Voter and Consumer Irrationality: An Extension

Questioning rational choice is not a sufficient basis for arguing for government intervention. But my past few posts questioning standard market propositions in the case of health care are not meant to promote every single government regulation. Rather, I wonder if we can try to further the study of economics and public policy by reversing the typical way public choice economics is applied.

When I think of public choice economics, I think of papers about how the marginal costs and benefits of lobbying lead to poor government policies or how decentralizing the functions of governments leads to a competitive equilibrium with better governmental policies. At its core, I interpret this line of analysis as using the success of markets to explain why governments fail. Markets succeed because marginal costs line up with marginal benefits, and thus governments fail because these costs and benefits are distorted. Markets succeed because there is competition between firms, and thus governments fail because there is no competition. What I want to propose is that we change the direction of this line of thought, and start from the vast literature on why governments fail, and see if that can teach us about why certain markets fail. 

This is why I think behavioral economics is so important. By reducing biases down to core psychological first principles, it highlights the connection between government and market failure. But we should also make sure to avoid thinking that policy makers themselves are perfectly rational. As John Papula pithily points out:

Why in the world do behavioral economists who study our flaws and irrational quirks advocate centralized power in the hands of a small group of flawed overlords? If people are irrational, so are government regulators, only they have corrupting monopoly power.

As such, I agree that it is important that behavioral economists grasp the concept of public choice, but also for public choice to grasp the concept of behavioral economics. There exist reasons for why the market often overcomes behavioral biases, but then there also exist markets in which those mechanisms don't function well. And most likely, we can trace these mechanism failures to arguments past public choice economists have made about the failure of government. That is my conjecture, and I hope to be able to pursue further analysis of health care under this framework.

Voter and Consumer Irrationality: Two Sides of the Same Coin

One strong argument against the government provision of services is that democracy is a very imperfect substitute for markets. As noted by Bryan Caplan, voters have a hard time rationally evaluating policy options, and thus the public choice economy may be a very bad one. The natural conclusion of this argument is that, to the greatest extent possible, the provision of goods should be pulled away from government and the responsibility should instead be allocated to the market. Yet this line of logic contains its own contradiction: if voters cannot rationally choose the right public policy, how can they be expected to make the right private choice?

Now, I don't mean to say markets never work. I am very happy with the way capitalism has treated me, both in terms of the large historic liberalizations that brought me to this country as well as in the small conveniences that improve my everyday life. I don't mean to reject this. But what I want to suggest is that the acceptance of the irrational voter hypothesis suggests that we should take a closer look at our acceptance of the perfect market equilibrium in many different sectors, in particular health care.

Bryan Caplan, in his work on voter irrationality, outlines four main biases: the anti-market bias, anti-foreign bias, make-work bias, and pessimistic bias. 

Anti-market bias refers to the public's systematic bias towards policies that interfere with prices and profits, such as farm subsidies or rent control. I like to think of it as the public's bias against simple supply and demand explanations of the market. 

Anti-foreign bias refers to the way the public tends to see people from other countries as fundamentally different from itself, and thus the bias causes people to under-estimate the benefit of free trade and immigration. People focus on the auto jobs outsourced by free trade and the native fast food restaurant replaced by immigrants, while paying little attention to the new opportunities and technologies granted by interaction with foreigners.

Make-work bias refers to the way the public confuses productivity with having a job. As Caplan cleverly states, "For an individual to prosper, he only needs to have a job. But society can prosper only if individuals  do a job, if they create goods and services that someone else wants." People tend to make the classic luddite fallacy, that new technology destroys more jobs than it creates, and that, as a result, technological growth actually worsens the standard of living.

The fourth and final bias, pessimistic bias, refers to the way the public tends to overemphasize the things the get worse over time, and forget the ways that life improves. Recession is confused for regression, and the massive technological advances of the markets, such as improvements in information technology and energy infrastructure, are forgotten.

Yet when I think about these biases, I would argue that they all are manifestations of a more fundamental psychological bias: a bias towards salience.

Salience refers to the way certain effects or phenomenon are more apparent and more obvious. We should be familiar with this idea of salience in our everyday lives. It's easy for me to enjoy the concentrated fun of watching old episodes of scrubs, it takes more effort to remind myself of the dispersed benefits of working hard on math homework. It's easy for me to catch up on the extra hour of sleep, it takes more effort to remind myself of the long-term benefits of exercising in the morning. As Katherine Baicker, Sendhil Mullainathan, and Joshua Schwartzstein like to joke, "Our research has definitively determined that running is unpleasant and donuts are tasty," and as such, people tend to choose the salient joy of tasty donuts and the immediate avoidance of hard running instead of the long term benefits of consistent exercise. Or as Thaler and Sunstein put it in their book Nudge, very rarely do people ever make new years resolutions to smoke more cigarettes or drink more alcohol.

I would argue that salience forms the basis of Caplan's irrational voter biases. Salience means that people tend to see the direct negative impacts of market liberalization, and neglect the role of the fallacy of composition in hiding the harms that arise from government intervention. In the case of the anti-market bias, the benefit from paying farmers is salient, whereas the cost of higher food prices are dispersed and not as apparent. In the case of the anti-foreign bias, the closed steel mills down the street are immediately visible, whereas the newly created jobs in the software industry and management consulting are not as visible. In the case of the make-work bias, people directly observe the seamstresses fired and don't see the new women, who on count of greater general prosperity, are then given the chance to go to college for a better life. And finally, in the context of the pessimistic bias, people tend to worry about the bad changes more than the good. People complain more about the rising price of gas and food, which they buy everyday, and not the fall in computer prices, which they rarely have the chance to purchase.

Re-framing the issue in terms of fundamental psychological first principles adds to the way we should interpret the irrational voter hypothesis. Rather than seeing the phenomenon of the irrational voter as an isolated problem that makes government policy ineffectual, we should see the irrational voter as a more general irrational person. As such, there may be certain (not all) markets, such as health care, that do not liberalize in the way other markets do. In the case of health care, because certain differences between doctors, such as bedside manner, are more salient than others, such as improved recovery times, a simple market liberalization may not always promote the best health outcomes. We need to think again, again, about why we need reform. Promoting price disclosure makes the costs consumers pay more salient. Moving away from an employer provided health care system towards individual insurance with an individual mandate makes the costs of choosing bad insurance more salient. But just kicking back and "letting the market (not) work", and treating healthcare just like "consumer electronics, telecommunications, computers" or cars is not a sufficient answer. And if we can move away from fiery rhetoric about socialism and capitalism and towards a grounded and pragmatic analysis of psychology and behavioral economics, only then can a meaningful dialogue on healthcare can occur.

........................
Update 11/18: I extended some of my thoughts on public choice and health care here.

Healthcare and Cars are Not Isomorphic

I've recently been doing some thinking about health care, and my thoughts aren't yet complete. However, when reading many of the conservative commentaries on health care, I was very peeved by their treatment of the market for health care just like the market for any other good or service, and this was aggravated by the fact that I agreed with most of their other arguments. As a result, I wrote the following NextGen article to collect my thoughts on this issue before I move on to more substantive analyses of some recent posts.

..................................................
"Healthcare and Cars: Why They're Not the Same"


How is the market for health care different from that for cars? And how does this difference change our understanding of health care policy?

Let us start from the basics: health care is a service, while a car is a good. While this does not prevent market competition from improving society's welfare on the margin, it does mean that there are fundamental limits to how far the market can improve. In particular, it means that one productive hospital cannot simply “export” its health care services across the country. And as we know from the car industry and international trade theory, that lack of export markets and distant competitors can impede the creative destruction that makes markets work.

But this is an incomplete justification. Haircuts are also services, but few people think barbershops require regulation. Yet, they differ in one key respect. While it is easy for me  to determine the quality of my haircut, it is much more difficult for me to determine what counts as “good” health care. A large part of this is linked to the uncertainty inherent in any kind of biological process. How does one determine if one is receiving sufficient care? Of course, if there is any gross negligence, it can be detected. Yet if the care is just slightly worse than it should be, there's no real way for the consumer to know. The problem is compounded by the fact that most people rarely get sick. If you go to get a haircut every month, you can quickly determine which barbershop is the best. Unfortunately for neoclassical economists but fortunately for society as a whole, people need catastrophic care substantially less often than they need haircuts, thereby impeding the market from finding the most efficient solution.

Chronic care does change the calculation, but it introduces other factors that again prevent the market for health care from functioning like the market for cars. To take a concrete example, consider the 40 million elderly individuals who suffer from arthritis. Does it make sense to ask them to “shop around” to find the perfect hospital? Does it make sense to ask them to trust in an online service like Yelp to decide where to get care? And when they are fatigued by chronic pain, does it make sense to think they can make the “optimal” choice, and not just the most convenient?

Other problems make it difficult for the market for health care insurance to function like a regular competitive market. Again, a parallel to the market for cars, in particular used cars, is helpful. A problem with used car markets is that it's not entirely clear whether the car you wish to purchase is good or bad. This depresses the market value of used cars, both good and bad. But then producers of the more expensive good cars exit the market, further lowering the average quality of a used car until the market spirals out of existence. The market for health care is subject to similar pressures, but in the opposite direction. For insurers, they cannot easily tell whether an applicant is healthy or not. As a result, they charge the same premium to all individuals, which leads to healthy individuals dropping out of the market because they value the insurance less. This process repeats itself until premiums spiral upwards and no market for health care insurance exists.

This describes what economists call adverse selection, and although it is a market failure, it doesn't mean the government always needs to step in. However, understanding how the market circumvents the problem helps us understand why a simple solution for the health insurance does not exist. For example, markets for used cars can exist because the salesmen can offer warranties, so that if the car does break down you can “turn back the clock” on the purchase and return the car. This reduces the incentive for used car salesmen to trick you into buying lemons, because if the car does turn out to be a lemon you can always return it.

But what would a similar system mean for health care insurance? In that market, it's the consumer who has the private knowledge, therefore it falls to the consumer to make the “warranty”. This would mean the consumer would have to be able to guarantee that he or she won't become more unhealthy in the future – an impossible task. Alternatively, insurance companies can “turn back the clock” and refuse to pay for the consumer on the ground that the consumer was unhealthier than was expected. This hardly seems like an efficient or moral outcome, and is precisely the reason why the protections for preexisting conditions is so popular in the ACA.

With these stylized facts, it should be abundantly clear that the markets for health care and insurance are very different from what the standard competitive model predicts. Hence, we need to be very cautious when we draw glib analogies from health care to cars. However, the implications for policy are mixed. On one hand, certain regulatory reforms such as cutting the employer health care deduction, reducing occupational licensing barriers, and relaxing privacy laws can result in substantial advances towards improving health care. Yet the presence of these government failures does not mean that there is no role for the government to ameliorate the market pathologies that I have listed above.

A notable example of this is the individual mandate. The individual mandate, by forcing people to purchase insurance, massively expands the market for health care. In doing so, this increases the incentive for firms to pursue innovative new projects to increase productivity. And even though consumers still rarely need catastrophic care, the larger market allows transparency and reputation to have an effect as hospitals are at greater risk to lose customers. For the insurance market, an individual mandate can help pull us away from the adverse selection death spiral by preventing self selection and thereby keeping most of the population insured.

In other words, the government is not perfect, but neither is the market. By noting that the market for health care is fundamentally different from that for other goods such as cars, we can recognize why the standard assumptions for regular markets may not apply in the case for health care. As such, we must move forward with caution, with an understanding that true reform lies not with complete liberalization or complete regulation, but rather with a judicious mix of the two.
......................

Update 10/24/12 -- I was pleasantly surprised to find that this post was heavily discussed in Reddit. I just wanted to put a few points to defend myself from a series of very informative critiques.

• I never oppose market reforms like HSA's or cutting the employer healthcare tax deduction. That's not in my post at all.
• The point of the post is to highlight some frictions in the healthcare market that justify some sort of intervention -- market based (preferably) or command and control (not ideal).

Take-the-Best Statistical Model

Why do we do multiple regression?

Multiple regression is the workhorse of econometrics.  Almost every empirical paper in economics relies on it, and it also forms the basis for a large majority of political science research.  But is this a valid model for prediction?  How sensitive are the results?

As it turns out, the answer is "very".  Multiple regression is sensitive to a host of issues, including normality, linearity, and low error data.  But if the real world doesn't always fit these assumptions, why do we try to use the model to predict the real world?  One thing to notice when reading the empirical papers is that they often tell you that a certain coefficient is statistically significant, but rarely does one see a given confidence interval for that coefficient.  Of course, listing confidence intervals for every coefficient, especially when there are so many, is quite cumbersome.  Yet this convenient omission often leads us to be too confident about our estimates.  How much do we really know?

Behavioral economists have long criticized this model of human decision making because there's no feasible way that we can run a regression in our head and then make a decision.  At least, I know I don't.   Although many of my friends may have used excel spreadsheets to decide where to go to college, they did not end up basing their decision on some kind of complex regression model.  It's just too computationally intractable for everyday use.

Furthermore, one key problem of multiple regression is ecological validity.  We know that the regression model predicts the sample pretty well, but does it predict the future with any accuracy?  Especially if the future is highly variable and uncertain, why should we trust our Gaussian methods that are highly sensitive to outliers? According to Gerd Gigerenzer, the most accurate rules are often not the high powered intensive statistics methods.  Rather, fast and frugal algorithms that actually limit the information they evaluate can create more accurate results.

One of the prototypical fast and frugal algorithms Gigerenzer describes is Take the Best.  While multiple regression would look at all the data and perform various tests on individual data's contribution to the dependent variable, Take the Best does a sequential evaluation of a list of key determinants.  For example, multiple regression would decide between two restaurants by looking at all the data: food quality, wait time, location, parking spots.  It would then weight each input carefully according to an equation, and then look at the results of the equations for the two restaurants.  Whichever restaurant returns the higher value is the restaurant that's chosen.  On the other hand, Take the Best would look at whether the food quality gap is high.  If so, then pick the restaurant with better food.  If the gap is not high enough, move on to the next rule and repeat this simple process.  Computationally, this would require M+1 evaluations, which is in linear time and computationally quite tractable.

Gigerenzer applied this heuristic to predicting Chicago high school dropout rates.  Given two high schools and all the associated statistics: attendance rate, proportion of low income students, social science test scores, and more, what's the most accurate way to predict which high school had a higher dropout rate?   Gigerenzer and his fellow researchers took half the population of schools and built a Take the Best model and a multiple regression model to explain the data.  No surprise, multiple regression did better, predicting over 70% of the pairs correctly while Take the Best only managed around 65%.  Yet when the two models were tested on the other half of the population, Take the Best had about a 60% accuracy rate and multiple regression barely manged around the low 50%'s.

Surprising?  Gigerenzer in Gut Feelings explains:

But why did ignoring information pay in this case? High school dropout rates are highly unpredictable-in only 60 percent of the cases could the better strategy correctly predict which school had the higher rate, (Note that 50 percent would be chance.) Just as a financial adviser can produce a respectable explanation for yesterday's stock results, the complex strategy can weigh its many reasons so that the resulting equation fits well with what we already know. Yet, as Figure 5-2 clearly shows, in an uncertain world, a complex strategy can fail exactly because it explains too much in hindsight. Only part of the information is valuable for the future, and the art of intuition is to focus on that part and ignore the rest. A simple rule that relies only on the best clue has a good chance of hitting on that useful piece of information.

Looking backwards can hurt; you might end up blindsided by what the future can hold. The data might show trends that are only valid for the sample, and not the population as a whole.  Your results won't be ecologically valid if all the data is taken into consideration.  Especially since correlations change substantially over time, Gaussian methods are more likely to offer the pretense of knowledge than knowledge itself.

This has massive policy implications.  From Gigerenzer:

According to the complex strategy, the best predictors for a high dropout rate were the school`s percentage of Hispanic students, students with limited English, and black students-in that order. In contrast, Take the Best ranked attendance rate first, then writing score, then social science test score. On the basis of the complex analysis, a policy maker might recommend helping minorities to assimilate and supporting the English as a second language program. The simpler and better approach instead suggests that a policy maker should focus on getting students to attend class and teaching them the basics more thoroughly. Policy, not just accuracy, is at stake.

Yet with these policy issues at stake, it's surprising that fast and frugal algorithms aren't used more in economics research.  One disadvantage of take the best is that it doesn't give much quantitative accuracy.  It only tells which value is higher, but not by how much.  But how much does that matter?  While multiple regression may give more statistically significant coefficients, do we really have the power to tune the economy that much?  Even the best of natural experiments don't result in parameters that don't change through time.  Romer and Romer beautifully estimate tax elasticity, but how arrogant would a person need to be to build our entire tax policy based on an estimation from an almost 80 year old data set?  DSGE quantitative accuracy is such a joke that peripheral ad-hoc models are needed to make them even somewhat useful.

These alternative statistical tools are likely to add much to our insight of models.  The development of robust heuristics will be critical in a complex world, in which calculation becomes increasingly difficult and Gaussian methods increasingly fragile.