Monday, June 8, 2009

More in the 'Road to Serfdom' Vein

A good sample of conservative worry, and the values which ground it.

Monday, June 1, 2009

Behavioral taxation

Awhile back Ross wrote that "efficient injustices will be harder to notice than inefficient ones" in reference to a New York Times article that documented how hikes in highway tolls meet less resistance from E-Z Pass drivers. I believe that the danger, as Ross sees it (I think), is that government could extract an unfairly high toll and the political opposition would be muted because no one would perceive it. Of course the danger is not in E-Z Pass which I think everyone, even conservatives, can agree is overall a good thing, but in opaque forms of taxation. The more complex the tax system is, the more difficult it is for citizens to perceive whether their overall tax burden is being raised or lowered. Just as credit card companies or banks insert all sorts of hidden fees that most users don't really perceive, the IRS could do the same thing.

I was reminded of Ross' post because of a recent blog entry on the Tax Policy Center blog titled: The Benefits of Opacity. The post cites the E-Z Pass research that Ross cited, but also some behavioral economics research which shows that people do not factor in sales tax in their purchasing decisions if the tax is not included in the sticker price of the item. However, as you might guess from the title, it comes to the opposite conclusion that Ross did. Taxes often create inefficiencies because they will distort the behavior of a rational economic actor. Fortunately, humans aren't rational economic actors, so if they do not understand how a tax works, or they do not bother to inform themselves about it, it will not distort their behavior as much. So a poorly understood tax is an efficient tax because it does not distort behavior. So it could be that the extraordinary complexity of our tax system is partially a good thing because most people do not understand it. By failing to understand it, people won't be as influenced by it, and it will be a smaller detriment to efficiency than standard economic theory would predict.

One problem with the TPC's analysis is that it doesn't explore the possibility that people think their marginal tax burden is higher than it actually is, and therefore inefficiently overreact to taxes. Joe the Plumber did not understand the difference between revenues and profits and so he thought that Obama's tax plan would raise his taxes if he bought a plumbing business with $250,000+ in revenues. He claimed that he was considering not investing in the plumbing business if Obama was elected. I wouldn't be surprised if there aren't many others like him.

Nonetheless, I think the TPC is correct about the overall trend and on the whole poorly understood taxes are probably more efficient. However, as I'm sure Ross will point out, efficient is not the same thing as just. Intentional tax complexity is no different than willfully tricking the citizenry into paying more in tax revenue, which raises some serious philosophical concerns. Also, even if the citizens do not completely perceive the taxes, they still effect the budget constraints of consumers. The additional tax revenue has to come from somewhere in the system, and if government tax revenue as a percentage of GDP goes up then ceteris paribus there will be fewer dollars spent in private commercial transactions.

Thursday, May 28, 2009

A Black President

The significance of a black president lies precisely in the fact that most people don't feel its significance. Thus I know it's an important sign in the history of this country, but I'm too young and too white and too West-Coastern to really get it. And it's because there are so many people like me that the sign is important, that it is not just a fluke. Lots of people voted for Obama, and many of those people did it because they didn't particularly care about his race; his election would not have the same significance if things had been otherwise, if it had in some way been forced on us (as with some of the black officeholders during Reconstruction). All that said, it is at this point more a sign than a cause of such sentiment; whether it will have causal force, and in which direction, depends on Obama himself.

Sunday, May 3, 2009

Are student loans spurring a tuition bubble?

The theory of federally subsidized student loans is that it allows more people to go to college. It probably does, but primarily it allows people to pay more for college and take on more student debt. Sound at all familiar? Housing and higher education are very different goods, but I think the price of both have been unreasonably inflated in recent years. And it cannot be a coincidence that people have had extraordinary levels of credit extended to them to purchase either a house or an education. Obviously the extension of credit for bubble-priced homes was ill-conceived, and I'm starting to wonder whether extending tens of thousands in loans to people in their late-teens or early twenties is also problematic.

The problem is that college tuitions at elite private schools just keep rising, no matter how ridiculous they become. At some very wealthy colleges, like Swarthmore, the high tuition functions more like a progressive tax system. The college charges a very high tuition, but only half the students (those who can afford to) actually pay it. Half the students in the school pay substantially less than the tuition, and the full tuition is more like a top tax bracket than actual price. However, at the many private schools that have high tuition but fewer resources for financial aid, the high tuition is more like a high price. Financial aid is available, but its resources are limited. Students who want to go to these private schools, but don't have the resources, typically borrow to pay for college, and the easy availability of loans for tuition makes it possible for these schools to charge the outrageously high tuitions that they do.

This leads me to wonder whether there is a bubble. The earnings premium for getting a college degree is substantial, and the increase in lifetime earnings and quality of life more than make up for the price of college on average, even at the most expensive colleges. But that's weighing an elite college against nothing, I wonder how it stacks up in a cost-benefit analysis against in-state tuition at a state school. My guess is that only the most elite and wealty private schools stack up well, and that the vast majority of private schools do not. But more importantly I wonder whether students, in the pursuit of a more prestigious degree, are taking on large debt burdens in order to overpay for their educations. Is it really worth $40,000 a year to get that private college education? Or are students overpaying for their education in order to get a degree with a certain prestige?

So, if there is a bubble in certain types of higher education, how could it plausibly crash? Well my guess is that what would happen is that some students would start to balk at being asked to take on tens of thousands of dollars of debt and attend communtiy colleges instead. This may start as a trickle, but as more and more strong students opt out of expensive elite colleges in favor of more affordable options, the signalling power of expensive colleges will wane. When that happens, the bubble will crash and some private colleges will find that there is no longer a market for their services at the prices they are charging. The other option is that the crash could come from the loan side. We could reach a point where both public and private lenders prefer to give loans for public college tuition rather than private just because it's more likely to be paid back. If the loans for expensive private education dry up, the demand for them will dry up also, and very fast. I'm tired, but later on I'll post about how this might effect academia.

Monday, April 13, 2009

A Serious Question

Would you rather be Tiger Woods or Kenny Perry? And kids growing up today, which are they raised to admire?

Tuesday, April 7, 2009

Tax distribution

The CBO just released their most recent data on distribution of tax burden. Above is a graph of what percentage each income quintile contributes to overall federal revenues (including income, social insurance and other taxes). So 5 represents the top 20% of households in terms of income and 1 represents the bottom 20%. One thing that stuck out at me is that the bottom 40% of households contribute less than 5% to federal revenue. That's partially because we have a somewhat progressive tax system, but mostly because the top 60% of households have way more money to be taxed. Still, people in the bottom 40% still pay considerable taxes as a share of their income (10.2% for the 2nd quintile), and paying those taxes requires some substantial marginal sacrifices for people in the bottom 40%. And yet, their contribution to the federal budget is just laughably small, it adds up to approximately $125 billion. That's about half of what we spend on interest on the federal debt, and less money than the federal government loses in revenue each year with the counter-productive tax exclusion for employer-based health coverage (~$160 billion). So here's a policy idea: get rid of the employer-based health insurance tax exclusion and offset the increase in tax burden by refunding all taxes (income, social insurance, other) paid by the bottom 40% of taxpayers. This might not be a good idea for reasons Ross has discussed way back in the day, but I still think it's worth considering.

Tuesday, March 31, 2009


In the comments of my previous post, Ilya makes the excellent point that data is really the only way to understand reality in a comprehensive way. I won't try to summarize his points, if you're reading this, you might as well just read them (they're short). Another point I'd add to his points in support of quantitative analysis is that data is the most free of human bias. It is the least likely to be affected by selective memory, prejudice, excessive optimism, or any other common human failure. In any case, although I agree with most of what Ilya is saying, I do think it is worthwhile to acknowledge the drawbacks of quantitative analysis of human affairs. The most obvious is that human events actually have to be quantified. That is sometimes either difficult to do, or just plain silly. It's easy and sensible to quantify how many points Kobe Bryant scores in a season, it's harder to quantify how he impacts the offensive production of his teammates, and it's silly to try to quantify how he impacts team chemistry and morale. That's a rather clear cut example, but this type of thing pops up everywhere as quantitative analysis becomes more popular. It's easy and useful to measure how much a degree at Swarthmore will improve one's earning potential, it's pretty silly to try to measure how 4 years at swarthmore will improve one's ability to appreciate literature, philosophy, and other intellectual hobbies. I could go on, but the basic point is obvious. In many decision-making situations for both individuals and society, there are many different factors to consider. Some of those things are quantifiable, some of them are not. The tendency for both society and individuals is to make decisions based upon the quantifiable factors, and not to give due weight to those factors that cannot or should not (because people do try) be quantified. So people take jobs that pay more even if they will enjoy them substantially less, teams overpay for people like Allen Iverson who scores a lot of points but saps the offensive production and chemistry from his teammates, and schools are shaped by high-stakes testing which force them to produce quantitative results in their students potentially at the expense of non-quantifiable learning. The impact of these biased decision-making processes often aren't felt until the decisions are irreversible.

Quantitative analysis is an important tool and it always will be. But would it be such a bad thing if people's decision-making stopped being biased so heavily in favor of quantifiable factors? I'm generally a big advocate of numbers, and I instinctively trust them more than anecdotal narratives or unsubstantiated theories, but it's important to recognize our own biases.