And Maybe That’s Just Not a Problem…

Earlier, Michael Cannon blogged about a recent discussion between him and Harvard’s David Cutler on the health outcome effects of increasing consumers’ price sensitivity for the costs of their care. (Translation: Have consumers deal directly with some of the costs of their care, using such mechanisms as co-pays, HSAs, etc.)

Cutler worries that increasing consumers’ price sensitivity will worsen Americans’ overall health. Though heightened price sensitivity has the positive effect of reducing the use of expensive health care of dubious value, it also reduces consumer use of health care that is of value — an outcome supported by the landmark Rand Health Insurance Experiment. The undesirable result, Cutler says, is worse health outcomes.

Cannon responds that a broader use of price-sensitivity mechanisms would invoke supply-side market responses such as lower prices. The undesirable result of worse health outcomes may thus be avoided (and, perhaps, better outcomes might result).

In following this discussion, I have a question: Is worse health outcomes necessarily undesirable, especially in this circumstance?

The value of having consumers deal directly with some of the costs of their care is not simply because doing so will reduce the use of dubious health care. The real value is that it increases consumers’ appreciation of the costs and benefits of their care and allows them to decide the tradeoffs between those costs and benefits.

Suppose an extremely expensive treatment would provide a consumer with a modest, but very real, positive health outcome. Some consumers may quite rationally choose to put their money toward other uses (ranging from necessities to a “Last Holiday”). On the “health outcome” measure, that decision would be a negative one, but on the “overall welfare” measure, that would be a positive.

Under a zero-price-sensitivity health care model, consumers wouldn’t have that choice. They would have already paid for their health care through their insurance premium (or worse, elected to forgo insurance because the premium was too expensive), and so any health care benefit they could receive under their health plan would be “use it or lose it.” So why not take the expensive treatment that yields modest results? Whereas, in a sensitivity model, consumers could quite rationally elect to keep their co-payments and HSA money in some situations.

This is not to say that people are wrong to worry about worse health outcomes, or about consumers making questionable choices. But the worriers do have an intellectual IOU outstanding: Do worse health outcomes necessarily mean worse welfare, if consumers can put their health care money toward other uses?

Better health outcomes are preferable to worse outcomes ceteris paribus. But, with price sensitivity mechanisms like co-pays and HSAs, the ceteris isn’t paribus. (My apologies to Latinists).

Global Warming Costs & Benefits

A few days ago, the British government released the Stern report, a voluminous study arguing that the costs associated with stabilizing carbon dioxide concentrations at 550 parts per million were far less than the costs associated with doing nothing. Although the study acknowledged rather large bounds of uncertainty, the median estimates therein suggested that business as usual (that is, we do nothing) would mean a loss of 5–10% of global GDP every year forever. Most of those harms, however, could be avoided if we spent 1% of global GDP to cut back on greenhouse gas emissions.

There are very good reasons to suspect that Stern’s estimates regarding the cost of cutting back on greenhouse gas emissions are too low and that the damages forecast by Stern are too high. The underlying assumptions of the analysis producing Stern’s estimates have been well dissected by statistician Bjorn Lomborg, climate scientist Roger Pielke, Jr., and economist Richard Tol. But for the moment, let’s put those complaints aside.

My colleague Peter Van Doren and I have done three present value calculations assuming that business as usual (BAU) will reduce global GDP by 2%, 5%, and 10% beginning in 2056 and then in each and every year through the end of time. Don’t worry about the silliness of such a proposition. Oddly enough, once you try calculating beyond 200 years, the numbers don’t really change much given the need to discount future costs and benefits by 5%.

First, we calculated the cost of using 1% of GDP every year through the end of time to reduce greenhouse gas emissions. The net present value of that cost is $15,541 per person in the United States.

Then, we calculated the benefits for U.S. citizens (global GDP figures are pretty dodgy, so we stuck with U.S. GDP figures for the purposes of this exercise). They amount to $36,447 is you accept -10% GDP as your BAU scenario, $18,239 if you accept -5% GDP as your BAU scenario, and $7,295 if you accept -2% GDP as your BAU scenario.

In other words, Stern’s investment advice makes sense only if you think that warming will hammer GDP by 10% a year. You don’t gain much at all from emission cuts, however, if you think GDP will only drop by 5% a year if we do nothing. And if you think warming will only cost the global economy 2% of GDP every year (the “concensus” belief among economists, which comes from a widely cited analysis from Yale economist William Nordhaus), then Stern’s investment advice is shere lunacy.

And that’s not even taking into consideration the fact that reducing greenhouse gas emissions might produce no benefits at all. The latest IPCC report — as all other reports before it — acknowledge that the evidence that anthropogenic emissions are primarily driving the warming we’ve detected is strong but circumstantial. Scientists disagree about how large the chances are that we’re wasting our time cutting greenhouse gas emissions, but there’s no disagreement within the latest IPCC report that there’s a chance that anthropogenic emissions are not particularly important factors in climate at present.

Is global warming insurance a good buy? Probably not. And that’s particularly true given the fact that the relative poor (us) will pay the premium so that the relatively rich (our children and grandchildren) will get the benefits if there are any. For example, since 1950 real GDP per capita has increased by about 2% per year. Given that growth rate, U.S. GDP per capita in 100 years would be $321,684 in current dollars, or more than seven times higher than it is at present ($44,403). If global warming cuts GDP by even 10%, then GDP per capita will be $289,515 in 2106 rather than $321,684. Would anyone, let alone liberals, ever propose a 1% tax on those who make $44,000 to create benefits for those who make $289,000?

Winning, But Losing

When the government accuses someone of a criminal offense, it typically proceeds to exert enormous pressure on the accused to surrender the right to a jury trial. Fewer than 10 percent of the criminal cases in America go to trial. Plea bargaining dominates the system.

Sometimes a person will insist on a trial. This is risky because if the government gets a conviction, it will mete out extra punishment because it was forced to go through the “trouble” of a trial. But if the jury sides with the accused, the state loses, right? Wrong.  The state can still unleash punishment after an acquital. 

Hard to believe, I know. Here’s a recent ruling (United States v. Ibanga) in which the Court is at pains to explain the law.

After an eleven-day trial, a jury acquitted defendant Michael Ibanga of all of the drug distribution charges against him and one of the two money laundering charges against him in the Indictment. The single count of which defendant Ibanga was convicted typically would result in a Guidelines custody range of 51 to 63 months. However, the United States demanded that the Court sentence defendant Ibanga based on the alleged drug dealing for which he was acquitted. This increased the Guidelines custody range to 151 to 188 months, a difference of about ten years. …

What could instill more confusion and disrespect than finding out that you will be sentenced to an extra ten years in prison for the alleged crimes of which you were acquitted? The law would have gone from something venerable and respected to a farce and a sham.

From the public’s perspective, most people would be shocked to find out that even United States citizens can be (and routinely are) punished for crimes of which they were acquitted.

[…]

The Sentencing Guidelines have accomplished much good in the course of standardizing the sentencing process. Similarly, the Fourth Circuit’s post-Booker presumption approach is a politically savvy parry to the thrust of those who call for more stringent measures, such as the expansion of mandatory minimums. However, it is a charade to say that the Sixth Amendment violations inherent in the Guidelines are cured simply by intoning the word “advisory.” Saying something is so does not make it so.

One of Charles Dickens’ characters, Mr. Bumble, famously observed, “If the law supposes that, … the law is an ass– an idiot.” Charles Dickens, Oliver Twist 463 (3d ed. The New American Library 1961). He was referring, of course, to a legal fiction that had no basis in reality. Many of our fellow citizens believe that Mr. Bumble was right — that the legal process is rigged through sleights of hand that defy common sense. It would only confirm the public’s darkest suspicions to sentence a man to an extra ten years in prison for a crime that a jury found he did not commit. (Italics added.)

This case stands out because the ruling is bitterly critical of this aspect of sentencing law. Most court rulings affirm this stuff all the time, without comment. 

I should point out that the state is powerless to do anything in the typical TV drama situation where there is a single murder charge that the jury is considering against someone. If there’s a single charge and the jury says “not guilty,” the prosecutor cannot do anything about that result. But that’s TV. Nowadays, when a case goes to trial, there are multiple charges. And if the jury comes back with a single “guilty” verdict, the government might still drop a ton of bricks on the defendant — even if the jury said “not guilty” on a dozen other charges.

Does the existence of such a power influence a person’s decision with respect to whether he ought to “waive” his right to a jury trial in first place — and accept a plea bargain? What do you think?

The constitutional right to a jury trial is on life-support and that’s where the government wants it. Go here for Cato articles related to sentencing.

P4P All Over the Private Sector

At yesterday’s Cato policy forum on pay-for-performance (P4P) in Medicare, I argued the Medicare bureaucracy should stay out of P4P largely because Medicare would ruin the idea. A Medicare-administered P4P program would be less flexible than private efforts, more likely to harm patients, and the very providers that P4P aims to discipline would have way too much say in a Medicare P4P program. I recommended confining P4P to private Medicare Advantage health plans. Read my full argument here.

Harvard’s David Cutler argued that Medicare should get involved in P4P because private insurers didn’t have the purchasing power to really force providers to change. At the time, I was unaware of this study by Meredith Rosenthal and her colleagues in this week’s New England Journal of Medicine. They report:

More than half the HMOs, representing more than 80% of persons enrolled, use pay for performance in their provider contracts. Of the 126 health plans with pay-for-performance programs, nearly 90% had programs for physicians and 38% had programs for hospitals.

That probably doesn’t match Medicare’s purchasing power. But it does suggest that P4P can gain a toehold through the private sector.

Global Warming Science — Utter Garbage

… at least, that’s what one might think about the state of the literature today after reexamining a paper by Isabelle Chuine et al. that was published a couple of years ago in the journal Nature (subscr. required). 

Chuine et al. claimed to have developed a method for estimating summer temperatures in the French wine region of Burgundy from 1370 to the present based on the dates that grapes were harvested. Using this method, the authors asserted that the summer of 2003 was the warmest summer in Burgundy since 1370. The study was offered up as yet one more piece of evidence that global warming is running amuck. 

But not so fast — it turns out that the estimates offered by Chuine et al. have absolutely no relation with observed temperatures and that no one ever bothered to check whether their estimates matched hard data when the two coincided. In a forthcoming paper for Theoretical and Applied Climatology, Douglas Keenan tears the Chuine paper limb from limb

More importantly, Keenan calls attention to the overall shoddiness of the scientific press today and the researchers publishing therein:

What is important here is not the truth or falsity of the assertion of Chuine et al. about Burgundy temperatures. Rather, what is important is that a paper on what is arguably the world’s most important scientific topic (global warming) was published in the world’s most prestigious scientific journal with essentially no checking of the work prior to publication.

Moreover — and crucially — this lack of checking is not the result of some fluke failures in the publication process. Rather, it is common for researchers to submit papers without supporting data, and it is frequent that peer reviewers do not have the requisite mathematical or statistical skills needed to check the work (medical sciences excepted). In other words, the publication of the work of Chuine et al. was due to systemic problems in the scientific publication process.

The systemic nature of the problems indicates that there might be many other scientific papers that, like the paper of Chuine et al., were inappropriately published. Indeed, that is true and I could list numerous examples. The only thing really unusual about the paper of Chuine et al. is that the main problem with it is understandable for people without specialist scientific training. Actually, that is why I decided to publish about it. In many cases of incorrect research the authors will try to hide behind an obfuscating smokescreen of complexity and sophistry. That is not very feasible for Chuine et al. (though the authors did try).

Finally, it is worth noting that Chuine et al. had the data; so they must have known that their conclusions were unfounded. In other words, there is prima facie evidence of scientific fraud. What will happen to the researchers as a result of this? Probably nothing. That is another systemic problem with the scientific publication process.

Unfortunately, few enviro-beat reporters take the time to critically examine the avalanche of papers crossing their desk claiming this, that, or the other. Fact-checking rarely if ever occurs. The Globe and Mail, for instance, was absolutely breathless about Chuine’s findings. After all, what a topical hook: global warming will screw up your pinot noir!

The lesson here is that you can’t assume that anything in the scientific literature has ever been given even a cursory critical review prior to publication. Peer-review means nothing. Swallow this stuff at your own risk. 

When Patients Change, Do Providers Change Too?

Harvard’s David Cutler visited Cato yesterday to participate in a small group discussion about cost-effectiveness in medicine, and also in a panel on improving quality in Medicare. (You can watch the latter event here in a couple of days.) My colleague Arnold Kling blogs about issues discussed at both events. 

I am struck by one issue that emerged, which has to do with price-sensitivity, provider behavior, and health outcomes. Cutler argued that when patients are more price-sensitive (i.e., when they have to pay for more of the cost of their medical care), they tend to cut back both on care that would have done nothing for them, and on care that would have helped them. He postulates that if we were to move all Americans into health savings accounts (HSAs), thereby making patients more price-sensitive, we would see worse health outcomes than we see now. 

I am skeptical of that prediction. I think that if the move to HSAs were confined to a small, randomly selected subset of the population (call it “Rand II”), Cutler’s prediction would be more plausible — though by no means certain. There is precious little evidence that suggests — and it does no more than suggest — that for some patients, greater price-sensitivity leads to worse health outcomes. 

However, even if we assume that Rand II would show that greater price-sensitivity leads to worse health outcomes, it does not follow that we would get the same result were the entire population made more price-sensitive. The reason is that with a population-wide shift, the supply side of health care markets would respond to the enormous change on the demand side. Faced with patients who are less eager to consume medical care, providers would have to do a lot more to sell their services, including:

  • conducting research on the usefulness of their services,
  • improving the quality of their services,
  • lowering their prices, and
  • educating patients about the value of their services.

These responses should enable patients to make smarter decisions about what to consume and what to avoid. Instead of having patients cut back equally on beneficial and useless care, they would cut back on useless care more, having more help discerning between the two. Downward pressure on prices should make cutting back on beneficial care even less frequent.

MIT economist Amy Finkelstein demonstrates that the supply side of medical care does respond to demand-side changes. For 30 years, economists believed that the expansion of health insurance (which reduced price-sensitivity) had a relatively small impact on the growth of health spending. That belief was based on the effects of a demand-side study (Rand I), which was too small to induce or measure any supply-side responses to the change in price sensitivity. Using a data set that does capture and allow her to measure supply-side responses, Finkelstein estimates that the effect that the expansion of health insurance had on health spending is six times greater than the demand-side-only experiment Rand I suggests. 

Casual observation suggests that supply-side responses are helping price-sensitive patients make better choices right now. At the same time that HSAs and other insurance options are making millions of patients more price-sensitive, insurers and entrepreneurs are furnishing more of the price and quality information that patients need.

It would be foolish to claim that the supply-side response to price-sensitive consumers would be so great that patients would have perfect information and would never make mistakes. Yet most opponents of making patients more price-sensitive make the equally foolish assumption that there would be no supply-side response to the new incentives coming from the demand side. I say “most” because Cutler and others are not in this group. If I understood Cutler, he acknowledges that there will be such supply-side responses, and that we have no way of knowing whether or how much they would improve health outcomes.

True enough. But it’s something like 50 percent of the debate over HSAs and health outcomes. T’would be nice to have opponents of HSAs and the like acknowledge and engage it.