You are currently browsing the category archive for the ‘Modeled Behavior’ category.
Via Mark Thoma comes an interview with economist Arindrajit Dube about his research on the minimum wage:
In an interview with The Real News, Arindrajit Dube, labor economist and Assistant Professor of Economics at University of Massachusetts, said that increasing the minimum wage in some areas has not reduced jobs as expected by the conventional theory.
Dube said the conventional wisdom surrounding minimum wage comes from research done before the early ‘90s. … Dube told TRNN that around the early to mid ‘90s some economists realized these studies were badly flawed, and began looking at local evidence instead of just national evidence. The famous work of labor economists David Card and Alan Kruger looked at the border of New Jersey and Pennsylvania when New Jersey raised its minimum wage. Within a year, he said, not only had employment in New Jersey not decreased, it had actually risen in some groups.
He said the report received strong criticism from the economic community, but Dube’s studies apply this technique across borders of all the states, over a twenty year period to track the effects in many cases, and for a much longer period.
Dube sort of creates the impression here that the current conventional wisdom is based on outdated research, which is not the case. While the conventional wisdom may have been founded on research from before the 90s, the majority of post Card and Kruger research, what has been called “the new minimum wage research”, supports the notion that minimum wages decreases employment. For instance, a 2006 paper from David Neumark and William Wascher summarizes the new minimum wage research like this:
The studies surveyed in this paper lead to 91 entries in our summary tables (in some cases covering more than one paper). Of these, by our reckoning nearly two-thirds give a relatively consistent (although by no means always statistically significant) indication of negative employment effects of minimum wages—where we sometimes focus on results for the least-skilled—and fewer than 10 give a relatively consistent indication of positive employment effects. In addition, we have highlighted in the tables 20 studies that we view as providing more credible evidence, and 16 (80 percent) of these point to negative employment effects. Correspondingly, we have indicated in our narrative review that, in our view, many of the studies that find zero or positive effects suffer from various shortcomings.
This is consistent with their 2008 book, Minimum Wages, which I don’t have on me at the moment.
In any case, I have not read Dube’s paper but it looks like an interesting extension of Card and Kruger. In the meantime, the majority of the new minimum wage research supports the hypothesis that the minimum wage increases unemployment.
As I’m sure most are familiar with by now, firefighters in Tennessee recently refused to put out a house fire because the family hadn’t paid the service fees, and despite their offers to pay them, the fire company let the house burn to the ground. While this is a terrible tragedy, there are several important lessons about public policy here, and -as the title to this post suggests- there are several lessons that aren’t here, which people seem to believe are.
Like Paul Krugman I do think that this case presents a somewhat apt example for the moral hazard of health care, and that if you can’t credibly refuse to deny a service to someone it makes sense to force them to pay for it in advance. The question is, can they credibly deny the service?
On the one hand I think the townspeople who hadn’t been paying their service fees probably got up to date pretty quickly after this incident, as the fire company very much demonstrated that they can credibly deny service.
On the other hand, given the public outcry against this and the fact that a neighbor’s house caught on fire as a result, perhaps the fire company has learned that they can’t credibly deny service.
An important question remains for the locals in that area: having watched a neighbor’s house burn down, are you prepared to wager that your other neighbors have learned their lesson? This gets at an underappreciated public goods aspect of the issue that is aside from clear danger externality: nobody likes to watch their neighbor’s house burn down. It is a giant, unignorable, tragic, and heartbreaking spectacle that I’m sure every neighbor of that Tennessee home would have, ex post, willingly paid to avoid. This means that aside from danger externalities, there is an additional reason why fire insurance would be under-provided. It may be that given a level of mortgage debt and risk preferences, the service fee the fire company charged was not worth it even for a rational homeowner. But taking into consideration their neighbor’s desire to not see the house burn, it is likely inefficient for them not to have it. This suggests a mandate or tax.
On the other hand, I strongly disagree with the contention that this tells us anything about libertarianism. Is a voluntary provision of public services more libertarian than a mandatory provision? Yes, on the margin. But saying this is a “failure” of libertarianism, or that libertarian thinking is to blame, is like blaming the huge debts of the U.S. Postal Service on libertarianism because the post office isn’t a completely free service paid for by taxes. It’s also like blaming the failure of Fannie/Freddie on libertarians because they were GSEs rather than fully government run. For many libertarians these may be second, third, or fourth best outcomes, but for far more government optimists they are first or second best outcomes; this is a failure of government optimism.
In addition, I have to believe that the fire company was simply behaving in accordance with the law, and that they weren’t responding to the fire because they weren’t allowed to. At the very least they had no monetary incentive to go put that fire out (one would think they had plenty of moral incentive, but apparently that wasn’t enough). In contrast, had they had been a for profit company free to behave in a profit maximizing way, they would have certainly gone to put out the fire as they could have perfectly price discriminated. This is a “sinking ship” scenario sometimes discussed in price gouging contexts, as discussed here by Richard Posner:
Suppose a ship is sinking, and another ship comes along in time to save the cargo and passengers of the first. The second ship demands, as its price for saving the cargo and passengers of the first ship, that the owner of the ship give it the ship and two-thirds of the rescued cargo, and the captain of the first ship, on behalf of the owner, being desperate agrees.
Clearly, this is a highly profitable and pareto improving scenario. Posner informs us, however, that such contracts are unenforceable under admirality law and common law, so this might limit a private fire company’s ability to profit here.
Nevertheless, unlike a government run fire company, a private one would have plenty of incentive to put out the fire.
Yesterday I wrote this:
If I had to play homeowner psychoanalyst I would guess that homeowners with a strong preference for homeownership saw cap rates were changing and believed house prices were heading towards a permanently higher plateau that would permanently price them out of homeownership. People who would want to buy houses in the future but were currently renting had this fear as well and rushed into the market. Risk aversion here thus did not lead to selling when prices rose as a simple model might predict.
I was drawing on the work of Todd Sinai and Nicholas Souleles who have shown that housing works as a hedge against rental volatility. My thinking was that because owners often have strong preferences for home-ownership and are not indifferent between renting and owning, that those who planned to buy a house in the future would see buying a house now as a hedge against house price risk, much in the same way as the Sinai Souleles model. Well yesterday after I wrote that I came across a brand new paper that makes a very similar argument:
Our contribution is to focus on the importance of ownership as a hedge against future house price risk as individuals move up the ladder. We use a stylized model to show that increasing house price risk acts as an incentive to become a home owner earlier in the life-cycle and, once an owner, to move more rapidly up the housing ladder.
Increases in volatility are shown to increase ownership and to increase the quantity of housing wealth conditional on ownership in earlier periods of the life-cycle. We then establish that these relationships hold empirically using panel data on families in different geographic markets in Britain and in the US.
The authors use data from the UK and US to provide empirical support for their model. This is an under-explored causal mechanism for the bubble: house price risk went up, people bought homes to insure against that risk, which drove prices up, which increased perceived house price risk, etc. The cascading nature of this is clear, and it’s not hard to see how this could create a bubble.
So housing risk makes people want housing even more. I’m not sure if this will comfort or aggravate people like Felix Salmon and Ryan Avent who have been arguing that households are too risky in their housing consumption/investment decisions, but it should help explain why what looks to them like overly risky behavior is in fact caused partly by risk aversion.
However, this explanation for the behavior does offer a potential solution. Local REITs could be created for very small geographic areas designed to help young households who want to insure against price increases in a specific neighborhood but do not yet want to take the large risks of buying a house. They could even be metro area REITs that are heavily weighted in a particular submarket.
If Felix and Ryan want to get households to stop overleveraging themselves in housing debt, maybe their best option is to start looking for venture capital.
Does education really teach you anything or is it just a status symbol that allows employers to sort individuals by pre-existing traits? Bryan Caplan has been debating this issue with his former professor, and now his coworker Tyler Cowen is joining in as well. All-in-all I find Cowen more persuasive, and think that for most countries, more education is probably better than less. However, I do want to question one aspect of his story. He argues that education allows people to reinforce their self-perceived characters in positive ways. Here is how he briefly puts it in his debate with Caplan:
Signaling models are important but they are not the only effect and of course a lot of signaling is welfare-improving for reasons of screening and sorting and character reenforcement.
It is only quickly mentioned here, but discussed at greater length in his most recent book, Create Your Own Economy (retitled, expanded, and in paperback here). The story is that education creates a framing effect which allows people to have a more positive narrative about themselves that induces better behavior. People see themselves as a college educated person, which makes them behave as they perceive college educated people to behave; perhaps reading more, working harder, attempting to be more cultured.
I completely agree with Cowen that the framing effect, or character reenforcement, of education is real. My problem with it is that it may be zero-sum, or even negative sum. It may be that the framing effect of college is simply to shed oneself of the burden of the negative framing effect of not going to college.
It is certainly a common refrain in TV, movies and yes, in real life, for non-graduates to explain some negative circumstance or inability to overcome a problem by saying “look, I didn’t go to college,” and then “I can’t…” or “I don’t…”, or some other restraint that fact places upon them. Absent such a framing effect they may believe their odds at achieving something are better, and so they would be willing to work harder at it.
One way to view this is that college sorts people into high-skilled and low-skilled pools, and then individuals perceive themselves as similar to their cohort. The perceived differences in the average skills of these pools is exaggerated because some of college is signaling, which allows those in the high-skilled pool to view themselves as higher skilled than they really are, and vice-versa for those in the low-skilled pool. For high-skilled people this induces positive behavior and for low-skilled this induces negative behavior.
This sorting would increase the disparity of outcomes between high-skilled and low-skilled individuals. Depending on the distribution of skills, a higher percent of the population going to college may increase the disparity in average skill levels between pools, and therefore the wider the disparity in framing effects. (Has this affect exacerbated growing income inequality?)
I’m completely prepared to change my mind about this, but as far as I can tell while it’s true that college creates framing effects and character reenforcement, it’s unclear whether the net effect of this is to induce overall better or worse behavior, and whether the positive effect on college graduates outweigh the negative effect on non-graduates.
I know it when I see it
~Potter Stewart, 1964
Can it really make sense to vote for the opposite party just because the economy happens to be bad?
I might count myself among this 10%. (I say, "I might", because I do not really know myself. Who does?) The "kick the bums out" mechanism leaves a great deal to be desired, though. It is not obviously better to exchange our current bums for new bums as a response to the behaviour of past bums. In this case, however, it is likely that a bum-swap will deliver divided government, a prospect that warms my anti-partisan heart.
That’s one explanation. Another is this. Voter’s don’t really know what they want. Indeed the one’s that matter – swing voters – don’t know even know what it is that they might want. That is, they are ignorant of the policy choice set. Whatever they want, however, they know this ain’t it.
Now, broadly speaking there are two possibilities
1) The horrible state that we are in today is unrelated to government policy
2) The horrible state that we are in today is related to government policy
Suppose that the world is in state (1). Then switching parties does nothing to relieve the horrible state we are in. However, its also fairly low cost, because government policy isn’t related to the horrible state we are in. It might be related to other stuff, but by assumption that stuff is second order.
Suppose the world is in state (2). Then there is a good chance that switching parties will switch policies, which in turn might result in an improvement in the horrible state that we are in.
Thus regardless of the true state of nature I should switch parties.
Now, my democratic readers will be quick to chime in that this analysis is incomplete. It could be the case that the government policy is related to the horrible state that we are in, but that in fact our current policies are making it less horrible.
This is just another way of saying things would be even worse if Republicans were in power. I can easily see why partisan democrats would be inclined to believe that.
However, suppose that I am just a regular voter who hears Dems say things would be even worse without them and the GOP say that things are as bad as they are because of the Dems. What am I to think?
Well that depends on how bad things are compared to the worst they could possibly be. The closer things are to the worst things can possibly be the less likely it seems that the Democrat’s argument is correct.
Note that I don’t have to believe that the Republican argument is correct in order to vote for them. State of nature (1) makes a vote for the GOP low cost. I only need to believe that the Dem argument is likely wrong.
Now lets check the tape
The current state certainly looks worst than anything ever before. At least anything any of us are old enough to remember.
That makes it hard to buy the Democrat’s argument. And, if you don’t buy the Democrat’s argument then you should vote Republican, regardless of whether you buy the Republican argument.
This is of great interest to me because I am fascinated by the fact that Democracy seems to be a highly effective form of government despite an almost necessary implication that policy will be determined, or at least largely influenced, by the least knowledgeable and indeed least policy interested people in society – swing voters.
So my instinct is that there is an invisible hand at work. A mechanism that leads people towards decent policy even when they have no idea where they are going. Throwing the bums out just might be it.
Arguing for more generous mandatory vacation laws, Ezra Klein writes:
Which goes to the reality of the situation, which is not that workers and employers “flexibly choose an arrangement that works for them.” Employer-employee relations are rarely so idyllic. Broadly speaking, employees with the power to demand more paid vacation do so, and employees without the power to demand more paid vacation get less — or in some cases, no — paid vacation. A law guaranteeing paid vacation would primarily tilt the playing field toward low-income workers, rather than against them, as is the case now.
The problem with this is defining the employer/employee “power” in terms of vacation setting only. If an employer has the bargaining power negotiate a deal where the employees total compensation is less than their marginal product of labor , then they will have the power to negotiate lower wages when laws mandate less days of work. Or they can just take those hours back by negotiating longer work weeks, shorter breaks, working harder, or something like that. Unfortunately you can’t write a law mandates wages be equal to marginal product of labor, and there will be unintended consequences of any law that attempts to restrict hours whereby employers cash in their bargaining power in some other form.
UPDATE: And if you think only zany libertarians believe that labor markets work, here is Yglesias arguing basically the same thing several years ago.
No, not that Family Ties… A new paper from Alberto Alesina and several coauthors posits that strong familial relationships increase the demand for restrictive labor regulations:
Flexible labor markets requires geographically mobile workers to be efficient. Otherwise, firms can take advantage of the immobility of workers and extract monopsony rents. In cultures with strong family ties, moving away from home is costly. Thus, individuals with strong family ties rationally choose regulated labor markets to avoid moving and limiting the monopsony power of firms, even though regulation generates lower employment and income…
There is a lot of intuitive appeal to this theory, and the authors provide empirical support as well. This is reminiscent of work which argued that being married reduced mobility and thus allowed employers to bargain for lower wages, although I can’t remember any particular papers on this to link to at the moment.
My question is whether this phenomenon partly explains the political persistance of the mortgage interest tax deduction? Baseless cosmopolitan liberals like Ryan Avent, Matt Yglesias, Felix Salmon, and myself argue that a downside of homeownership is that it increases geographic immobility, which is desirable in a labor market. Thus it is in societies interest overall not to encourage homeownership with this tax deduction. But if 51% of people won’t ever want to move regardless of local economic conditions because of close family ties, then we could end up with this policy even though it is against everyone’s economic interest. This policy magnifies the labor immobility effects of family ties by encouraging more people to undertake large sunk costs in housing than otherwise would. In addition, regardless of family ties, people will support the law once they become homeowners, which expands the constituency beyond those with family ties.
More reason to be skeptical that the mortgage interest tax deduction is going anywhere anytime soon.
Jodi Beggs, aka Economists Do It With Models, argues that paternalism need not be justified by assuming irrational agents, but can simply be an efficient response to an informational problem:
Any Economics 101 course will tell you that a required condition for markets to be efficient (read, value-maximizing for society) is that consumers have full information about the products they are considering consuming. In this way, the calorie-labeling legislation is helping to push the fast food market in the direction of efficiency as much as anything else. What’s so behavioral-y about that?
One counter to this is that markets should be supplying the amount of information that consumers prefer, and that the reason we don’t observe a lot of menu labeling and other information from restaurants is because consumers don’t want to know. Of course, you could argue that they don’t want to know only because of what they don’t know…Wait, what?
Ok, bare with me. Pretend I had a sealed envelope that contained a letter from someone telling you telling you exactly why they hate you. But say you believe that everyone who you care about doesn’t hate you, therefore you assume it’s from someone you don’t care about, and since you don’t want the annoyance of reading hate mail from someone you don’t care about, you choose not to open the letter. But, say that letter is from your wife, who secretly hates you. Well you would want to know why your wife hates you, but since you believe your wife can’t possibly hate you, you won’t get information you want. Basically what I’m saying is that your current information set determines your demand for information.
So what does this have to do with menu labeling regulations? If we assume markets are working, then the level of information we observe is the amount demanded by consumers, which efficient. In this case menu labeling laws would make people worse off by giving them information they don’t want. That is unless the amount of information they are demanding is based on their assumption that restaurant food is kind of unhealthy, but not as unhealthy as it really is. If they had any idea how bad it was, they would want to know. In this case menu labeling laws could make people better off.
Determining the source of the lack of information is critical to knowing whether menu laws are efficient or not. This is especially important because the amount of information can affect demand, which contra Jodi, can change the choice set. She writes:
…the consumer has exactly the same set of choices available to her regardless of whether calorie counts are on the menus or not. Because of this feature, it’s hard to argue that this sort of legislation is significantly bad for anyone- here, the worst-case scenario is that some people keep eating unhealthy food but are no longer blissfully ignorant and instead feel guilty.
But what could happen is that when people are no longer able to be blissfully ignorant, which they prefer, they consume a healthier but lower utility set of products. This in turn could change restaurant supply decisions, which would mean a different choice set.
So what is it: is our demand for ignorance efficient, or is our ignorance causing us to demand an inefficient amount of ignorance?
Daniel Indiviglio at the Atlantic has another piece on how credit is bad for poor people. Like Daniel’s previous writing on credit and the poor, his reasoning amounts to partial partial equilibrium analysis that misses the big picture and really doesn’t capture the effects of credit on the poor.
One way to clarify the various effects he introduces is to think of them in terms of labor supply curves. First, he argues that credit makes people feel richer, and so decreases wages:
Credit pacifies those with lower incomes to make them feel like they’re better off than they actually are. If you can use a credit card to buy an iPad or new shoes, then you are more satisfied than you would be based on your income alone… The more content you are, the less need you’ll feel to try to increase your income.
So credit makes people feel wealthier, and because leisure is normal good that you consume more of when you’re wealthier, people want to consume more leisure. Therefore the labor supply curve shifts left as people choose leisure over work, and thus wages go up. That part is missing from Daniel’s story. Importantly, it’s not obvious that income goes down in this scenario; it depends on the slope of the labor demand curve.
Ok, so maybe that’s the wrong way to frame it, and credit simply works as an income multiplier instead as a wealth effect. This means individuals get $1.2 worth of consumption from every $1 of income, which increases the returns to income. But this would increase the value of work relative to leisure, which would shift labor supply right, which would decrease wages but increase hours worked. Again, the effect on income depends on the labor demand curve.
So from the start, the effect on income are unclear and depends on the slope of the labor demand curve and the relative sizes of the wealth and income effects. But then Daniel makes the analysis even less clear by undoing the effects he just established:
The poorer you are, the more expensive your credit, but the more credit you’ll feel like you need…And they’re paying relatively high interest rates, which further eats into their relatively lower income, reducing their wealth potential.
So he just told us that credit was increasing peoples perceived wealth and income, but now he’s saying it’s lowering their actual wealth and income. Are people completely unaware of this? Over their entire life that is a highly unreasonable assumption. In fact, most people surely understand from the get-go that borrowing money is not magically making them richer, and that it must be paid for in the long-run. I sincerely hope Daniel isn’t going to argue that most borrowers think they won’t have to pay it back.
So people are choosing to consume more now rather than later, and while ability to smooth consumption should increase their utility, it is far from clear using just the effects that Daniel has laid out whether it increases or decreases their lifetime income and wealth. But when you include other effects, like the increased ability to borrow and invest in positive NPV investments like, say, a college education, it becomes obvious how credit could increase lifetime wealth and income of poor people. Wealthy people don’t need to borrow to make positive NPV investments like that, but poor people do. So the benefits disproportionally benefit poor people.
As I’ve previously written, more credit not only increases investment in college, but also high school and grade school. Given the importance of credit in allowing poor people to make important human capital investments that increase lifetime wealth, income, and well-being, Daniel is going to have to work a lot harder to make a convincing case that credit is bad for poor people. His partial partial equilibrium analysis is not even close.
[Update: broken link fixed]
In particular legends Joey Chestnut and Kobayashi are pretty fit guys
But of course its just calories-in minus calories-out. No equilibrium modeling required.
Both concede it is difficult to find a mechanism that explains the relationship, which is also why I am skeptical. Ezra offers up this theory:
One theory that’s made intuitive sense to me is that the problem is not just the demand for credit but the accompanying supply of idle money. When someone making $25,000 a year gets a raise, they spend it. When someone making $2,500,000 a year gets a raise, they invest it. And the more money there is sitting around, the more demand there is for high-yield investments, which means the more reward there is for people who can invent new investment vehicles with high yields. Hence, you have explosive innovations in weird financial instruments that look good for a while because the risk is underpriced but end up making the system more fragile when their risks come clear and everyone flees.
This doesn’t seem correct to me. When they demand for investments goes up the price should go up, and since price of an investment is the inverse of the rate of return, the market rates of return on investments should go down . So if there are a lot of people with a lot of money to invest they bid down the required rate of return on a given investment.
This means that when Ezra says that the more investment demand there is the “more reward there is for people who can invent new investment vehicles with high yields“, I think he should be saying low yields. For example, if an investment can generate a 5% return, then when required market rate of return is above 6% the demand is zero. In contrast, a 10% investment will be in high demand when the required rate of return is 6%. So when investment demand increases, and thus rates fall, the marginal investments that become profitable should be increasingly low return, since high return investments were profitable in the first place.
Let me give a concrete example. If I have a factory that costs $1,000 to build and generate a $20 a year in profit, then it has a 2% return on investment. I will only be able to find investors for this factory when the market rate of return is 2% or lower. In contrast, if I have a factory that costs the same amount to build but generates $200 a year in profits, then it has a 20% rate of return, and I will be able to find investors as long as the market rate of return is 20% or lower. As market rates lower, factories and other investments with lower rates of return should increasingly be supplied.
I’m not completely averse to Ezra and Krugman’s proposed relationship between income inequality and recessions. But I don’t think this mechanism is correct, and I will be skeptical until I hear a believable story.
Daniel Indiviglio has a strange post at The Atlantic that I have been trying to ignore, but the website editors there keep putting it on their front page in a box called “Things you might have missed”. His argument is that credit should be criminalized because that would make the economy more stable. Daniel recognizes that the effects would be devastating in the short run, but claims that eventually it would only mean that the economy grows “a little slower”. This is not just my ungenerous interpretation:
It would involve an extraordinarily difficult transitional period, including massive job losses, deflation, and a deep recession as the government and population adjust. But if Congress managed to embrace a credit ban, we would end up with an overall economy that grows a little slower, but is incredibly stable. All that systemic credit risk? Gone. That reward would be well worth the cost. [Emphasis mine]
Would it really be worth the cost? Let’s consider what criminalizing credit would really mean. In reality it would mean lending would be pushed into the black market, because people would not be ridiculous enough to abide by a law that prevents borrowing to make investments with positive net present value, or making loans with a positive expected return.
To the extent that people don’t circumvent the law, a credit ban would mean a lot of terrible growth destroying stuff in the long run, not just during a “transitional period”. One growth destroying impact would be that people couldn’t borrow money to go to college, which would gradually erode the human capital stock of the country. People also couldn’t borrow money to buy cars to drive to their jobs, which would mean settling for worse jobs that are closer by, thereby inefficiently distorting the allocation of labor. They also couldn’t borrow money to make simple investments in durable goods that would save them money in the long run, like buying a washer and dryer instead of spend money at a laundromat.
It would also mean that when someone suffered a temporary and unexpected income shock that wiped out their savings they would have to begin selling off their possessions instead of using credit cards to make it through. Of course that’s only if they have valuable possessions to sell. The evidence in developing countries suggests that when poor families have no or limited access to credit, they are more likely to respond to income shocks by pulling their children out of school and putting them into child labor. So we’d probably see some child labor, which also represents a massive destruction of human capital.
Of course rich people wouldn’t see their access to education or mobility limited, nor would their resources to weather income shocks be noticeably diminished, because they have lot’s of wealth and savings to draw on. Thus banning credit would mostly hurt the poor, and thus would significantly increase income inequality. The ability to borrow money is much more important to low income people, and for so many it is an absolute necessity if they are to pull themselves out of poverty.
So no, even in theory, even on paper, we should not criminalize credit.
Karl responded to my request for a GM bailout proponent to analyze a BP bailout, and he highlights some crucial differences between the two scenarios. By the commodity nature of oil, a BP competitor would know they could take over BP wells and immediately keep pumping, and not face as much uncertainty as a potential buyer of a GM plant would. This is a big point in favor of GM vs BP.
Karl also argues:
…all three US manufacturers were tied together through the supply chain. This not only magnifies the intensity of a collapse but also has unique consequences for labor. US manufacturers used UAW labor while foreign manufacturers do not. This makes the two imperfect substitutes and would have contributed significantly to a difficult transition.
With respect to the supply chain impacts, that depends on whether the bankruptcy disrupts the supply of oil enough to effect prices. A disruption in supply chain of oil would be much worse than one in autos because oil is non-durable, so you can’t just use yesterday’s oil again today like you can with a car, and because it’s an input into almost every industry and product. James Hamilton has presented persuasive empirical evidence that oil shocks can cause recessions, while no such evidence exists for auto shocks. Regarding the UAW issue in this paragraph, I’m not sure why defunct union status would add frictions to hiring for whoever buys GM’s factories and brands. I suspect that former union workers would probably pretty quickly accept new jobs at lower wages. It’s not like the alternative employment prospects in the rust belt present many alternatives.
He also argues that because oil is a classic perfectly competitive product, the market should respond without problems:
BP on the other hand produces the classic perfectly competitive product. Its completely fungible. Its traded on an exchange. It has a healthy futures market. Its exactly what a textbook economic product should look like. Shortages will be sorted out in the markets, prices will respond and equilibrium will be restored.
As Hamilton shows, though, high oil prices can cause serious macroeconomic problems. The key question, I think, is whether a BP bankruptcy has any chance of disrupting oil production enough to cause a significant spike in prices. As Karl points out, the commodity nature of the product suggests that new owners could keep pumping and selling oil without much of a problem. However, I can imagine scenarios where supply would be disrupted. When Texaco went bankrupt in 1987, credit and supplies were being cut off:
In an affidavit filed in a Texas appeals court, Texaco outlined in detail the pressures it was under. Chase Manhattan had demanded that Texaco maintain new minimum balances in its accounts before the bank would transfer funds to satisfy commercial obligations. Worse, Manufacturers Hanover Trust had canceled a $750 million line of credit.
At the same time, some of Texaco’s suppliers were refusing to do business, or setting tougher terms. According to the Texaco affidavit, Venezuela’s state-owned oil companies had at least temporarily stopped pumping oil for Texaco. (Venezuela denies that it has cut Texaco off.) Southern California Edison started requiring Texaco, its largest customer, to pay its electric bill every week.
The BP bankruptcy would be inherently more uncertain because litigation damages have a much higher potential upside than Texaco’s worst case scenario of $10.2 billion. As Andrew Ross Sorkin reports, the final tally for BP could possibly (although unlikely) be in the hundreds of billions. In addition, we have an administration and a populist electorate that has shown itself willing to intervene in bankruptcy preceding. All of this could deter mergers or buyers. And selling off assets may be complicated by long-term leases, regulatory approval and other legal matters. Given all of these uncertainties, and difficulties in merging of divesting to a solution, I am not convinced that a disruption in production is not a possibility.
Karl also points out correctly that we were in the middle of an extreme credit crunch in 2008, and we aren’t right now. I grant that that is the strongest argument for the GM bailout. So my question is this: if Greece defaults or some other disaster like that occurs, and we begin once again to teeter on the brink of a credit crunch, would there then be a case for a BP bailout?
Narayana Kocherlakota, the newish president of the Minneapolis Fed, has a new paper out discussing how to devise an optimal bank risk tax. The creativity of his solution shows the value of having a theoretical economist, or at least a creative economist, heading up the Fed.
He begins with the observation that bailouts are inevitable, there is no mechanism to prevent this, and that any bailout possibility is going to make banks behave riskier. An ideal solution forces banks to pay for the implied guarantee at a risk adjusted rate. This is in contrast to something like deposit insurance, where every bank pays the same fees regardless of how risky or safe. So how do you do this?
Kocherlakota suggests that every financial institution must issue a “rescue bond”, for which the coupon is equal to 1/1,000th any bailout funds paid to that financial institution. In most years this bond will pay zero, and only when they receive bailouts will owners of them receive any money. Thus the price of the bond represents 1/1,000th of the discounted expected value of the transfers of the funds to the bank, and so the government should charge the financial institution a tax equal to 1,000 x the value of the bond.
It sounds like a good idea to me, but I worry about a couple of things. First, there appears to be a recursion problem. If a bank is an expected $10 billion short of solvent, and markets understand that, then rescue bonds will demand that the banks be taxed $10 billion. By the time this is knowable, the government can’t really credibly take the $10 billion from the bank, because then they will just be $20 billion in the hole, and they will be taxed $20 billion, etc. The only way this works is if it sounds the alarm far enough in advance that banks are solvent enough to be taxed and turn their behavior around.
For instance, if the price of rescue bonds indicated 2 years in advance that AIG was headed towards insolvency and caused the government to tax them at higher and higher amounts until were forced to turn their behavior around, then it might work. However, if in August of 2008 the price of rescue bonds indicated that AIG was going to be bailed out to the tune of $90 billion by the end of October, it’s not like the government could have actually taxed them $90 billion discounted by two month. And if they did the bailout would just have gone up to $180 billion.
So to the extent that this plan works, it has to mean that the financial crisis reveals itself sort of slowly, rather than coming on quickly and unexpectedly. Then again, if it does happen too fast the government just won’t follow through on the tax, so maybe it’s worth a shot.
One complaint about congestion pricing is that it would push people out of the market. But in theory this is not necessarily so. The common sense story is that the price of driving rises, so the quantity of driving demanded decreases, and thus less people drive. But it depends how people shift their behavior in response to higher prices. There are many ways to respond to higher driving costs that don’t mean less total people travelling by car.
For instance, more individuals could begin carpooling. A new study points to an increasing trend in carpooling, and finds that gas prices are indeed inversely related to carpooling. The charts below show that the trend over the past few years has been of decreasing single-drivers, and that carpooling has been the predominant alternative to single-drivers, with the percent carpooling to work doubling from 4% to 8% (and shame on the author for not using 0% as the Y-axis intercept!):
Another way people could respond to higher prices is to drive during times when prices are lower, that is if prices are set in relation to congestion. Neither this nor carpooling represents any decrease in quantity, if quantity is taken to mean the number of people using non-public transportation automobiles for travel.
In addition, there are currently people who would prefer to drive places, but due to the time-cost of money decide not to drive there because they don’t want to wait in traffic. Driving could increase among these people.
In the end the number of people traveling by car in response to congestion prices is an empirical question, and there very well may be conclusive evidence that, contrary to the arguments I’ve presented here, the relationship is negative. I would be interested to see any such studies.
Via Ryan Avent I find this comment by Frank McArdle on congestion pricing:
“The plan put forth by Charles Komonoff to price some taxi trips and other trips off the road through price increases gets to the heart of the equity debate over congestion pricing: who gets pushed out and who gets the benefit. …Using the money to subsidize further mass transit may be the hook that gets the government to impose the system,but there can be little question that the faster trips will be to the benefit of those who can pay and those who can’t will be asked to shift their trips to other modes or other times.”
Of course this is going to be true of removing many inefficiently low price floors. When you allow the consumption of good to be rationed by non-price means, moving to price rationing means that those with the lowest willingness to pay will be pushed out of the market. For normal goods the income elasticity if demand is positive, and so income and willingness to pay will tend to be correlated, and less wealthy people will be pushed out of the market. If equity concerns you in the context of this below market price floor -which is exactly what lack of congestion pricing is, a price floor at zero- then what price floors won’t you support? Rent control? Price floors on food? Certainly hard rules against price gouging should be supported.
All the people commuting by car seem to think they will hate it. And that makes me think that they probably will. It will make the traffic flow more quickly, but it will also cost them something like $200 a month. That’s a lot of money for most people, to do something faster that they can now do more slowly for free.
There’s obviously something intuitively compelling about this argument, and the high level of unpopularity of many proposed congestion price plans does make me wonder whether the prices they are proposing are set too high, and whether planners are overestimating the time value of money in this context. In the end though, I think we would probably observe the massive unpopularity even if congestion were priced efficiently, because price controls tend to be fairly popular ideas. Price gouging and the other examples of price floors that I mentioned above are good comparisons. Even when they understand that non-price rationing will lead to things like waiting in line for gas and store shelves being emptied, people tend to be in favor of rules that try and prevent price gouging, and so presumably think they make them better off. And yet I think most economists, and certainly Megan, would agree that price gouging rules and other types of non-price rationing do in fact make people worse off despite the fact people don’t believe it.
So while I agree that in general if people believe something will make them worse off those concerns should be taken seriously, when it comes to the benefits of markets and price rationing people can just flat out be wrong.
In addition to global warming, congestion, geopolitical costs, oil spills, and health problems like asthma and allergies, we now have another externality to gasoline consumption to justify a pigouvian tax: drunk driving. A new study by Chi et al (6 co-authors!) uses data from Mississippi to show that lower gas prices are related to drunk driving related accidents. The authors claim the study is the first to examine this relationship, and is important because from a theoretical perspective the relationship could be positive or negative.
The ways that gas could inversely relate to drunk driving are obvious: lower prices make it cheaper to drive and give people more disposable income, which means it’s less expensive go out drinking and driving and people have more money to do so. In addition, the marginal cost of driving a to a farther away bar decreases. Also, higher gas prices may cause people to shift to different modes of transportation, like walking or taking the bus, which (freakonomists aside) are less likely to result in drunk driving accident.
A positive relationship is less obvious, but could result if gas prices increases enough that the negative wealth effect (more expensive gas makes you poorer) is severe enough that it creates economic hardship, which can lead people to drink more.On the face of it, the positive relationship seems much less likely than the negative relationship, and this is what the empirical evidence found in this study suggests. The chart below shows the indexed values of drunk driving accidents and gas prices.
These results increase the growing gap between the nominal price of gas and the true cost of it, and strengthen the case for a pigouvian tax… not that externalities, efficiency, or empirical realities seem to matter much in the political debate on this issue.
A caveat though: these results should not be taken as dispositive but rather suggestive. The empirical analysis is pretty simple, does not get into a really serious attempt to examine causality, and has some fairly serious omissions. For instance, the authors do not control for weather in their analysis. They mention that it would be difficult to aggregate to the monthly level, but I think average temperature would probably suffice. Second, and relatedly, they do not control for seasonality. This is pretty important in a time-series context where you are very likely to see both drunk driving and gas prices increase in the summer and decrease in the winter. Finally, and this is a more minor econometric point, they choose between a poisson and negative binomial regression models by selecting the one with the higher log-likelihood, which I do not believe is a sufficient means to determine whether there is enough overdispersion in the data to warrant the use of negative binomial over poisson. More importantly, they don’t tell us whether the use of negative binomial, or OLS for that matter, affects the results compared to poisson, which would have taken 30 seconds to determine and would tell us something about the robustness of their econometric results. Given that the relationship between prices and accidents disappears for males when the analysis is partitioned by gender, it is not hard to believe that the results are potentially not robust.
All that said, the authors do argue that nobody has empirically examined this issue before, and the results are highly theoretically believable. In fact, I find the theory alone strong enough to conclude that a relationship is likely. At the very least when thinking about gas prices we should consider that drunk driving may be yet another cost of low gas prices, and this study should definitely be enough to prompt more research into this.
The pool of teachers in this country is almost certainly less risk averse than the pool of financial company executives. One important reason for this is that each job selects for certain risk profiles in the way that employee contracts and payments are structured. Teachers are offered (up until recently) nearly guaranteed employment for life with pre-specified and certain raises. People who don’t like risk do well to become teachers.
Finance executives, in contrast, are paid heavily in stock options in order to align their incentives with their shareholders. As Murphy and Hall argue, paying employees in options means paying them with undiversified and illiquid assets, which means that the value of the options to risk averse employees is half of what the option costs the company. So the more risk hungry someone is, the more they value being paid in options, the more they will want to be a finance executive. Given this selection, it’s no wonder we end up with risk hungry finance executives. In fact, as Murphy and Hall also demonstrate, this extends beyond executives because it’s not just executives that get paid in options, but a large share of employees; 90% of stock options are granted to employees below top-level executives. This means that the financial sector as a whole is selecting for risk hungry workers.
If we think this is a problem, then there is good reason to think that it is going to get worse. Efforts to make options more aligned with long-run performance make these options less liquid, and thus more risky. In the short run this may make the financial system less volatile by incentivizing executives for long-term rather than short-term performance, but in the long-run it may make it riskier by selecting for more risk hungry executives.
There is a balance to be struck between selecting for risk-hungry and risk-averse workers. I think the education system probably needs to select less for risk aversion. But I think you can also make the case that the finance system should select more for risk aversion. My question is this: will the increase in risk from this selection problem mean that efforts to align incentives with long-term performance will actually make the financial system riskier?
1. One complaint I received about the slew of salt regulation blogging last week was, “it doesn’t matter”. Well, tell it to the New York Times, who ran a story on the ongoing battle over salt regulation in the front page of the Sunday paper. The article is well worth reading and full of many interesting facts, including an informative case study of the roll of salt in a cheez-it, and this warning about unintended consequences:
Joanne L. Slavin, a committee member and nutrition professor at theUniversity of Minnesota, told her colleagues that reducing salt in bread was difficult and warned of unintended consequences. It is an argument also made by food companies.
“Typically, sodium, sugar bounces around,” she said. “So you take sodium down in a product and then sugar a lot of times has to go up just for taste.”
2. Relevant to my previous post on the use of rewards card data by supermarkets is this story, also in the Times, about Sam’s Club and other stores creating custom tailored discounts for consumers. The article compares the technology they use to the predictive analytics used by Netflix and eHarmony. Here is one quote which is suggestive of how stores using this type of analysis and their customer’s can benefit:
“I got like $2 off the bananas, a $1 off the Toasted,” he said, referring to a package of crackers in his cart. Mr. Mayoral, 36, said the best eValues deal yet was $300 off a $1,200 television.
“I remember that day,” he said later. “I came to buy food, and I bought two TVs.”
In response to my last post on salt regulations, several commenters expressed some further points that are worth addressing. Much of it amounts to the argument that the regulations are, from the consumer’s perspective, a free lunch.
Commenter JzB argues that, for the most part, lowering that lowering salt in packaged foods will not result in significant taste changes, and to the extent it does, it can be fixed by adding salt:
So some of the salt is there for good reason, but the amounts used are overkill. I’m suggesting reductions of 25 to 50% will be transparent, or close to it. And also adjustable on the plate, if so desired.
The notion that reductions in sodium of the scale argued for by the Institute of Medicine can be achieved without significantly impacting the food as experienced by the consumer is, I believe, incorrect. In agreement with me is the actual IOM report:
Current and ongoing industry reformulation has demonstrated that substantial reductions in sodium can be achieved based on existing technology and science. However, given the need to significantly reduce the sodium content of the food supply to achieve recommended population intake levels, additional innovations and research will be necessary to secure reductions while maintaining product taste, texture, safety, and shelf life.
These innovations will be necessary because, as the study points out, companies have already been taking advantage of the low hanging fruit in sodium reduction:
…some of the “easy” food reformulations to reduce the sodium content of processed foods have been achieved by the major food manufacturing companies, and in these cases, efforts to continue lowering the sodium content now require more creative and intense efforts.
You certainly get the sense when reading the report that the authors do not believe that the regulations will be simple, costless, or that our understanding of if and how the regulations will work is anywhere near certain. And remember, these assessments are from advocates for the regulation; it is almost certain that representatives of the food industry would be even less sanguine.
Another point that supporters of the salt regulation have made is that if you slowly reduce your salt intake, you won’t notice the decrease in saltines of foods. The IOM report, however, cautions that this a) this is far from certain, and b) may not apply to all foods:
…while data from perceptual studies may point the way to quantitative levels at which changes in the presence of a substance may not be perceived, much is yet to be learned about the application of such work to the wide range of food products and to other practical considerations in the real world….
Elsewhere they offer even more reasons to worry that sodium cannot be reduced without consumers tasting the difference:
First, the time course of changes in preference for salty foods in response to changes in salt intake is not well understood. Second, there are questions on the extent of a salt reduction that can be accomplished in a single reformulation without greatly altering the palatability of food…Third, it is unknown whether individuals are able to acclimate to lower-sodium foods when some high-sodium foods remain part of their diet.
The last point in the above quotation is worth unpacking a little. The report cites some disagreement among IOM committee members about whether exceptions to salt regulations should be made for certain foods. This discussion highlights that when faced with the possibility that their relatively lighter-touch regulation won’t work, some of the study’s authors would be willing to recommend more draconian measures:
…it is not known whether sensory accommodation would occur if salt were reduced in a single product category such as soup of bread or if the majority of the diet were low in sodium but consumers occasionally consumed foods that might be exempted from sodium reduction (anchovies, olives, etc.). This gap in current knowledge has been a concern for some committee members in determining whether exemptions should be considered for salty foods consumed in small quantities.
In previous quotations, the authors worry that it may be impossible to lower the sodium to a level acceptable in some foods, and here they recognize that the availability of these foods could prevent people from having their salt tolerance lowered. If foods like anchovies or olives aren’t amenable to sodium reduction without significantly altering their palatability, some of the committee members seem okay with those foods being effectively banned. This shows that the failure of a relatively lighter-touch regulation may simply lead policymakers to take a harder line with more draconian regulation. If you’re looking for the next slippery slope, this is a good place to start.
The problems mentioned above do not even get at the possibilities of public choice problem of allowing special exemptions, higher food prices due to higher R&D costs for food producers, increased barriers to entry, increased incentives for industry consolidation, lower levels of future innovation in new food choices, and that regulators will make “mistakes” and set suboptimal levels.
If you want to argue that the benefits of these regulations outweigh the costs, that’s an argument to have. But let’s be realistic about the costs. The old maxim is a useful one: there is no such thing as a free lunch.
Sciences other than economics use game theory, but they are loathe to use rational agent assumptions frequently favored by economists:
The academic tribes, however, will hesitate to accept the gift of game theory from economists if they are told that it comes with the rational actor model. Not everyone wants to shoulder the obligations that model entails….Psychologists, for instance, analyze decisions in terms of (often unconscious) cues and heuristics, and are not likely to switch to the paradigm of Beliefs, Preferences, Constraints and Expected Utilities that underlies the rational actor model. Why should they? In evolutionary game theory, they can enjoy the full panoply of behavioral experiments without the restraints imposed by the loitering presence of the rational actor. Strategies (that is, programs of behavior) need not be the product of rational decisions. They can be copied, for instance, through learning or inheritance.
Of course non-rational behavior can always be explained rationally, if unsatisfactorily, simply by postulating that the individual prefers to act however they are acting:
Gintis embraces another approach, explaining cooperation by a human preference for what he terms character virtues (such as honesty, trustworthiness or fairness). But every behavior can be interpreted as a preference for some virtue.
I think such explanations tend to irritate non-economists interested in explaining behavior.
Ryan Avent thinks we should start thinking about safety nets that are targeted at structural unemployment. Many of the safety nets we have in place are best suited for unemployment bouts resulting from aggregate demand shocks. Workers collecting unemployment and spend it, which makes them better off and supports aggregate GDP, but doesn’t really do anything for structural unemployment. There are retraining programs and such, but there’s not a lot of evidence that these are effective. So what policies could we pass to make the unemployed better off and incentive them in a way that speeds up the structural unemployment adjustment process?
One idea is relocation vouchers. If you offer relocation vouchers to unemployed workers who move a minimum distance from their current residence, then you could incentivize labor to move where it is needed away from where it is no longer needed. The demand for this type of voucher can be seen in the piece from Catherine Rampell on structural unemployemt that Avent was commenting on:
Ms. Norton has sent out hundreds of résumés without luck….She has one prospect for part-time administrative work in Los Angeles — where she once had her own administrative support and secretarial services business, SilverKeys — but she does not have the money to relocate.
“If I had $3,000 in my pocket right now, I would pack up my S.U.V., grab my dog and go straight back,” she says. “That’s my only answer.”
This is someone who could clearly be made better of by a moving voucher. In contrast, unemployment payments for her would do nothing to incentivize or even allow her to move, which would mean she remains in the labor market for which her skills are not needed at a salary she will accept.
One way to pay for this program would be to allow workers to front load their unemployment. Take a full three months worth of unemployment at once instead of spreading it out as long as the person can verify that they are relocating a minimum number of miles away from their current residence.
Senator Tom Harkin has proposed an amendment to the financial reform bill that will set ATM fees at $0.50. Many consumers may celebrate this law, but as they teach in econ 101, if you set a price ceiling, quantity will go down and there will be a shortage. This means banks will put in less ATMs, and people will have to travel farther to get to shorter supply of them. In the end, consumer welfare very well may go down if total travel costs are greater than the gains from lower fees.
The fact that regulated fees lead to a lower supply of ATMs is fairly obvious, and supported international comparisons. The UK and France regulate ATM fees, and have 968 and 761 ATMS per million inhabitants respectively. The US and Canada don’t, and have 1,335 and 1,630. In addition, ATM supply has grown faster since banks started charging surcharges to other bank’s customers to use their ATM 1996. The number of ATMS per capita in the US was growing at an annual rate of 9.2% from 1991 to 1996, and then at 16.7% from 1996 to 2001. This paper provides empirical evidence that consumers in counties with high travel costs experienced welfare gains from these increased ATM fees, while the effect is negative for those in low travel cost counties. This paper, which the above statistics come from, argues that bank profits are lower and consumer surplus higher when banks can charge ATM fees.
To the extent that poorer areas tend to be underbanked, and are most likely to experience higher ATM growth in the future than wealthy areas, the impacts of this law will have the worst impacts on poor people.
…a homeowner who moves has a cross-location hedging opportunity against sale price risk if he buys a new home at another location. Instead, a homeowner who moves to a rental house may only have been intertemporally hedged against price changes. This paper investigates the quality of both hedging opportunities.
That is from a new paper on the sales price risk of homeownership, and the value of homeownership as a hedging strategy against future price rises. The idea is that if you buy a home today and plan on selling it in the future and buying another home, you are facing two sources of risk: the risk in the uncertainty of the future sale price of the house you’re buying today, and the risk in the uncertainty of the future buying price of your next house. However, if you stay in the same neighborhood, or move into a neighborhood with correlated prices, then these risks may cancel out such that you are are hedged against this uncertainty. The authors use home sales data from the Netherlands to confirm the existence of this hedging opportunity.
This does suggest that if you are planning on buying a house in the future you are exposed to house price risk even if you don’t buy a home today. Heding against future price rises is a benefit of buying a home today.
However, the authors offer two words of caution for those who would seek to capitalize on this in today’s market. First, they note that
…the risk position of households may be aggravated, because income shocks and house price shocks are usually positively correlated.
And second, their empirical evidence shows that the ratio of risk to return is higher during periods of economic decline:
Hence, especially during an economic downturn the risk per unit of return is relatively high. In particular, one euro of return in 2000 was associated with between 0.6 and 1.6 euros spread in returns across types of houses. In 2003, this range was between 1.5 and 3.7 per euro return, and it was even higher in 2008 with the coefficient of variation ranging from 2.8 to 4.3. Hence, risk per unit of return may be between two or three times higher during an economic bust than during an economic boom.
Nevertheless, if you are certain you are going to buy a house in the future, than you are exposed to house price risk. Buying a house today can allow you to hedge against that risk.
One popular narrative of the subprime/foreclosure crisis is that many borrowers did not understand the loans they were getting into, and that subprime lenders took advantage of this by offering loans that were doomed to fail and difficult to understand. A new study from the Atlanta Fed provides evidence in support of the borrower ignorance part of the narrative:
We ﬁnd a large and statistically signiﬁcant negative correlation between ﬁnancial literacy and measures of mortgage delinquency and default, and the ﬁnding is robust to the inclusion of controls for income, education, risk aversion, and time preferences, thus ruling out a broad set of potential biases from omitted variables. The point estimates are remarkably robust, and quantitatively important: 20 percent of the borrowers in the bottom quartile of our ﬁnancial literacy index have experienced foreclosure, compared to only 5 percent of those in the top quartile. Furthermore, borrowers in the bottom quartile of the index are behind on their mortgage payments 25 percent of the time, while those in the top quartile are behind approximately 10 percent of the time.
Concerns of a lack of basic financial literacy have led to calls for… well, more financial literacy. But this study suggests the problems go deeper than the inability to understand how to discount, or what an exploding balloon payment is, to a fundamental lack of numerical ability:
We include as control variables measures of other aspects of ﬁnancial literacy and a general measure of cognitive ability, but ﬁnd that the correlation is highly speciﬁc to one aspect of ﬁnancial literacy: numerical ability.
Provocative question of the day: should mortgage applications come with a short IQ test, where potential borrowers receiving a score below a certain level are required to undergo extensive counseling to make sure they fully and completely understand the mortgage, payment schedule, and the all the issues it is assumed a borrower should understand?
Ezra Klein has some, to my mind, contentious ideas about how to fix Wall Street. He’s worried about the size, and power of big banks, both of which feed and are in turn fed by their profitability. Unless you can make them less profitable, he argues, they will be able to influence regulators and legislators; Wall Street is simply too profitable to exist in our corruptible democracy. He writes:
…I don’t believe you can effectively regulate the financial industry so long as it’s sucking up about a third of domestic profits. The incentives to take massive risks will just be too great. The power to bribe Washington to dismantle regulations and legislation will be irresistible over time.
His solution is that we need to take advantage of the brief moment of public interest in financial regulation to tax Wall Street profits down to a size at which they can’t buy as much regulatory influence. So if I understand him correctly, we can’t trust future regulators and legislators because
There’s money, expertise and interest on one side of the ledger, and the other side is likely to be spending its time on other things. How long till one party or the other needs to fund a tough reelection campaign and cuts a quiet deal with the financial sector?
And trying to hold Wall Street’s massive profits down via legislation and regulation will increase or decrease the amount of money, expertise, and interest that they are willing to commit to influencing legislators? Say you do manage to hold Wall Street’s profits down by $50 billion a year with $50 billion a year in taxes. You’ve just given Wall Street a $50 billion incentive to influence financial laws and regulations. They can now spend $5 billion a year trying to get that tax repealed and make a profit on that effort if it takes them less than 10 years.
My guess is that giving Wall Street maximum incentive to focus on changing whatever FinReg laws are passed this year isn’t going to make it easier for future legislators to leave those laws in place; it’s going to make it impossibly hard. I know it would feel great to pass a law that Wall Street hated by trying to maximize the damage to their profits, but laws designed to cost Wall Street money are laws designed to face a rough future.
In addition, an even worse possibility than future banks simply getting the massive profit tax laws repealed, is the threat that they manage to use the massive profit tax laws to their advantage; carving out loopholes for themselves, using the tax to erect barriers to entry, turning it into a protectionist policy that taxes foreign banks more than domestic banks… these things could all help Wall Street become less competitive, more centralized, and thus more dangerous.
This whole idea strikes me as unprecedented, highly speculative, politically costly, and more likely to make things worse than better for future regulators, legislators, consumers, and taxpayers. So no, you can’t tax Wall Street into submission.
A new NBER paper argues yes:
…we investigate whether an individual’s height is associated with criminality. Recent economic studies have uncovered positive associations between height and labor market outcomes, which suggest that taller individuals experience labor market advantages and, therefore, have fewer incentives to turn to crime (Persico et al 2005; Case and Paxson 2008)… This paper finds, consistent with several labor market studies, an inverse relationship between crime and adult stature, even after controlling for several features likely to influence an individual’s decision to select into criminal activity.
Specifically, we find that the hazard of prison entry for individuals in the fifth quintile of stature was 20 to 30 percent lower than for individuals in the first quintile…. Nineteenth-century criminals were short.
Policy implications anyone? Perhaps Mankiw’s taxation of height will gain steam now. I think a subsidy for shortness would get more traction.
I came across this old post from Katja Grace where she provides an argument for smoking bans based on game theory. The basic story is that people like conformity, and there is a coordination problem that the government can fix:
Imagine everyone is doing A. Everyone likes doing B more than doing A, but not as much as they like conformity. There would be a huge gain to a coordinated shift to B, but nobody moves there alone. In some such situations those involved arrange coordination, but often it is impossible. If there are many equilibria like this, and no means to move to better ones, intervention by someone with the power to force a coordinated move could be a great thing.
She provides an example where everyone goes to the same crappy bar because that’s where everyone goes. If any individual person, in her example Roger, decides to go to a better bar, they are worse off since they prefer being around people. But if everyone went to a better bar, they would all be better off than the status quo. Here is the normal form game she uses to model the situation:
Payoffs for Roger in choosing a nightclub
In this game, Southpac is the crappy bar. If Roger, or any individual, goes elsewhere when anyone else is still at Soutpac, their payoff goes from 2 to a 1, so he is worse off. If everyone goes somewhere else though, then the payoff to Roger and everyone else is 3, so they are all better off.
She uses this game to argue that the government could sometimes make everyone better off, for instance in the event of a smoking ban. But I think this doesn’t work as a justification for smoking bans or any bans that attempt to fix coordination problems like this.
The failure of this model is obvious if you consider that once everyone is in the higher payoff equilibrium, there is no incentive to diverge, and the status quo becomes the dominant strategy for everyone. In her example, this means that once everyone goes elsewhere, they never go back to Southpac.
This means that a temporary smoking would suffice to fix the coordination problem. Anything more than that is unnecessary, and it runs the risk that this model is incorrect and you are holding people in a lower payout equilibrium. Thus this model can never justify permanent bans like we see.
This model also offers an easy empirical test which I think intuition suggests it would fail. If you enacted a temporary smoking ban and then removed it, would bars and restaurants revert to smoking or would they stay nonsmoking? If Katja is correct, they should stay nonsmoking. I would bet by and large though this would not be the case.
With education reform I tend to be a fan of what Chester Finn, the editor of Education Next, calls “blowing up the system”. That said, I think Gail Collins criticism of the reforms that Florida tried to pass are worth considering:
Can I digress, people, and say that while it’s important to make teachers accountable, telling them their jobs could hinge on their students’ grades on one test is a terrible idea? The women and men who go into teaching tend, as a group, to be both extremely dedicated and extremely risk-averse. The stability of their profession is a very important part of its draw. You do not want to make this an anything-can-happen occupation, unless you are prepared to compensate them like hedge fund traders.
It may be that we do want to pay teachers like hedge fund traders, but there are 3.5 million teachers in the country right now, and whatever reforms we put into place must consider not only the stock of teachers you want, but the stock you currently have. Given that the job selects for risk aversion, injecting too much employment risk into the job, especially too fast, is probably going to do more harm than good.
Matt Yglesias agrees with William Galston that the democrats would do well to switch their focus from immigration to the economy. I don’t know -or care- what’s best for the democrats politically, but I think immigration policy is actually one of the few levers that the government can push on to improve the economy at this point.
Most economists would agree that an increase in house prices right now would be a good thing. It would increase household wealth and bank balance sheets, which would potentially stimulate consumption and lending. A large enough increase can send a clear signal to hesitant homebuyers that the bottom has been reached, and lead to further increases in demand. Construction output increases to meet demand… and so on and so forth. These things are obvious though. The question is, how do we increase prices at a low enough cost to make it worthwile?
The most efficient way I can see to do this is to increase immigration. I think increasing all immigration would work, but even more targeted to economic growth would be a very large increase in H1B visas for skilled workers. This is a more efficient means to increase house prices than any other stimulus/housing policy plan out there, because it
1. requires no market distortions (you’re welcome Arnold Kling)
2. does not encourage buying over renting (Ryan Avent and Felix Salmon, that’s for you)
3. is Keynesian stimulus, because the spending on housing, transportation costs, and new furniture associated with immigrating would not have happened otherwise (DeLong, you can celebrate)
Now if you auction these H1Bs at a couple thousand dollars a pop and then use that money to build those crybabies in Arizona their border fence then you’ve got a Pareto improvement. What about lower wages for high skilled workers? I can believe that the increase in output from more immigration might be enough to actually increase the marginal productive of skilled labor, and thus wages. If someone wants to work out the general equilibrium result there please do so.
If you’re concerned about the welfare of the emigrating nations consider that remittances are often sent back, and the ability to leave the country and get higher returns on human capital raises the incentives for citizens of that country to accumulate human capital. This in turn would increase the supply of human capital generating institutions like higher ed, which the non-emigrating population benefits from. I can’t cite specific literature on this, but I believe the work done on this generally supports the notion of a positive impact of emigration.
Perhaps then we should be considering pro-emigration policies too. For instance, as Dean Baker has suggested, if we let the elderly take their medicare to Europe they will get more for their money and we will spend less on medicare.
In a global sense more immigration is just allowing labor to move to where it is most productive. For all the urbanists who criticize pro-homeownership policy for reducing labor mobility, this should be an obvious win. Let’s lower barriers to entry and exit, and not let nativism get in the way of economic recovery.
A new gallup poll illustrates the relationship between smoking, education and income.
The relationship mirrors health issues
The real challenge, however, is to tease out the causal relationship between these four factors. Are people poor because they are unhealthy? Are they uneducated because of the same social forces that drove them to smoke? Are there genetic factors underlying all of this?
A new paper from NBER tests Max Weber’s Protestant work ethic, and other hypothesis about whether religious affiliation affects economic behavior:
We find that Protestantism increases contributions to public goods. Catholicism decreases contributions to public goods, decreases expectations of others’ contributions to public goods, and decreases risk aversion. Judaism increases worker reciprocity in a bilateral labor market gift-exchange game. We find no evidence of religious identity effects on disutility of work effort, discount rates, or generosity in a dictator game.
The authors sought to create exogenous variation in religious identity using a laboratory experiment:
The priming instrument… is a sentence-unscrambling task where subjects are asked to drop the irrelevant word in a five-word group and rearrange the remainder to form a four-word sentence. For example, “yesterday it finished track he” becomes “he finished it yesterday.” Each subject unscrambles ten sentences.
The sentences vary depending on whether the subject is in the religion-salient condition or the control condition. Five of the sentences unscrambled by religion-salient subjects contain religious content. These five sentences are: “she felt the spirit,” “the dessert was divine,” “give thanks to God,” “the book was sacred,” and “prophets reveal the future.” None of the control subjects’ sentences contain religious content.
How valid is this instrument? One worrying result is that the authors fail to replicate an experiment done by the originator of this particular priming instrument. So Caveat lector, but still, interesting results and approach.
Kevin Drum and Matt Steinglass both object to the Spirit Airlines decision to charge for carry-ons on the grounds the sometimes you want the luxury of a simplified decision that bundling provides. Drum complains:
Choice is good. Most of the time we want it, and economically it’s often beneficial. But it can also hide things and make prices hard to compare. Is the Spirit flight really cheaper? Better do a close comparison! Is dinner at Joe’s the same price as dinner at Mary’s? If Joe charges for bread, maybe not.
I generally hate a la carte pricing for things that used to come bundled. I especially hate it when it forces me to make complex decisions about things that are trivial and irritating. I hate spending discretion. I want to be offered a price for a service with the normal accoutrements of that service, not ten prices for ten different versions of that service in terms that take me an hour to understand.
They both have a point that simplicity can be the consumers friend and complexity can be a device used by producers to extract more surplus from consumers… or to put it more plainly, to charge higher prices. As Drum points out, it is valuable to a consumer to be able to compare prices between suppliers. If you can’t compare the total price and benefits of two otherwise similar goods (or bundle of goods), then the goods become less like homogeneous, easily substitutable goods, subject to a competitive market price and more like heterogeneous goods that can’t easily be substituted for each other. This means that making goods complex or confusing can allow them to garner a monopolistically competitive price.
In the real world, there are many examples of this. Mattresses for instance, are notoriously and purposefully overcomplicated. The same exact mattress is sold under different names in different stores just so that you can’t do price comparisons. Even the free market economist must bristle somewhat at this… or at least they will when they have to buy a new mattress.
However, Steinglass should be more wary of bundling, which he praises, since it can easily be used as a tool for confusion and price obfuscation. This is exactly what Toys’R’Us did when faced with increasing competition from warehouse clubs like Costco, Sam’s Club, and BJ’s. They forced manufacturers to only sell bundled toys to the warehouse clubs, thus making their prices incomparable so that they couldn’t be undercut by their lower prices. This example also illustrates why all producers don’t just unilaterally make their prices more confusing: people like simplicity, and are willing to pay for it. If Toys’R’Us has just put all of their own products into complex bundles, then consumers would switch to the warehouse stores because they prefer simplicity. Thus, Toys’R’Us pressured manufacturers into foisting complexity onto their competitors.
For a more formal discussion of this phenomenon, check out this paper from Xavier Gabaix and David Laibson which shows that purposefully confusing complexity can exist in a competitive market:
In this analysis, complexity is itself an endogenous variable chosen by each ﬁrm. For example, a ﬁrm can create an unnecessarily complex fee schedule, which makes it harder for a consumer to determine the true cost of the good. We show that ﬁrms will generally prefer such excess complexity. A small amount of excess complexity has only a second-order negative impact on the intrinsic quality of the good, but generates a ﬁrst order increase in the (confusion-driven) demand for the good. So ﬁrms choose ineﬃciently high levels of complexity.
In the interesting debate on libertarian paternalism at Cato Unbound both sides have mentioned the issue of smoking bans. In the anti-paternalism lead essay, Glen Whitman uses smoking bans as an example of a slippery regulatory slope: first it was just the airplanes, then it was restaurants, what’s next? Richard Thaler counters in the pro-paternalism essay that smoking bans aren’t even really nudges:
First, most of the anti-smoking laws are based on externalities, not paternalism. People do not want to fly, eat, or work in smoke-filled environments. Indeed, many smokers favor such laws. Note that while smoking bans are not nudges, they are shoves…
Scott Sumner chimes in, disagreeing with Thaler that smoking is an externality at all:
Doesn’t this argument violate the Coase Theorem? For example, let’s take the ban on smoking in the workplace. Where is the externality argument? Doesn’t the employer already have an incentive to put in place the smoking rules that minimize his productivity-adjusted wage bill?
One important problem not discussed is that if we are nudging (or shoving) people away from a behavior that is (or could be) an externality, what are we nudging them towards? The unintended consequences are potentially serious here. For instance, the literature on smoking bans suggests that they increase drunk driving and exposure to secondhand smoke for the children of smokers, both of which are more obvious and more egregious externalities than the one we were trying to get rid of in the first place.
Negative unintended consequences may occur even when there is no externality. For instance, a new paper in the current edition of AEJ: Applied Economics uses a field experiment at a fast food chain to show that if not properly designed, a nudge that successfully reduces an individual’s caloric intake of sandwiches may be offset by an increase in their caloric intake of soda and side orders. Depending on the kind of sandwich, side order, and soda involved, this may make the individual worse off. Shifting from less of a turkey wrap to more french fries with trans fats would clearly make the person worse off.
Both examples show that when weighing nudges (or shoves) we should consider the potential unintended consequences, otherwise we may end up pushing people in a bad direction. When externalities are potentially involved, this is especially important. The smoking bans also provide a warning that even if we identify a very serious unintended consequence down the road, the legislation won’t necessarily be undone. We should be extremely cautious about passing these laws if we are uncertain about our ability to identify unintended consequences in advance; by the time we know for sure it could be too late.
Glen Whitman over at Cato Unbound attacks mild paternalism using behavior economics, arguing essentially that they put us on a slippery slope to more intrusive paternalism:
First, it is well-established that people exhibit extremeness aversion: a tendency to avoid positions that are presented as extremes…. The mere presence of an extreme option makes the middle option seem better. The new paternalists, intentionally or not, have exploited this same tendency by presenting their position as a middle-ground between laissez-faire and heavy-handed paternalism.
This would be no great concern, were it not for the tendency of the middle ground to shift over time. A newly adopted middle-ground quickly becomes the status quo. Then a more intrusive option takes center stage, and what used to be the middle-ground becomes one of the bookends….
Sound paranoid? Anti-smoking regulations followed a similar path. Once upon a time, banning smoking on airplanes seemed like the reasonable middle ground. Now that’s the (relatively) laissez-faire position, smoking bans in bars and restaurants are the middle, and full-blown smoking bans have come to pass in some cities.
Behavioral economics is typically used to justify paternalism (“people are irrational, let’s help them”), and the usual counter-argument is a rational agent one (“no they’re not, so you can’t”). It is, to me, a novel framing of the debate to see Whitman attack paternalism on behavioral grounds. Will Gary Becker get a voice in this debate?
Two days ago my past self discussed a paper that argued we might want to consider our current selves as distinct from our future selves. The idea was that “when we make choices that affect both our current selves and our future selves, we think about the preferences of our future self in the same way we think about the preferences of a different person”.
My past self, being more interested in entertaining than informing, didn’t really discuss some critical problems that could arise if we took this framework seriously. My current self is less frivolous than that.
First, is the issue of whether suicide is murder. teageegeepea points out in the comments that if you treat future and present selves as separate people, than things that we obviously consider acceptable also become crimes:
If suicide is murder, then spending in the present is theft from a future self, sex is rape and a boxing match is battery.
The crucial difference here is that on average the preferences for these things should align fairly well between current and future selves. A priori we don’t really have any way to know whether the current boxer’s future self would want him to not box, or not spend the money, or not have sex. His future self may in fact wish his past self did more of each.This is true even for behaviors more destructive than those; if we enjoy those experiences today, our future selves may enjoy the memories of those experiences enough that they are worth doing.
The reality is that in most situations the relative preferences of future selves and current selves are difficult to guess, and the current self probably knows better than policy-makers or anyone else what his future self wants. If we don’t now know the preferences of future selves over current actions, than we can’t call those actions crimes against them.
With suicide, in contrast, it is a fairly safe bet that the vast majority our future selves would really, really, rather our current selves not commit suicide. In this case the distinction between future and current self is useful, because we know today how to balance current actions to better reflect future desires: we simply try to stop suicides. You can object that suicide shouldn’t really be called murder, and that’s fine; since our future selves and current selves would obviously be different “selves” in a way that is distinct from how you and I are different selves, and I’m not concerned with whether or not the word can literally be used that way. What’s more important than the semantics is the notion that suicide is like murder in this framework.
A second objection comes from Sister Y:
The “successive selves” idea can never genuinely catch on, true as it may be, because then we couldn’t lock people up for rapes and murders for long periods of time.
We can still justify locking people up even if we grant their future selves as separate from past selves. First off, for practical reasons alone we must all be somewhat responsible for the actions of our past selves. The alternative is a world of intolerable level of crime. Even if we granted that, in-and-of-itself locking a future self up for the crimes of a past self is not ideally just, the downside to not doing this is obviously too great. I’ll certainly grant though that the notion of a future/current self distinction certainly complicates the notion of justice.
The second reason we could lock future selves up is that, as I said before, in the vast majority of cases the current and future selves preferences tend to be highly related. In the short run, most people will not be able to credibly claim that their current selves aren’t criminals even though their past selves were. However, we already accept this notion in the long run and release people from jail if we believe they are truly “reformed”, which really means that their current self is sufficiently different from their past self.
These issues do show the difficultly in considering future and current selves as literally distinct selves. However, the points I’ve raised show that we already accept many of the logical conclusions of doing so. What I find appealing about this notion isn’t that it justifies some new, radical, ideas about how we should think or policies we should enact, but rather how it justifies common sense policies without violating notions of rational individuals. We should try to prevent suicides and some very destructive behaviors; with some choices we tend on average to not consider the future as much as we should; in the long-run we can become very different people. These are not radical ideas.
Then again, the extent to which this idea makes us freer from responsibility for past actions isn’t really appealing to me. And if society in general accepts many of the conclusions of thinking like this without actually thinking like this, then would thinking like this move the general attitudes about crime and free will too far in the direction of paternalism and no personal responsibility?
Overall I am open to persuasion on this issue.
A common framework for economic analysis is to assume that people attempt to maximize lifetime utility subject to a discount rate. But should economic analysis be concerned that your future self has different preferences than your current self? Are you really two different selves whose desires should then be weighed separately? A new paper from the Boston Fed argues that neuroeconomic evidence suggests that the answer may be “yes”, because when we make choices that affect both our current selves and our future selves, we think about the preferences of our future self in the same way we think about the preferences of a different person:
…our central idea is that essentially the same brain system is utilized when individuals attempt to predict or empathize with others and when they attempt to predict or empathize with themselves in the future.
So if the way we think about our future selves is as distant as we think about strangers, then is our current self really adequately considering the wants of our future selves when they make choices that affect our selves in the future? This naturally raises the question of whether and to what extent the state should protect our future selves from the actions of our current selves.
If one seriously considers the future self as a separate self, it seems to me a serious challenge the Szaszian idea that mental illness is just extreme preferences and that suicide should be respected and allowed as a legitimate exercise of choice; if our future selves are separate selves, then suicide is murder.
One way to incorporate this into policy making would be to consider not just the individuals who be affected today by policies, but also how their future selves will ex ante view the policy. For instance, we would consider how people who were “nudged” into quitting smoking by high cigarette taxes feel about the desirability of those taxes both before and after they quit.
Behavioralists who are pleased to see a new justification for their favorite paternalism-lite policies should not celebrate so quickly. Behavioral economic studies that demonstrated seemingly irrational choices may no longer be inconsistent with rational agents when you consider present and future selves separate agents.
People who think that economists are intellectual turf-grabbers probably aren’t going to like this one.
Robin Hanson recently said that we should count satisfying the preferences of the dead as a benefit in cost benefit analysis. I disagreed, arguing 1) that it would lead to horrible outcomes, and 2) the dead don’t know whether their preferences have been satisfied, so satisfying them benefits nobody.
In response, Robin has offered some comments. First he states that efficiency is not morality, so we should not be surprised that using efficiency criteria would deliver us outcomes that are not moral. That’s fine, but the conclusion I draw then is when efficiency would have us select obviously highly immoral outcomes -like, say, slavery, holocausts, and orphan eating- we reject efficiency in that circumstance. Robin, in contrast, bites the bullet: bring on the orphan eating!
Robin’s second objection is that we should definitely care about outcomes that we can’t see, like having our preferences satisfied after we’re dead:
Many of us want things we will never experience directly; we want our children to prosper after we are gone, for example. This is especially true of our moral wants; we want our donations to Africa to actually help real Africans. So we are understandably wary of deal-making frameworks which explicitly suggest that they seek only to achieve the appearance, not the substance, of our wants. So yes, a deal-finding analysis tool should definitely count unseen wants!
What he’s saying is that people prefer to believe that their preferences are going to be satisfied, even the ones they won’t see themselves. This is an argument for satisfying unseen preferences but only to the extent that it allows us to believe our own unseen preferences are satisfied.
This is really a classic time-inconsistency problem: we’re all better off if we live in a world where we can trust our unseen preferences are satisfied, including preferences for things that happen once we are dead. But the most efficient outcome is to commit to satisfying unseen preferences, and then later reneging on them. Of course once you break that promise once, it will be hard to credibly commit to keeping them in the future. Thus breaking the promise now will mean that current and future generations will not believe their preferences will be satisfied, which may be an inefficient outcome relative to not breaking the promise in the first place.
Of course we never promised our ancestors anything. They observed what we are observing today, which is the fact that “Our contract law system refuses to enforce many win-win deals between distant generations.” So by not caring about the preferences of our ancestors, we are not reneging on any promise or refusing to fulfill commitments they expected to be fulfilled, which means we could still credibly commit to honoring the preferences of the future dead. So while future people should defer to us, we should definitely not defer to our ancestors, because to do so would be inefficient.
And what if our future ancestors renege? It doesn’t matter as long as we die believing that they wouldn’t, and they can find some way to convince themselves they’ve credibly committed to their future generations. It’s all a con game where our own irrationality can make us better off. How perversely un-Hansonian of a conclusion is that to derive from a Robin Hanson argument?
The proposed Consumer Financial Protection Agency was supposed to regulate payday loans -those extremely high interest rate short term loans- but it now appears that it may not. This could turn out to be a good thing for consumers, since the best way to help payday borrowers might be to deregulate the industry…. Bare with me, I know that sounds ridiculous.
A recent paper from Robert DeYoung and Ronnie Phillips at the Kansas City Fed provides some cautionary results about potential negative side effects of increasing regulation, and suggests possible positive impacts of deregulation. What they find is that more competition among payday lenders can decrease the exorbitant interest rates on payday loans. Increasing competition decreases prices; this is not so surprising.
The authors even go so far as to suggest increasing competition by removing regulations that limit the ability of local banks, thrifts, and credit unions to offer payday loans. This makes sense, since reputation is probably more important to local banks, thrifts, and credit unions than it is to payday lenders, they would be more likely to offer actuarily fairly priced payday loans and less likely to try and manipulate borrowers with confusing contracts, etc. In fact, DeYoung and Phillips provide evidence that payday lenders with large franchises are less likely than mom-and-pop stores to engage in exploitative pricing behaviors. Getting banks and other financial institutions into payday lending could help prevent a “race to the bottom” in lending standards that might otherwise result from increasing competition.
Another caution provided by DeYoung and Phillips is that setting a rate cap may provide a Schelling point for payday lenders to collude around, so that for many borrowers rates could actually go up. What this means is that the payday lending industry has the market concentration to collude, but without explicitly communicating with each other (“Hey, if you charge 1,000% and I’ll charge 1,000% too) they cannot find a stable equilibrium price point to settle on, thus the resulting market is somewhat competitive. A rate cap provides them with a natural price point to collude around. So if the government says “You can’t charge more than X%”, then that rate becomes a Schelling point and all lenders begin collusively charging X% for all loans, which can actually be a higher price point than they were previously charging for many loans. The evidence they provide is not conclusive, but it is consistent with studies that have shown similar responses resulting from credit card rate ceilings.
I would not say this paper is conclusive enough to declare that these impacts are what will occur if we get national rate capes for payday loans. But as we debate whether payday lending should get regulated as a part of the CFPA, these possible outcomes and deregulatory means to improving the industry are worth thinking about.
It seems to me that the authors of this study have identified but not really controlled for the endogeneity problem that gun control regulations are both caused by and the cause of the gun prevalence. They report:
Access to lethal weapons is an important risk factor for suicide. Our study suggests that general barriers to firearm access created through state regulation can have a significant deterrent effect on male suicide rates in the United States. Permit requirements and bans on sales to minors were the most effective of the regulations analyzed.
My instinct is that decreasing access to guns will just lead to more non-gun suicides, which is what one study they cite actually found:
An evaluation of the 1996 National Firearms Agreement (NFA) in Australia documents a decline in firearm suicides after the implementation of the agreement (Klieve et al., 2009). However, these findings may be confounded with an overall decline in gun ownership that preceded the NFA. Additionally, there was some evidence of increased suicides by hanging.
However, I can imagine some people might deterred from a temporary impulse to kill themselves, and that the extra time that buys them might be enough to get help, change their mind, etc. Nevertheless, it is an interesting question, and I think some clever scientist out there should be able to think of an actual instrumental variable to get a better answer to this question.