Excessive whining about uncertainty

This article is one of many that complains about the inhibiting effect of uncertainty on business investment and other good things. Business people are supposed to be used to the idea of taking risks. The biggest risk they take is the decision to start a business at all, given the number that fail. For small risks, like variations in tax rates, it would seem that expected-utility theory is the way to go. You judge the probabilities, multiply them by the outcomes of various options, and pick the option with the highest weighted sum. They used to teach this in business schools. It is simple enough to teach in high school. A business that thought this way would, statistically, come out ahead of its competitors.

But a second problem is that people (i.e., those in business, overreact to salient risks and ignore the huge risks that are off screen. Compared to those huge risks, uncertainty about taxes is almost trivial. I'm reminded of a story about someone who presented a precise economic prediction of the price of oil over the next few years. One of the listeners asked, "Er. What about war in the middle east?"

A final comment. Some of this is not about risk so much as ambiguity, the psychological state of feeling that you are lacking crucial information. This is, as I have argued, something of an illusion, as we are always lacking crucial information whenever we cannot predict an outcome with certainty. We over-react when certain missing information, like what Congress will do with taxes, is called to our attention.

Action, omission, and the health-care law

Judge Henry E. Hudson has just said that the insurance mandate of the new health-care law is unconstitutional. The center of the opinion, as I understand it, is that the constitutional power to regulate interstate commerce extends to economic "activities" but not to omissions of economic actions, such as not purchasing health insurance. It seems that Hudson's reasoning on this is empirical, namely, that prior court decisions on the scope of the commerce power have all concerned actions. Because no case has concerned an omission, then it must be that the power does not extend to omissions. Apparently - and I stand to be corrected if wrong - no prior court decision has explicitly said that omissions don't count. (Apparently the most relevant part is around pp. 20-21 of the opinion, which is available in the NY Times web page.)

A simple alternative explanation of the empirical facts is that most economically significant activities are actions, so an omission has never come up before. The judge here seems to be making his own distinction out of what seems to me to be thin air. (Again, I stand to be corrected if wrong.)

The decision seems to attest to the psychological salience of the act/omission distinction. As the lawyer for the health-care bill pointed out, the application of that distinction is in any case unclear, especially in this case. (Jonathan Bennett, in "The act itself", might be seen as saying that is unclear more generally.)

Hindsight bias and Obama's "capitulation"

Paul Krugman and many other Democrats are saying that President Obama and the Democrats made a mistake by putting off until now the question of what to do about the Bush tax cuts. Clearly - as David Leohhardt, Nate Silver, and others have argued - the Democrats are not in a strong bargaining position now that the Republicans have threatened utter chaos (hold up all the other legislation, kill all extensions of all tax cuts, etc.) unless the Democrats extend the unproductive and unfair tax cuts for the rich.

But the wisdom of hindsight is wonderful. Where were the pundits saying "Now is the time to bring up the extension of the tax cuts"? And, looking back over the last two very difficult years, even now, can someone point to a specific time when the Democrats could have done this? My hunch is that there were always more important things.

So long as one of the two major parties is willing to sacrifice the well-being of the nation and the world for its misguided ideology, we have a problem that is not easy to solve. In this respect, the U.S. is much like Israel, where the equivalents of the Republicans are in control.

Added later: Seems I'm wrong. Here is what should have been done, in hindsight. But the congress had to do it, not the president.

Provenge

Medicare will now pay for Provenge, which costs $93,000 and prolongs life for about four months, although this is the median not the mean. This comes to about $279,000 per year, and probably more for a quality-adjusted life year. At the hearing where this was decided, may objected to the possibility that Medicare would consider costs, calling it "rationing". What precedent does this set, and what else could Medicare cover if it were to value life so highly?

Consensus statement on morality

http://www.edge.org/3rd_culture/morality10/morality_consensus.html is a consensus statement of a few people invited to a meeting.

Near the bottom of
http://www.edge.org/discourse/morality.html
is my attempt to say everything I believe about moral judgment in 500 words.

World government

I keep reading about problems that are best solved at the world level, or at least at some level that involves many nations: climate change; overfishing; illegal drugs (if they are a problem); deforestation (which is related to the importation of wood); trade; protection of endangered species like whales; terrorism; and on and on. One of the lessons that Howard Raiffa, Max Bazerman and others have taught us is that negotiation about several issues at once allows log-rolling, that is, trade-offs in which each party gives up something it cares about less in return for something it cares about more. One way to prevent this kind of win-win strategy is to negotiate one issue at a time, but that is exactly what we do at the world level. We have organizations for fish, organizations for climate change, and so on, but no supra-national organization that permits nations to make trade-offs with all these issues on the table. The lack of this sort of world government is a serious impediment to world progress.

I was reminded of this issue this morning when I discovered a project called Democracy Unbound, which aims to study the prospects for supranational government from the perspective of several scholarly disciplines.

Action and omission in the U.S. Supreme Court

This interesting article shows how the Court accepted a distinction between action and omission as relevant. Harm caused by omission becomes a matter of "positive liberty", which is apparently not protected by the Constitution. In particular, even after a child welfare agency involved itself in a case, the agency cannot be sued for failing to protect the child from terrible harm. Of interest is that this is not the sort of omission that anyone can commit; the agency was already involved, hence, by everyday moral standards, "responsible".

Direct democracy

I was supposed to be in a video for a group that advocates direct democracy (initiative and referendum at the national level in the U.S., among other things). I said I had doubts. ("You see Switzerland. I see California.") But they wanted to do it anyway. In the end we couldn't schedule it. I had thought about what I would say, and here it is.

My first worry comes from my work (with Ed McCaffery) on the isolation effect. When people are asked about a given proposal by itself they isolate that proposal and ignore its side effects, its costs, and to some extent the alternatives to it. This leads to inconsistencies.

For example, majorities seem to favor lower government spending, lower taxes, a balanced budget, but more spending on health care, education, and social security. This happens when they are asked about these things one at a time. Yet the combination is impossible, even if we take into account their willingness to cut military spending and other parts of the U.S. budget.

Or, for a slightly more subtle case, they favor progressive taxation, couples neutrality, and marriage neutrality. Couples neutrality means that couples with the same income pay the same tax. Marriage neutrality means that marriage does not affect the total taxes paid by the couple. These three things are mathematically impossible to have at once.

The first problem is thus that referenda do not protect us against proposals that are incompatible with other things that we want.

The second problem is that, when we consider one proposal at a time, we prevent "log rolling". A better term might be mutual back scratching. This is what legislatures do, when they work as they are supposed to work. Each side gives in on something that it cares about less, in order to get something that it cares about more.

A recent example is the energy bill that the U.S. Senate almost got to debate. The idea was to get a price on carbon, for one side, in return for increased development of nuclear power, and increased off-short drilling, for the other side. It would still be a good deal, even without the off-shore drilling. This sort of thing could not happen unless someone happened to put together deals like this in an entire referendum proposal, which is unlikely, because most of these proposals come from interest groups on one side and they don't negotiate with their opponents.

I'm sure there are places where referenda can do some good, but we need to consider these problems too.

Presumed consent for research

Although presumed consent for organ donation
may not be a panacea, it does seem to help. (Showing that presumed consent is correlated with other determinants of high donation rates does not determine the direction of causality.) Why not do the same for certain kinds of research?

A case in point is the use DNA samples for research on the genetics of disease. In a recent case, the University of Arizona returned DNA samples to a Native American tribe, the Havasupai, and paid them $700,000, to settle a lawsuit about informed consent. The tribe members had donated DNA for what they thought was a study of diabetes, but the DNA was used for other purposes. (The dispute about what they were actually told was not settled. See the news article in Science, April 30, 2010, p. 558.)

The main issue here seems to me to be property rights. Some of my recent research has shown that people think of property rights as dictated by deontological (non-consequential) moral rules. But economic and utilitarian analyses of property argue that it exists because of its social value in terms of beneficial consequences, such as encouraging development of property, production, and innovation.

In the case of organs and DNA samples, it is hard to see how functions like these are relevant. The social value of not having property rights in these cases seems quite high, much higher than any benefit of property rights might be.

The idea of presumed consent does not completely remove property rights. It removes them unless they are positively asserted. This limitation allows people who want property rights an option to take them. In general, we might think of such desires as anti-social. However, some people might not trust others to look out for their welfare. They might believe that organ donors are allowed to die more quickly or that DNA donors will be found to have some condition that others can use against them. The option to assert property rights may help to make the users more trustworthy, lest they become more generally untrusted.

My hunch here is that perception of the violation of rights is highly dependent on social norms and conventions. For example, when we write in blogs we assume that our writing is in the public domain. It is polite to attribute quotations but not illegal not to do so, unlike other writing. I don't know the law on this, but this is what I assume, and I suspect that others do too. In another example, when I go to a tourist attraction I do not think I have the right to consent to being in someone's photograph of the scene. So long as we know what our rights are, these expectations are generally harmless.

More generally, in cases like these, people get upset when they think that social norms and conventions actually give them certain rights. If they did not think so, they would not get upset. This argument applies, of course, only to cases where there is little or no justification for providing a right in the first place, in terms of the social value of such provision.

Judgments in the Goldman Sachs case

An article in today's (4/20/10) New York Times on-line edition ("A difficult path in Goldman case") discusses Goldman Sachs's likely defense against the fraud suit against it, outlined in two letters written by the company: "... The letters went on to argue that, contrary to the S.E.C.’s assertions, Goldman disclosed all information about the deal that was material. In particular, the letters drew a sharp distinction between information about the security, which the company said it provided in full, and information about Mr. Paulson’s role.

"The second letter said, 'It is this concrete information on the assets -- not the economic interest of the entity that selected them -- that investors could analyze and use to inform their decisions.'"

"To win its case, the S.E.C. must prove that Goldman was not merely silent about Mr. Paulson’s role but actually gave investors the wrong impression, experts in securities law said."

Implicit here is the idea that the investors' had access to everything that John Paulson knew when he chose the securities in question. Thus, they should be able to judge their value. But the investors did not know that Paulson had chosen the securities exactly because he thought that they were particularly overvalued, compared to other securities he might have chosen. Indeed, the investors did not know which securities Paulson rejected. The question is whether these things could have changed any investor's mind.

From a psychological (not necessarily legal) perspective, such missing information seems relevant. In cases of difficult judgment, people are known to be influenced by the judgments of others, and rightly so. Paying attention to others' judgments often increases accuracy. Thus, knowledge about why Paulson chose the securities should have, and probably would have, influenced the decisions of investors.

I know of no literature on the effect of someone else's rejected choices on judgment. Perhaps the S.E.C. should do (or fund) some psychology research on this point.

Mammograms under 50

A recent news article in Science (2/19/2010, vol. 237, pp. 936-938) reviews the issues concerning breast-cancer screening for women between 40 and 50. What seems clear is that NOBODY involved in this controversy has done even the most rudimentary decision analysis, yet the question is obviously a good one for such analysis.

According to the article, the major issues are these:

1. Many cancers detected by mammograms in younger women will be detected anyway, even without mammograms. Of the cancers detected without mammograms, most will be cured. The upshot is that, while mammograms do detect cancers (and many women who have had cancers cured after detection with mammograms think that the mammograms saved them), they do not reduce deaths from cancer very much.

2. Still, mammograms do prevent cancer deaths. The effect is small, and its size varies from study to study. The lowest figure seems to be that cancer deaths (which are rare anyway) are reduced by 15%. Or, conservatively (it seems), one death is prevented for every 1900 women screened routinely between ages 40 and 50. (The article is a little unclear about this.) Screening at age 50 prevents one death for roughly every 1300 women screened.

3. So why not? The article puts aside the cost. Let's say the cost is $100. This comes to $100 times 1900, which comes to $190,000 to prevent one death. A very good deal if you ask me. (I suspect that standard insurance policies cover treatments that cost $100,000 to save one year of life. Surely prevention of cancer death in a 45-year-old will save many years of life.) So cost does not seem to be the problem.

4. The article also puts aside the increased risk from x-rays from repeated screening. I'll take their word for it that this is a small factor. (Wikipedia agrees, and cites a reference. Someone correct me if I'm wrong.)

5. The big trade-off according to the article is the risk of false positives. And the big problem with false positives is anxiety. Note that a biopsy is typically not immediate after a positive mammogram, so the anxiety can go on for some time, possibly weeks. The article gives a ratio of 5 biopsies for every case of cancer detected, which means that 4 are false positives. If 10% of these cancers would be fatal if undetected, then this means that the risk of a false positive is about 40 times the risk of death from undetected cancer. (I'm guessing at the fatality rate. It may be much lower.)

It is conceivable that some women would rather take the risk of undetected cancer than the much higher risk of anxiety from a false positive. I would think that research is needed that compares the relative utilities of death from cancer and anxiety from false positives. It may turn out that women differ enough so that they differ in the option that would maximize utility for each patient.

It seems likely to me that very few women would have a sufficiently high disutility for anxiety as to tilt the decision toward no screening after age 40. But the research remains to be done.

Again, what strikes me here is the total absence, so far as I can tell, of analysis of this question in terms of expected utility. Why is this idea so distant from even scholarly discourse?

I welcome comments by email, and I may post a follow-up: baron@psych.upenn.edu.

Troubles understanding health-insurance reform

The health bills passed by the U.S. house and Senate are complex. Some polls suggest that citizens like them a lot better once their provisions are understood. Both bills contain a logical sequence:

1. Prevent insurers from discriminating on the basis of health risks, such as pre-existing conditions.

2. If this were the only change, insurance costs would go up, and low-risk people would drop out, leading to more increases as insurance covered only those at risk, more drop-outs, more increases, and so on. This has happened when compulsory auto-insurance laws have not been fully enforced. Thus, to make sure that this does not happen, we must have a mandate. And that means a penalty for not having insurance.

3. But the cost of compulsory insurance would hit the near-poor very hard. (The poor are already covered by Medicaid.) Thus, they need a subsidy.

4. To provide this, we need additional funds, either through taxes on the rich (the House bill) or on expensive employer-provided health insurance (the Senate bill).

In sum, any adequate reform has these four elements:

E - equality of those who are healthy and at risk;

P - a penalty if people do not get insurance;

L - help for low-income people;

T - a tax, or reduction of a tax deduction, to pay for L.

To examine people's understanding of these issues, I gave a questionnaire to a panel of people who do questionnaires on the World Wide Web for Pay. I waited for 90 responses. The panel is typical of the U.S. population in income and education, although most are women, and some are foreign.

After defining "health" and "at risk", I started with a basic description of a full plan, EPLT (for all four elements):

"Insurance companies must treat at-risk and healthy people the same (for example, cannot exclude people with pre-existing conditions or charge them more), people must pay a penalty if they do not have health insurance, the government will help low-income people pay for health insurance, and the government will recover the funds by some combination of increased taxes on those with high incomes and reduced tax deductions for expensive employer-provided health insurance."

Then I presented 7 different modifications of the plan: E, EP, EPL, EL, ELT, L, and LT. Then I presented EPLT again. After each of the 9 plans, I asked the several questions, which I list at the end. All the questions, except those asking for an overall opinion, were introduced with: Compared to the present system, how would this proposal affect each of the following?" (I plan to present more details in a later post.)

The most interesting result was the one for question about effects on "The number of healthy people who have health insurance". For the E proposal alone, which specified, "Insurance companies must treat at-risk and healthy people the same (for example, cannot exclude people with pre-existing conditions or charge them more) AND NO OTHER CHANGES", many subjects thought that the number of healthy people with health insurance would INCREASE (40 thought it would increase, 12 decrease, p=.0001). This is diametrically opposed to the economic argument. That argument requires thinking ahead about how insurance companies will respond to the new regulation, and how consumers will respond to that response. It is like thinking two moves ahead in chess. Work I have done with Ed McCaffery (McCaffery and Baron, 2006) suggest that many people rarely think beyond the primary or intended effect of a policy.

It seems that many of my subjects did think through to the first move, however. When asked about "The overall cost (including insurance) to a healthy person who has insurance now" of E alone, 54 said it would increase and only 9 said it would decrease. Thus, the problem seemed to be in the understanding of the effect of price on demand. As the price goes up, fewer people are willing to pay it.

Conclusion: Opposition to the health proposals is based in part on lack of understanding of the four-step sequence outlined above. People may be sensitive to argument about the need for a mandate to prevent people from dropping out because of higher prices ... and to
keep prices lower by making sure that healthy people stay in the pool.

#################################

All questions:

What do you think of this proposal, compared to the present system:

much worse -- somewhat worse -- not sure --somewhat better -- much better

Compared to the present system, how would this proposal affect each of the following?

The number of healthy people who have health insurance:

fewer -- not sure -- more

The number of at-risk people who have health insurance:

fewer -- not sure -- more

The overall cost (including insurance) to a healthy person who has insurance now:

less -- not sure -- more

The overall cost to a healthy person who do not have insurance now:

less -- not sure -- more

The overall cost (including insurance) to an at-risk person who has insurance now:

less -- not sure -- more

The overall cost to an at-risk person who does not have insurance now:

less -- not sure -- more

The cost to taxpayers who already have insurance, including workers who already have insurance through their employers:

less -- not sure -- more

The cost to a person with lower income:

less -- not sure -- more

The U.S. government's budget deficit:

lower -- not sure -- higher

Now that you have considered all these effects, what do you think of this proposal, compared to the present system:

much worse -- somewhat worse -- not sure -- somewhat better -- much better

Reference:

McCaffery, E. J., & Baron J. (2006). Isolation effects and the neglect of indirect effects of fiscal policies. Journal of Behavioral Decision
Making, 19, 1–14.