnecessary when simply knowing that "dynamic price discrimination" is a keyphrase I can google to figure out how to maximize profits in my new airline.

To some extent, the educational system is very behind the times in this sense. There SHOULD be more overview classes where you just learn what is going on and skip the details. On the other hand, knowing what is out there is useless if you're not capable of looking it up and learning it on your own. School needs to make you very good at logic and math and self-teaching, and the best ways of doing that all require delving into details. Practice makes perfect and you practice on specific examples. Then worry about building up familiarity.

I think I just talked myself into putting effort into figuring out a better way of "glancing" at lots of information for my fields of interest (since overview classes don't really exist).

Saturday, August 8, 2009

Bureaucracy in research

Good editorial on human subject regulatory boards for medical research.

I don't have experience with human subjects boards in medical research obviously, but they are certainly ridiculous in economics. Somehow at my undergraduate university, a small private research institution with little bureaucracy (the registrars and deans knew me by name...), I never once had to deal with approval for economic experiments. Presumably there was standing approval for experiments of a certain type. Makes perfect sense. Economic experiments are basically paying people to play silly computer games. There's no risk whatsoever.

Yet at Berkeley, anything at all involving human subjects has to go through a grueling paperwork-laden bureaucratic approval process that takes months. And once you've collected all your data and have no more contact with subjects, you have to continue submitting paperwork just to analyze it. All for things as simple as fake auctions and maybe a few questions about age and educational background. I don't know how much of that is because Berkeley is an enormous public university or because it's politically particularly "liberal" in the "paternalistic" sense of the word, but I don't see why it has to be such a huge pain in the neck to do a simple experiment here and not other places.

It absolutely hinders research: you can get a heck of a lot more done when you can run an experiment three days after coming up with the idea, and then run slight modifications of it over the next couple weeks, following and refining the results which almost never exactly align with pre-formed hypotheses that were the basis of the experiment design, without redoing all your approval paperwork.

The U.S. is already having to make up for a comparative disadvantage in scientific research due to pro-life and other religious objections to various types of research. Unless the government gets out of the way of scientists doing their jobs, bureaucracy is going to be a much bigger problem.

Sunday, August 2, 2009

Wisdom of Statistical Crowds

Jordan Ellenberg is a mathematician I adored based on a talk on philosophy and mathematics that he gave at Caltech six years ago, who turned out to also be a delightful Slate columnist and author, and who has a blog now! (Thanks for the heads up, Ben.) Also, his list of books read is what inspired me a couple years ago to keep track of what I read, which is why you are periodically subjected to sets of three book 'impressions' (not exactly reviews) on my own blog.

Anyway, I liked this observation of his on the netflix contest (to improve their recommendation system by 10%):
One of the really interesting lessons of the competition is that blendings of many algorithms seem to work better than any single algorithm, even when there’s no principled reason to do the blend. It’s sort of a “wisdom of crowds of computer programs” effect.
And this reminded me of my experience trying to explain CDS spreads (basically, the price of debt default insurance) of emerging market sovereigns when I worked on Wall Street. I came up with a nice model based on current CDS and economic data and found a few key factors that determine the credit risk. I then set about backtesting my results with CDS data from 1998 and forward. As it turned out, using a giant list of every possible factor to predict future spreads is much more profitable than using a more robust, smaller model! (For those unfamiliar with econometrics, it is not recommended to have 18 explanatory variables to analyze a dataset of 24 countries because the model becomes "overfit" and useless - every spurious difference between two countries predicts a corresponding difference in the future that usually doesn't pan out.)

In this instance, while there is no principled reason to blend so many methods of prediction of CDS spreads, the fact that it worked so much better than the 'sensible' thing made me very curious about the mechanism of price determination in the financial sector. Prices are thought of as magical reflections of the truth (and I certainly don't want to undermine that fundamental trait of free markets - it is incredibly powerful and robust). But the probability of very rare events is hard to know and hard to bet on and hard to verify and adjust. The market price follows the beliefs of the buyers and sellers more than anything else, and those buyers and sellers were trying to make educated guesses using the same tools I was. Even if something in the data is spurious and not relevant to predicting future prices, if someone else sees it and made a decision based on it, it's useful for me to consider it too. So it makes some sense that a needlessly complicated model would be more profitable than an economically sane one.

I wish this is something I could study rigorously, but good luck getting proprietary research methods released from enough major players in even a small niche market as emerging market sovereign insurance. You would have to use a vastly simplified lab setup where you throw useless information at people and see if its perceived importance becomes self-fulfilling. (I think this has been done actually but I'm too lazy to look up the reference.)

And the alternative is that there really is some wisdom of statistical crowds and statistical methods. It certainly looks like there is in the netflix study, because I can't think of a similar self-fulfilling mechanism that might be in effect. (A recommendation system will of course determine what movies people are more likely to view, but not what they will think of them, and none of the proposed systems are actually used yet anyway.) That's pretty darn cool too.