An Opportunity For A Pharmaceutical Policy Experiment?

The Toronto Globe and Mail ran a story ($$$ subscription required) a little while back about cancer patients treating themselves with DCA, the drug which, according to reports out of the University of Alberta shows promise as a cancer treatment, based on its effectiveness in lab rats. Orac, who blogs at Respectful Insolence, ran a post here cautioning against getting your hopes up too high, based in part on the fact that it's easy to cure cancer in rats. If cancer were just a disease of rats we'd have wiped it out by now. DCA hasn't been tested in human as a cancer treatment, indeed hasn't yet been put through clinical trials as a cancer treatment. Still, and not surprisingly, cancer patients for whom all other treatments have failed are eager to give DCA a shot.

The complicating factor about DCA is that it's been around for a long time, and is now off patent. That means that if it's discovered that a particular dose of the stuff is effective against cancer, anyone who wants to will be able to manufacture it. That in turn means that there will be a lot of competition in the market supplying it - all the generic drug firms will make their own version - but it also means that nobody would be able to recoup the costs of the research necessary to find out whether it really does work on humans. And as Orac reminds us, over the years there have been lots of breakthroughs in cancer treatment in the lab, most of which turned out not to be effective in people. But even if it does work miracles, somebody has to pay for the testing, and Big Pharma is taking a lot of flack for not being willing to fund clinical trials on a drug that there's not just no hope of making a massive profit on, but not even a hope of recovering research costs on.

I'm eventually going to get around to making a suggestion about the economic issues involved, but first let's look at some background, from the Globe.

The story in question is by Anne McIlroy, published on 1 June / 07, and headed:

The battle over a cancer pill

A few snips:

[cancer sufferers] started self-medicating this spring after University of Alberta researchers announced that DCA (dichloroacetate) dramatically shrank tumours in rats without damaging healthy cells. But the Edmonton team was having trouble finding money to see if DCA works in humans.

DCA already has been in use for a long time to treat rare metabolic disorders, so it can't be patented as a new drug. And without a patent even a wonder drug won't make huge profits – major pharmaceutical companies won't invest the hundreds of millions of dollars required to test it and bring it to market.

Note the hundreds of millions of dollars bit. Let's be a bit more clear than the Globe - not only will they not make massive profits on it, under the current research incentive structure (and that's what I'll eventually come back to) they'll take a pretty hefty loss, guaranteed.

Dr. Evangelos Michelakis, the principal investigator, has urged patients to wait for valid results, warning that they could poison themselves by taking DCA ordered over the Internet. The University of Alberta so far has raised more than $200,000 toward a small, initial clinical trial.

But from Liverpool to Louisiana, people with cancer say they can't afford to wait.

They are experimenting on themselves, ordering DCA, sharing their results on a website and putting together a database in an attempt to figure out if it is working.

Their do-it-yourself approach illustrates the Internet's growing power to help patients circumvent – and perhaps undermine – traditional medical research on drugs. The Edmonton scientists have warned that by taking it on their own, patients are jeopardizing the chances of a real clinical trial ever taking place: What if anecdotal reports spread that DCA doesn't work, or makes people sicker?

“It's destroying the efforts to do this right,” Dr. Michelakis recently told the science journal Nature.

He is so opposed to patients dosing themselves that he would not be interviewed for this story, because it includes interviews with people who are taking the drug.
.....................................................................................................
But the patent issues were a huge obstacle when it came to getting financial backing. Since the DCA molecule can't be patented, a researcher can claim intellectual ownership only on how a drug is administered to fight a particular disease. “Use” patents are generally considered weaker than standard patents, and are much less valuable to pharmaceutical companies.

When the researchers couldn't interest any drug companies, the University of Alberta started fundraising. Contributions, anywhere from $5 to $1,000, have been steadily flowing in.

I read somewhere that the University of Alberta held a Use patent on DCA but let it lapse, figuring that the return they would make on it wasn't worth the fees they would have to pay to keep the patent up. Dr. Michelakis has filed a use patent on it at some point.

Helping cancer patients treat themselves with DCA clearly wasn't what Dr. Michelakis had in mind when he filed the patent, or when he published his results in January in the medical journal Cancer Cell.

But he may have underestimated the power of the Internet, and the appeal of the DCA story. In January, he gave interviews to reporters from all over the world, but now he doesn't want to talk to the press about DCA.

“Dr. Michelakis is a highly ethical man and feels a great deal of responsibility for his research and the health of patients,” University of Alberta spokeswoman Jo-anne Nugent said in an e-mail.

“We have found that every time we participate in such publicity, it plants a seed of hope and results in more patients taking DCA.”

Taking it in doses based on the details which Dr. Michelakis reported in the patent - patents give a certain amount of protection, but the flip side is that in order to get one you have to put a lot of information into the public domain.

OK, so big pharma won't fund research into it. But what about the generic manufacturers? One thing you don't often hear mentioned in the drug debate is that generics' markup of price over cost is often higher than the markup for brand-name drugs: possible since the generic manufacturers don't have to pay research costs. And they can be pretty profitable - that's why, back in 2005, Novartis was willing to pay over $8 billion to buy two European generic manufacturers.

So what about in this case? They don't seem to have been asked, nor to have volunteered, to date, but the Globe piece does include a bit of relevant evidence:

If it seems hard to believe that a promising and inexpensive treatment for cancer would never get tested, listen to Judes Poirier, director for the Centre for Studies in Aging at McGill University.

“I have been there before,” he says. “This is a recurring story. It happens in many different fields.”

Researchers are increasingly going back to look at older drugs to see if they might be good at treating more than one disease, he says.

In 1996, Dr. Poirier found that a heart-disease drug called Probucol showed promise in keeping Alzheimer's disease at bay. It worked in lab animals, and a small human trial conducted in the late 1990s was also encouraging.

But there is no patent for Probucol – like DCA, it has been around too long. Dr. Poirier filed a use patent and naively hoped that Probucol would be tested soon in a large clinical trial because it already had been proved safe in heart patients.

He couldn't interest any pharmaceutical companies. They wanted him to change the drug's chemical formula just enough to get a new patent, but that would mean time-consuming tests to prove that the new version was safe.

He turned to a generic drug company, one that sells cheaper versions of name-brand drugs once their patents have expired.

Too risky, he was told. Their business approach didn't involve trying an old drug against a new disease.

But if someone else wants to run the trials, they'll make the drugs. Generic manufacturers manufacture, they don't research.

(Actually, that's too broad a statement. There are a few generic drug companies, in Canada and India in particular, which are doing their own original research, and more power to 'em.)

OK, that's the background. Now, what about the economics?

A lot of people argue that the patent system is not a good fit for the pharmaceutical sector. They see the monopoly power it gives patent holders as a major source of loss of economic welfare. There have been a number of alternatives suggested, but one of the most popular ones is the prize approach.

Basically, somebody (meaning the government) offers a monetary prize to, in this case, the first firm which can show that DCA is useful for treating cancer in humans, and finds a formulation in which it can be brought to market. Once the conditions of the prize have been met, the government owns the patent on the cancer treating version of DCA, and allows any generic firm which wants to to manufacture the stuff.

Surely DCA is a good case to use to test drive the prize idea. The drug has already shown promise in the lab, and because it's been around for a while it won't need to go through Phase I testing (that's the phase of drug testing where you give a bit of the stuff to a handful of humans, to see whether simply taking it kills them). So it can go directly to Phase II: the phase at which you try it on a small sample of cancer patients to see whether the promise it showed when it was given to rats seems still to be there when you give it to humans. If a drug fails Phase II trials, there's no point taking it forward to Phase III, where it's tested on large numbers of cancer patients.

Since we're designing a policy experiment here, let's simplify the structure a bit more. Let's ask some level of government (in practice wither the feds or the government of Alberta) to fork out a paltry few million dollars fund the Phase II trial, and only start the prize competition if DCA passes Phase II.

So if it gets to Phase III, the deal is that the government will hand over a large, precommitted pile of dollars to the first drug company which can bring DCA successfully through Phase III trials to marketability.

How large a pile of dollars? Well, it'll have to be large enough to cover the costs of the Phase III trial. And since the drug company won't make any profit off marketing the stuff, it'll have to pay them something like the rate of return they would have made had they been able to patent the drug. That rate of return will have to include a risk premium because I've said that the prize will be paid to a company which brings DCA successfully through Phase III trials - if it turns out that the Phase III trials show that it doesn't, in fact, work in humans, nobody wins and any entrants swallow the trial costs as a loss, just as they would if they were working on a patentable drug and that drug didn't pan out.

Still, critics of Big Pharma are always claiming that the industry exaggerates the costs of research, and that a big part of their reported costs are really marketing costs in disguise. In this case there won't be anything much in the way of marketing costs - the generic manufacturers will pump the stuff out, if it works, and the news coverage of successful Phase III trials will be all the advertising the generic version needs.

So how big will the prize have to be? That'll be an experiment in itself - the government can keep raising the amount of the prize until it's large enough to attract entrants.

Ok, so there are still some details which would need to be fleshed out. But there are lots of researchers out there who have done theoretical work on the prize notion, and there's probably a game theorist or two who might be able to make a contribution.

So there it stands. Proponents of the prize idea should sort out the details of a workable prize. They should then start lobbying governments, not just in Canada, to conduct the experiment. If it works out as some people think it might, you get a new cancer drug. If it bombs, you at least get one real world test of the whole prizes instead of patents idea. And whatever the outcome, the questions about DCA are answered.

What more could you ask for?

From Florida, A Little Bit of Applied Health Insurance Theory

The South Florida Sun-Sentinel ran a piece not too long ago which illustrated quite nicely a bit of the standard economic model of insurance. The story actually dealt with a decision by Florida State University to require all students to have health insurance - if they don't have it through other sources they have to get it through the university. The bit of the story that relates most directly to the way we teach the economics of insurance is actually just one line, late in the piece. Even that one line's worth taking a look at, though, for what it says about the economics of health insurance.

First, though, the background. The story's from the 28 May/07 Sun-Sentinel, was written by Bill Kaczor, and was entitled:

FSU to be first public university in state requiring student health insurance

A few snips:

A broken bone, cuts and bruises weren't enough to keep a Florida State University student out of class after being hit by a car while riding his bicycle three years ago.

His $30,000 hospital bill, though, forced him to drop out because neither he nor the driver had insurance, recalled Leslie Sacher, director of the university's Thagard Student Health Center.

The student eventually worked out an arrangement to pay the bill, and he returned to school the next semester. It was the last straw, though, for health-center officials who annually process hundreds of medical withdrawals.

"We were determined that was going to be the last student who suffered in that way," Sacher said.

It took longer than she had hoped, but Florida State this fall will be the state's first public university to require health insurance, starting with freshmen and other new students.

If students lack their own coverage that meets university criteria, Florida State will provide it at a minimum annual cost of $1,449. Financial help will be available for low-income students.

Mandatory insurance is a growing trend among the nation's universities. It's nearly universal among private schools, and the number of public universities requiring it has grown from 25 percent to 35 percent in two years, figures from the American College Health Association show.

And here's the bit that catches an economist's eye:

The turning point for Mary Coburn, Florida State's vice president for student affairs, came when only one company bid on the school's voluntary health insurance offered to students. Too few students were enrolling and those who did typically had health problems, driving up the risk and cost.

Ah yes, adverse selection. The thing about insurance is that it pools risk. A whole lot of people contribute to the pool, but only those who suffer an insured-against event collect. If everyone in the pool has the same probability of being hit with a bad event, then they will all pay the same premium and the amount of money in the pool will just cover the stream of payouts that have to be made through the year. But suppose some people are more likely than others to suffer that bad event. If everyone knows who falls into the high and who into the low risk group, one solution is to charge the high risk individuals a higher premium, one based on the higher likelihood that they'll be making a claim on the mutual pool. Smokers, for example, could be charged a risk adjusted premium.

If, on the other hand, there are both high and low risk individuals in the pool but nobody at all knows who's high risk and who's low risk - perhaps we're talking about a genetic disease for which there's no known marker - then everybody would pay the same premium, calculated so that the total funds in the mutual pool would still cover the expected annual payout.

But suppose there are high and low risk people in the pool, and suppose that everyone knows who's who, and suppose that the law requires that everybody be charged the same premium, not differentiated by risk. So people who know that they're fundamentally healthy pay the same premium as those who are fundamentally unhealthy, and both groups know it.

In that case, the healthy are likely to reckon that they're not getting their money's worth out of their insurance and, if insurance is voluntary, drop out. That raises the riskiness of the remaining pool, since there will be a smaller proportion of fundamentally healthy folks in the pool, which means that the average premium will have to rise. That premium will drive more basically healthy people away, and the process continues until you wind up with a pool of fundamentally unhealthy people paying very high premiums for their insurance. the process is sometimes referred to as an insurance death spiral. This is exactly what seems to have happened with FSU's student plan in recent years (in the case of university plans, things are complicated by the basic perception of that age group that they're invulnerable, so they may well get the probabilities wrong when they do the calculations). Mandatory insurance is one approach to dealing with the issue, basically forcing the healthy to cross subsidize the unhealthy.

My own preference in this would be a mixed system. I'm willing to compromise my fundamentally libertarian view of the world and argue that it should be mandatory to carry insurance against catastrophic bills (but not comprehensive insurance; not the sort which would pay all your doctor's bills). I do think, though, that there should be a bit of risk rating associated with habits which are known to lead to higher health bills - I have no problem with smokers having to pay an extra premium.

Anyway, if you're teaching a course on the economics of insurance, the Sun-Sentinel's FSU story's a nice one to build a lecture around.

Some people think that risk rating of access to care is unfair, and argue that what the US needs is a government run health care system, under which everybody would be treated exactly the same way, but nobody could leave so no death spiral would be possible. Would that it were that easy.

Consider this story, from the UK's Sunday Times, 3 June/07, about recent decisions in Britain's National Health Service:

NHS trust to warn smokers: no surgery until you quit
Sarah-Kate Templeton, Health Correspondent

SMOKERS are to be asked to give up their habit before they are put on the waiting list for routine operations such as hip replacements and heart surgery.

National Health Service managers say smokers take more time to recover from surgery, blocking beds for longer and costing more to treat.

One primary care trust will launch a consultation on the new curbs this summer to coincide with the ban on smoking in public places to be enacted on July 1.

Rod Moore, assistant director of public health at Leicester City West Primary Care Trust, said it should become the norm for patients to stop smoking before all routine surgery.

“If people give up smoking prior to planned operations it will improve their recovery,” Moore said. “It would reduce heart and lung complications and wounds would heal faster. Our purpose is not to deny patients access to operations but to see if the outcomes can be improved.”

NHS managers want patients not to have smoked any cigarettes for a full month before surgery. But as they would be expected to take about two months to stop, operations could be delayed by up to three months.

And presumably the option exists of no quit, no operation. A couple of years ago, the BBC website carried this piece:

Three Suffolk primary care trusts have ruled patients with a body mass index (BMI) over 30 will not get operations like hip and knee replacements.

A person of average weight would have an BMI of between 18.5 and 24.9.

Dr Brian Keeble, a director of Ipswich PCT, said: "We cannot pretend that this work wasn't stimulated by pressing financial problems."

Under new guidelines surgery will not be performed unless "the patient has a body mass index below 30 and conservative means have failed to alleviate the patient's pain and disability".

Interestingly, the 30 May/07 Guardian newspaper ran a brief piece on a poll of the British public's opinions on such restrictions:

A majority of the public wants the NHS to deny obese people surgery until they lose weight but allow smokers an automatic right to treatment even if their condition stems from the use of tobacco, an ICM opinion poll reveals.
.................................................................................................

The results .................... show 66% of those questioned thought obese patients should be made to lose weight before any operation for a weight-related illness. Restrictions on treating the seriously overweight were introduced in parts of England last year. They were widely seen as an economy measure to ration care, but NHS chiefs said slimming before surgery could bring medical benefits, including reduced pressure on joints and organs.

There are similar arguments for making smokers quit before heart or lung surgery, but more than two-thirds in the poll said smokers should not be required to alter their behaviour before an NHS operation for smoking-related illness.

Interesting response on smokers. Perhaps they reckon that smokers are innocent victims of brain-washing by evil cigarette company ads. Or maybe they got a sample which had a heavy proportion of smokers in it.

According to this BBC piece, during a debate in the House of Lords about this sort of rationing:

Lord Tebbit asked whether the minister would consider it acceptable if similar requirements were applied to people whose sexual habits "make them vulnerable to particularly unpleasant sexually transmitted diseases".

Lord Hunt dismissed the question as "quite ridiculous" adding: "This is a completely different issue".

That would be an interesting essay topic to set.

Incidentally, since we've wandered a long way from Florida State University, you know how lots of folks blame fast food restaurants for the rise in measured obesity? Seems a British researcher has a different culprit in sight. From the BBC website:

Microwaves may be to blame for kick-starting the obesity epidemic, a UK scientist suggests.

Professor Jane Wardle says obesity rates started to rise soon after 1984 - around the time of the rapid spread of microwave ownership.

The mid-1980s also saw the first ready-meals appearing in shops.
..................................................................................................

Professor Wardle who is professor of clinical psychology at University College London said: "I looked at the figures showing rates of obesity in the population over many years and it seem very clear it began between 1984 and 1987.

"So then we looked at what changes were going on in the food and activity world at that time and one of the striking changes was there were differences in the speed with which we could prepare a meal as a consequence of the introduction of microwaves."

She added that food also became cheaper and ready meals began to appear on supermarket shelves.

"I'm not trying to demonise the microwave but it was emblematic of a change that took place in the 1980s in terms of the availability of food - a real change in the disincentives for eating."

And, from the same story, on that supermarket thing:

............ Professor Tim Lang, professor of food policy at City University in London argues that the introduction of the supermarket is to blame.

"Co-op introduced the supermarket retail format to Britain, heralding the late 20th century food revolution in which prices have tumbled, car use rocketed, physical activity plummeted and the NHS was born which picks up the pieces."

Unless you're an obese smoker, it seems.

Bottom line (finally): health insurance isn't as easy a thing to sort out as lots of folks, from all parts of the political spectrum, would have you believe.

According to that original Sun-Sentinel story, the Board of Governors of the Florida public university system is considering extending the requirement to the entire system. I wonder if anyone's thought of taking it one step further, and letting Floridians continue to pay into the system after they've graduated? Perhaps with tax breaks (Federal, since Florida doesn't have an income tax, or perhaps on the state's property tax) and a guaranteed renewability structure, to discourage the healthy from dropping out? It just might be worth giving some thought to.

US Health Insurance: Identifying Its Problems

From the South Florida Sun-Sentinel, an article which contains some observations which should be kept in mind by anyone thinking about how to reform the US health insurance system. Interestingly enough, it's not an article on health insurance reform - those have a nasty habit of ranging from vapid to pie-in-the-sky, especially during an election season (which, these days, means pretty much all of the time). Rather, it's an article by Carolyn Bigda, from the Chicago Tribune, talking about things to keep in mind if you're thinking of taking a job at a small company in the US.

It was published in the Sun-Sentinel on 3 June/07, headed:

Health coverage can be hard to find at smaller companies

Always nice to see an article start with an indisputable fact. Even if you do regard the "can" as a pretty broad qualifier. A few excerpts:

If you're considering taking a job with a small employer, take a hard look at the benefits: Health insurance may not be included.

A March study from the U.S. Government Accountability Office found that the number of employers offering coverage has declined, particularly among small businesses. In 2006, 60 percent of employers with between three and 199 workers provided insurance, down from 68 percent five years earlier.

It's not that small businesses are stingy. But because they have fewer employees to contribute premiums, the insurer takes on more risk, which drives up the cost for the business or workers (or both).

Then, if one employee becomes seriously ill, premiums skyrocket.

"It's not just the rate, but the unpredictability of the rate," said Bill Lindsay, president of the Denver Benefits Group of Lockton Cos. LLC, an insurance broker. "You don't see this volatility in prices in any other part of an employer's operation, except maybe fuel prices today."

Health-care costs have jumped 20 percent over the last two years for Avon Lake, Ohio-based Phoenix Products Inc., which has 29 full-time employees. The company has had to trim benefits, such as discontinuing a prescription drug plan.

Look at that bit about small employers having fewer employees to contribute to the plan. That's the key to a whole lot of US health insurance issues.

Much US health insurance is small pool insurance - when a company has a small workforce, its health insurance plan can't spread risk widely. Should one member of the plan suffer a costly illness, it has a much greater impact on the plan than it would were it a large pool plan - in a large pool plan the costs of that one individual's major illness would be spread over a much larger pool of risk-sharers.

In addition, small pool insurance is a major reason the US has such high administrative costs - the costs of administering an insurance plan are pretty close to fixed, so they're very much the same whether it's a large or a small pool plan (they'll rise with plan membership, but there's still a heavy fixed cost element). But that's in total. Take very similar total administrative costs and divide them over, in the one case a large workforce and in the other a small workforce, and they'll be much higher per capita for the small workforce. they'll also, for the same reason, be a higher share of premium revenues. If you look at the health insurance plans of very large employers in the US, you'll find that they're pretty similar, as a percentage of revenue, to the costs of government run plans in other countries. There are economies of scale in administering an insurance plan, and small pool plans can't take advantage of them. (There are, by the way, limits to those economies of scale, so going beyond a few thousand members - up to, say, the entire population of a country - doesn't buy you much if anything more in the way of cost savings.)

Employer-based health insurance had its pluses, especially in the early days of health coverage. Nowadays, though, the problems associated with it are pretty major, and they arise not, as some folks would have you believe, from the for-profit nature of so much insurance but rather from the fact that so much of it is small pool. (And Canadians should keep in mind the fact that most of us get our dental and, very importantly, our drug insurance through the same kind of employer-based mechanism.)

Anyone who wants to reform the US health insurance system should start by tackling the employer-based nature of the system. Move the insurance plans out of the individual firm - simply pool a whole lot of employer-based plans into a handful of large pool plans. There's no reason for this to affect employees at all, at least at first. Employers can go on making premium payments on behalf of their employees, they just make those payments to an outside, large pool plan. Since there would probably be several such pooled plans in any state, the next step would be to let employees pick which plan they'd want their payments to be made to. Ideally, each of the large pool plans would be permitted to offer a catastrophic health expenditure plan: in effect a plan with a high deductible, so that members would pay their small bills out of pocket. Membership in a plan wouldn't be tied to staying with a single employer, so the employee could keep the same coverage no matter how often he changed jobs, and there would be lifetime, guaranteed renewable coverage, meaning that the individual and the insurer would both lock into continuing the insurance - that covers members who develop chronic illnesses (the pre-existing condition problem).

Even the self-employed, or unemployed would be allowed to buy into these plans, with low income individuals getting a government subsidy for their premiums, so that the individual insurance market (boy, talk about small pool) would be absorbed into these pools.

And finally, that handful of preventive procedures which have actually been shown to work (pap smears, mammograms) would be absorbed into state public health programs and funded out of general tax revenues, the way flu shots often are, so there'd be no concern about people deciding not to have these tests for financial reasons.

The US health insurance system is fixable. As with any cure, the first step is in getting the diagnosis right. Small pool insurance is a major element the US health insurance system's problems, (and not because it's for-profit insurance, despite what you're always hearing) but it's an element which really wouldn't be all that difficult to fix.

Pigou With A Twist

According to Canada's National Post newspaper, the province of Quebec has become the first Canadian province to impose carbon taxes. But, according to this story the plan has some slightly unusual details.

The story, from the 7 June/07 Post, is by Kevin Dougherty and is headed:

Quebec the first to announce carbon tax

And at first glance, all seems well:

Quebec will have the country's first designated "carbon tax" to help fight global warming, it was announced yesterday.
.................................................................................................

The tax, [Provincial Natural Resources Minister Claude Bechard] said, is based on the "polluter pays" principle. "That is not negotiable," the Minister said.

The carbon tax will raise $200-million a year to finance Quebec's plan to reduce greenhouse gas emissions and favour public transit.

Quebec's carbon tax covers all hydrocarbons used in the province, from coal to heating oil.

The amount of the carbon tax varies according to the amount of carbon dioxide each fuel produces.

For gasoline, the tax is 0.8 cents a litre, the charge for diesel is 0.9 cents, for light heating oil 0.96 cents, heavy heating oil one cent a litre, coke used in steel making 1.3 cents a litre, coal $8 a tonne and propane 0.5 cents a litre.

The twist's in that non-negotiable polluter pays bit:

Provincial Natural Resources Minister Claude Bechard, who announced yesterday that a 0.8-cent-a-litre carbon tax will come into force on Oct. 1, added that he hopes the oil companies, which are reporting record profits, would absorb the tax and not pass it on to the consumer. Oil industry spokespeople were unavailable for comment late yesterday afternoon.
.................................................................................................

"We hope at 0.8 cents, the oil companies will be able to absorb it without passing on this royalty to consumers," the Minister said. "Especially when you realize that refinery profit margins have gone in the last three, four months from 8 cents a litre to about 19, 20, 22 cents a litre."

Asked why he thinks the oil companies will absorb the carbon tax, Mr. Bechard said, "Well, we count on the goodwill of the gas companies." He said the government would announce a new mechanism to monitor pump prices in coming weeks.

Mr. Bechard has also threatened to impose a ceiling price on gasoline. Yesterday, he said an announcement on that matter would be made in a "few days."

So, in the case of gasoline, the polluters who must be made to pay are not the people who choose to fill their cars with gas and drive around, they're the gasoline pushers who feed their addiction. But notice that this isn't a pure profits tax, so it will be distortionary. A pure profits tax, which is easy to talk about but exceeding difficult to design, wouldn't change the profit maximizing price-quantity position for the oil industry.

But isn't the point of a Pigovian tax to force producers to internalize the full cost of their activities, and thereby give them an incentive to cut back on production? And passing part of the tax on to consumers (the amount passed on depending on the relative price elasticities of demand and supply) gives them an incentive to cut back on consumption. So isn't the whole idea to reduce consumption of gas?

Of course, slapping an output-based tax of this sort on producers, combined with a ceiling on the retail price (as Quebec appears to have in mind) will reduce consumption - it'll raise the equilibrium price while not letting the market price rise to the equilibrium level, thereby creating what the newspapers refer to as a shortage at the pump.

The CBC's website has a bit more detail:

Natural Resources Minister Claude Béchard said Wednesday he hopes the petroleum industry will pay the tax without passing on the cost to drivers when they fill up their cars at the pump.

"We all have a responsibility. Every Quebecer has a responsibility. It's important for every Quebecer. So I hope that all those companies will have the same sincerity that we have, that Quebecers have," Béchard said.

And who could say fairer than that?

Meantime, next door, in the province of New Brunswick, we have another example of Canadian environmental policymaking. According to one CBC report:

The New Brunswick government has released a plan that it hopes will help to cut the province's greenhouse gas emissions to 1990 levels by 2012.

The plan includes more green power generation, a greater use of public transit and programs to help residents become more energy efficient.

Premier Shawn Graham says the five-year climate change plan will require everyone in the province to voluntarily do their part.

Require everyone to voluntarily do their part. Does that come under the heading of "unclear on the concept"? And then there's this:

Graham says emissions can be cut by 10 per cent below 1990 levels by 2020, but admits that doesn't take into account a proposed new oil refinery in Saint John.

So we can get our emissions down below their 1990 level as long as we don't count some of them. Apparently the rules of government financial accounting have been extended to environmental accounting. One more bit, from another CBC story:

[New Brunswick Environment Minister Roland Haché] says mandatory limits and penalties aren't necessary to hit the target.

"We didn't put anything in there that was wishful thinking or anything, and we feel that those targets are very realistic and we will reach those targets."

The province says it's up to the federal government to create and enforce mandatory emissions limits, and the provincial plan will work regardless.

So there's nothing mandatory about these voluntary targets which the premier says you're required to hit. Mandatory limits, being nasty things which might annoy voters, are up to the feds to impose.

So what do you think? Is it true that you get the government that you deserve?

Every Little Bit of Evidence Helps

It's always nice to be able to back up an assertion drawn from theory with evidence, even if the sample size is limited (to 1, perhaps). The Albertan Dr. Al, who blogs at What's Wrong With Health Care provides some anecdotal support for something that I regularly tell my health economics class is an implication of economic theory, but for which there's no good Canadian evidence. Al's post was inspired by an article in the March 25/07 Calgary Herald (no link, sorry) on the role of Physician Assistants in the US health care system. In my lectures, I usually talk about Nurse Practitioners because they are the non-physician providers most often talked bout in the Canadian policy context, but despite the fact that PAs and NPs follow different training programs, by the time they've completed their training their potential scopes of practice are pretty similar.

I preface the lecture by pointing to news articles which report that some provincial government or other (pick one, they all do it) has just announced that it has plans for making greater use of NPs in that province's health care system. Most recently, I've been using articles on Ontario's latest version of community health centres, but that's just a new variation on an old theme (apart from the fact that the way they plan to pay doctors looks pretty much like the way Finland did in their family doctor health clinic program - they had to change the payment mechanism after a few years because the one they were using led to significant increases in wait times for access to care).

The announcement can be counted on to make a big thing about this health care innovation, and can also be counted on not to mention that the Americans have been using NPs and PAs for decades. (Sounds much better to say that we're doing something new than to say that we're trailing by a decade or so.) I point this out to my students, and tell them that some of the earliest experiments with NPs were conducted in Canada in the 1970s, and ask why it is that we lag so badly at introducing them into the health care system.

So here's a bit from Al's story on his own experience with PAs, back in the 1970s:

It was about that time that the Calgary medical school had, as their Dean of Medicine’ a young, bright acquaintance of mine, Dr. L. McLeod. Another young family physician and I met with him and proposed a two to three year program at the Calgary medical school for physician assistants. I had a nurse working for me at the time (Pat W), that I had taught to put on and remove casts, do minor suturing, do prenatal visits, well baby check-ups and assist my office nurse in arranging consultations, investigations, etc. Pat had a group of people in certain chronic disease areas (obesity, diabetes, etc) that she worked with regularly and councelled, and made house calls on palliative care patients. This expanded the number of patients that I was able to accommodate in my practice by approximately thirty percent and I feel, improved the care that I was able to give overall in my practice. Unfortunately, our universal healthcare system provided no means for charging for her services (I could only charge the system if I, as the attending physician, gave the service personally), so when our great leader, Mr. P. E. Trudeau, brought in wage and price controls and “extra billing” was outlawed, I had to discontinue the program because it represented an overhead that I was unable to recapture.
With this practical experience, I felt strongly that we should move in the direction of physician assistants provincially, with a billing system specifically for their services. Unfortunately, Dr. McLeod felt there would be too much conflict with the department of nursing and other jurisdictional headaches, and did not pursue establishing a school for Physician Assistants.

(There's a lot more to the post - well worth going across and reading the whole thing.)

A quibble - it's the Canada Health Act that's the legislative source of the problem: it's what blocks extra-billing, and, more importantly, virtually prohibits anyone but the government paying for a Medicare-insured service. I say more importantly because that's what prevents an MD billing the patient directly for the services of a PA or NP who was employed by the practice.

There's been a lot of talk over the years about making more use of Non-Physician Providers (NPPs) like NPs and PAs in Canada, including lots of talk about how it would reduce our health care costs. The Barer-Stoddart report's view of a restructured Canadian health care system included cutting back on the number of doctors in Canada and introducing NPs as providers of primary care. As is typical of government run systems, we've produced endless talk and reports while the Americans have gone ahead and made NPPs a key part of their primary care system.

What those Calgary doctors were doing back in the 1970s was not just adding bodies to his practice, he was also invoking the fundamental economic principle of comparative advantage or, if you prefer, specialization and division of labour. That's the principle that's at the core of the potential value of NPPs to our health care system and their actual value to the US system. A lot of people will tell you that NPPs will provide care more cheaply than GPs because they cost less to train than GPs do. Rubbish. What will happen is that their incomes will rise to the point where the ratio of their incomes to GP incomes equals the ratio of their productivity to GP productivity in the production of care. What that boils down to is to say that when you compare like with like - looking at what NPs and GPs earn for supplying exactly the same number and mix of services - NPs are no cheaper than GPs. Their annual incomes are less, but that's because they supply fewer services and don't supply the more complicated ones which GPs do and for which GPs earn higher fees. That's already happened in the US, and American supporters of NP and PA programs have pretty much dropped the argument that they're a cheaper source of supply.

Getting back to Dr. Al and comparative advantage, what happens when you add NPPs to a GPs practice is not just that you've increased the supply of health care labour, it's that you've freed the GP to specialize in looking after more complicated cases, leaving the simpler ones to the NPP. this is a more efficient use of the GPs time than having him look after every case that walks into the office, regardless of its level of complexity. Costs will rise, because the NPPs have to be paid, but the practice's productivity will rise by even more, because the different types of labour (GP and NPP) will be able to specialize in the direction of their comparative advantage.

What those doctors were doing back Calgary in the 1970s was engaging in what has come to be labelled entrepreneurial medicine, and in the Canadian context, entrepreneurial medicine is evil because it might give someone an opportunity to make a profit. In the Canadian health care context, making a profit is defined as taking advantage of someone else's misfortune for your own selfish gain, it's not defined as being rewarded for figuring out what someone needs and coming up with a way to make it available to them. So the system prevents doctors from experimenting with NPPs, or from expanding the range of services which they provide in their offices by investing (another dirty word) in private minor surgical clinics, for the simple reason that such actions might yield revenue in excess of their cost.

So we have to wait - decades, in the case of NPPs - for government to commission reports and agonize about the political ramifications (as in "we can't put one of those clinics in Moncton if we're not also putting one in Saint John") and maybe, eventually, to move, so long as the budgetary implications don't look too nasty.

But that's the Canadian way, isn't it.

Once Again, We See People Responding To Incentives

The New York Times ran an article on April 17/07 ($$$ subscription required) on the costs of treatments for drug addicts under New York's Medicaid program. It contains a couple of nice examples of optimal responses to incentives, or at least, individually optimal responses conditional on certain broad objectives.

The article's by Richard Perez-Pena and is headed:

Revolving Door for Addicts Adds to Medicaid Cost

Its focus is on the costs of drug and alcohol treatment in New York state:

Through its Medicaid program, New York spends far more than other states on drug and alcohol treatment, including more than $300 million a year paid to hospitals for more than 30,000 detox patients. One reason for the high cost is that $50 million is spent just on the 500 most expensive patients, at a cost of about $100,000 a person. These patients check in and out of detox wards, on average, more than a dozen times a year — a practice that experts say would not be tolerated in most states.

In the state’s 2004 fiscal year, one patient was admitted to such units 26 times at 17 different hospitals around New York City, spending a total of 204 nights, Medicaid records show. In fiscal year 2005, there was one patient who spent 279 nights in detox wards, at a cost of about $300,000.
.................................................................................................
In other states, most addicts who go through detox programs do so on an outpatient basis, while in New York the vast majority are inpatients. Medicaid rules in New York also encourage hospitals to provide the most expensive kind of inpatient detoxification, though it is often not medically necessary, while many other states favor a less expensive form of inpatient treatment.

The story discusses various factors behind New York's heavy use of inpatient treatment, including the rate of homelessness, and then we get to the first of the examples of incentives at work:

The most intensive form of treatment, “medically managed” withdrawal takes place in a hospital, usually involves some sedation, and requires a great deal of care by doctors and nurses. The next level, “medically supervised withdrawal,” can be done in a hospital, or sometimes on an outpatient basis, and requires less medical intervention and less staff.

In New York, Medicaid pays an average of more than $100 a day for outpatient medically supervised withdrawal, and close to $400 a day for the inpatient version.

But it pays more than $1,300 a day for medically managed detox — and state officials estimate that more than 40 percent of that is profit for the hospitals. Hospital executives say the margin is not that high, but they concede that the most expensive form of detoxification is a significant money-maker.

As a result, many hospitals offer that program, but not the cheaper ones. By law, hospitals cannot turn away emergency patients, and drug or alcohol withdrawal is considered an emergency. So about 80 percent of the detox patients handled by hospitals in New York are treated at the most expensive level — often because it is the only one available.
...................................................................................................
Federal officials say they do not keep state-by-state Medicaid records, but experts and state officials say it is clear that New York spends far more on drug treatment than any other state, because other states mostly provide outpatient treatment. Figures compiled by the Department of Health and Human Services support that claim, showing that New York has more hospital admissions for drug or alcohol abuse — whether paid by Medicaid or someone else — than California, Texas and Florida combined.
.................................................................................................

Spitzer administration officials say the state needs to pay less for the top level of care, and possibly pay more for the others, to spur the development of those services. That fits with the governor’s plan to review what Medicaid pays for all medical services, with an eye to encouraging less expensive forms of care.

But those officials also know that George E. Pataki tried twice as governor to change the detox payment system. Each time, the hospital industry, which has been losing money over all, persuaded the Legislature to protect one of its few sources of profit.

So there we have it, right? It's that evil for-profit American health care system.

The trouble is, this is New York state that we're talking about. While it doesn't outright ban for-profit hospitals, its rules about hospital ownership are so tight that corporate ownership is virtually, if not entirely, absent. So these are non-profit hospitals which are worried about losing one of their remaining sources of profits.

Non-profit, remember, does not mean not making a profit. It's a tax status, meaning that the IRS has agreed not to tax whatever profits you make. In the case of non-profit hospitals in the US, favourable tax status is granted in exchange for the hospital's providing care to the poor and uninsured. There have been questions raised recently about how many non-profits are really for-profits in disguise, doing the bare minimum they can get away with and still keep the IRS off their backs, but that's not relevant here - the same economic issues arise in the case of hospitals which live up to the spirit of the tax deal in full.

The big issue is what's termed uncompensated care - care which hospitals provide to people who don't have insurance. There's really very little genuinely uncompensated care - what there is is a lot of care which is paid for through an absurd system of cross subsidization. Non-profit hospitals are supposed to cover the costs of this care through the profits they make on care provided to people who have coverage which will pay a price which exceeds the cost of the care they receive. And that means that, in order for a hospital to treat uncompensated care patients, it has to be making a serious profit on its other patients. Including ones whose bills are paid through government programs. Drug addicts on Medicaid, for example.

Consider these patients. The hospitals could shift them to less profitable forms of care, but if they did that, they'd lose the funds they use to cover some of the costs of the uncompensated care they provide. The treatment these addicts are receiving may be unnecessarily resource intensive, but there's nothing in the article to suggest that it's harming them, and they do get a few (or in some cases a lot of) nights room and board at the state's taxpayers' expense. So shifting the addicts to less intensive, and less profitable, forms of treatment would not significantly improve the addicts' lot and would reduce the hospitals' capacity to supply uncompensated care.

So a fully altruistic non-profit hospital has every incentive to squeeze every cent it can out of the Medicaid fee schedule.

But there's more - a second incentive at work, this time on the part of the addicts themselves. From the NYT piece:

The system suits the most frequent patients — most of them homeless, mentally ill, or both — who see the programs as a source of shelter and food. And the most expensive treatment, which usually involves some sedation, can reduce the discomfort of withdrawal better than other methods.

Some drug users, especially those on opiates, also set out to clean their systems so they can reduce the dose needed to get high, according to addicts and those who treat them. For a homeless addict, the cost of each dose is a major concern.

Focus on the second of those paragraphs. What it means is that over time, the optimal consumption pattern for these guys is saw-toothed. For whatever reason, they choose to start on a drug. Over time, they become habituated, meaning that they have to consume more and more of the stuff in order to get the same level of effect from it as they got in the past. As their consumption level increases, so does the size of the chunk of their budget which they have to spend just to get the same level of high as before. When their spending hits a critical level, they check into a New York hospital for inpatient treatment, paid for by the state's Medicaid plan. That takes their level of addiction, and of habituation, back down to zero (or to some low but non-zero level) and they can start again at a lower level of consumption, and let the pattern repeat. The price to them of the treatment is zero, and they get a few nights shelter and food. It would actually be rational, conditional on their having decided to take up the consumption of an addictive commodity, and on a decision not to go cold turkey after the first treatment (see why economists make so many "on the one hand, on the other hand" statements?) for them to plan in advance to do this. Rational addiction behaviour.

So the hospitals profit by supplying only the most expensive form of treatment (but do so in order to put their profits to good use) while on the demand side, addicts benefit from the availability of the treatment (and are not harmed by being treated at a much more resource intensive level than they really need). And one little part of New York's Medicaid costs go through the ceiling.

But if the state government cuts payments on these profitable services, the hospitals' capacity to supply uncompensated care is damaged.

(Incidentally, note again that these are non-profit hospitals, and that, since we're talking about Medicaid patients, we're looking at what is essentially a single payer system.)

The American way of covering the health care costs of the uninsured is hopelessly inefficient. It's not absurd in its origin - it comes out of the American belief in personal, rather than government, charity. If your neighbour is in trouble, you don't wait for the government to help him, you (and your community) do it yourself. An eminently sensible form of social safety net in a frontier land, where any level of government other than the local one is so far away as to be pretty much useless when local emergencies arise. The UN regularly castigates the US because its government doesn't give enough to international relief agencies when disasters occur internationally, but that always ignores the massive amount of aid which Americans give privately.

So the system of providing care for the uninsured out of local hospital revenues has a respectable pedigree - historically, your local community hospital would price discriminate, charging the wealthier local residents more than it really cost to provide the care they used in order to be able to subsidize care for the poorer local residents.

That's a system which, however well it worked in the past, can't go on. It's time to rationalize the way uncompensated care is paid for, perhaps by pooling the mess of funds currently spent on it into some kind of an insurance pool. That might well involve eliminating the hospitals' un-taxed status and using the tax revenues to fund the pool. It'll require measures to prevent free-riding, perhaps by following the pattern set by Romneycare in Mass.: require everyone to show, on their income tax, proof of health insurance. Anyone who doesn't have private care would be required to pay into the state pool, with premiums being subsidized for low income people.

It might not cut US health care costs, but I'm inclined to argue that the first step towards rationalizing the US health insurance system is to figure just exactly what channels the money which is currently being spent is flowing through, taking a good firm hold on those channels and straightening them out.

But That's Not What We Wanted To Have Happen

A bit of evidence, this time from the Wall Street Journal, on the matter of whether cigarette smokers respond to tax-driven price increases. The answer is that they do, but often not in the way the tax increasers wanted them to.

The story ($$$ subscription required) is by Amir Efrati, from the March 2/07 Wall Street Journal and is headed

Cigarette-Tax Disparities Are a Boon for Border Towns

Basically, it's a story about cross elasticities of demand, although in this case we're not looking at how a change in the price of good A affects the demand for good B, but rather how a change in the price of good A at place 1 affects the demand for good A at place 2. In this case, places 1 and 2 are on opposite sides of state borders:

LARCHWOOD, Iowa -- Since Jan. 1, this town of 800 in the northwest corner of Iowa has been deluged by thousands of visitors, and they haven't come to see the sights. It's for the smokes.

On the first day of 2007, the cigarette tax in the neighboring state of South Dakota jumped from 53 cents a pack to $1.53, making a carton cost at least $11.70 more than a carton in Iowa. Immediately, hordes of South Dakotans crossed the border into Iowa, many of them heading for Larchwood, a 20-mile drive from Sioux Falls, South Dakota's largest city.

........................ in the month of January, Larchwood Mini Mart sold $160,000 of cigarettes, triple the amount sold all of last year. "It's a gold rush," says owner Larry Zuraff.

The Iowa/South Dakota example's not an isolated one:

Today, other recent boomtowns attributed to cigarette-tax hikes include Sunland Park, N.M., population 16,000, where the difference from neighboring El Paso, Texas, as of Jan. 1, is about $5 a carton. Texans have descended on Sunland Park's only tobacco shop, which has seen a fivefold increase in cigarette sales, up to 4,500 packs a week. That helps the town due to a two-cent-a-pack municipal tax on top of a 0.7% municipal sales tax.

Hudson, Wis., a town of about 11,000 located 20 minutes from downtown St. Paul, Minn., has seen an influx of smokers from Minnesota since mid-2005, when that state's cigarette tax jumped to more than $7 a carton higher than Wisconsin's. The visitors from the Twin Cities, where smoking at restaurants is banned, are taking advantage of Hudson's smoking-friendly haunts.

Big winners are also located on the low-tax side of metropolitan areas situated on borders, such as Bellevue, Ky., ($3 tax per carton), across the Ohio River from Cincinnati ($12.50 tax per carton); and northern Delaware ($5.50 tax per carton), just south of Philadelphia ($13.50 tax per carton).

If you're estimating a demand function for cigarettes on state level US data and you don't take account of border-crossing effects, you'll wind up with badly biased estimates of the own price elasticity of demand. Think about the South Dakota case: if you don't factor in the effect of Larchwood, Iowa, being 20 miles from Sioux Falls, you'd observe an increase in the cigarette tax in South Dakota and a large drop in sales in South Dakota, and you'd conclude that the tax increase had been a pretty effective public health measure. In fact:

State governments often tout cigarette taxes as a public-health initiative -- a way to force some people to quit and to reap revenue from those whose habit leads to smoking-related illnesses. Most states raised cigarette taxes as a way to balance their budgets during hard times.

And while they might have been touting the public health benefits, what odds that the prime objective was revenue raising? After all, if you reckon that the demand for cigarettes is pretty price inelastic, they become a very tempting target for the tax man.

In Having It So Good, his history of Britain in the 1950s, Peter Hennessy reports that, when firm evidence first appeared in the UK linking smoking to lung cancer, the government of the day decided not to take any serious anti-smoking measures in part because tobacco taxes were funding most of the welfare state. But back to the WSJ piece:

States raised a collective $14 billion from cigarette taxes for the fiscal year ended last June 30, but studies have pegged the percentage of revenue lost due to smuggling, Internet sales or other forms of tax evasion at anywhere from less than 10% to more than 20%.

Southeastern states have always had the lowest cigarette taxes, which has led to long-distance, organized smuggling from North Carolina ($3 tax per carton), for example, to states such as New Jersey ($25.75), where they are sold to retailers. The black market also includes counterfeit cigarettes from countries such as China.

Indian tribes, which aren't required to charge state levies, have been major beneficiaries of the tax increases. For years they have enticed nontribal members to buy from stores on their land or on the Internet, depriving states of revenue.

Most states have laws prohibiting residents from bringing in more than a stipulated amount of cigarettes or liquor on which the home state's tax hasn't been paid. In South Dakota, bringing back even one pack from Iowa for personal use is technically illegal, though it's hardly enforceable. On the west side of the state, smokers are making the 120-mile round trip from Rapid City to Beulah, Wyo., where cigarettes are at least $8.70-per-carton cheaper.

At the South Dakota Department of Revenue and Regulation, Michael Kenyon, director of property and special taxes, says the state is still assessing the problem. And while South Dakotans assess, Iowans prosper.

The border crossing effect is well known in the health economics literature. It might perhaps have been worth South Dakota's assessing it before that state raised its taxes. But all is not lost, at least from South Dakota's point of view:

At Larchwood's only grocery store, overall sales are up as much as $1,500 a week. Mayor Leonard Vanden Bosch says Larchwood could take in up to $15,000 in extra sales-tax revenue this year.

But Larchwood's luck may turn. The Iowa legislature could soon vote to raise the cigarette tax by up to $10 a carton, to $13.60, which would almost negate any savings for South Dakotans.

OK, any bets on whether Iowa's used the right elasticities in calculating the revenue impact of a sales tax increase?

Florida Hurricane Insurance Again

Florida Governor Charlie Crist was in DC in February trying to get support for a national disaster insurance fund - according to Insurance News Net:

The idea behind the proposed national fund for natural disasters is that policyholders all over the nation would share the risk should a major disaster occur anywhere part of the country: a major earthquake in California, a hurricane smashing into Texas, a twister in Kansas or flooding along the Mississippi River.
...................................................................................................
The CAT fund would be modeled after the Florida Hurricane Catastrophe Fund where annual contributions by insurers will be used to pay claims when total insured damage exceeds more than $4.5 billion.

The governor, as is the way of politicians, expressed optimism that such a fund could be created.

According to the South Florida Sun-Sentinel in a story dated 27 Feb/07:

Florida's Republican governor said he was "encouraged" after personally lobbying House Speaker Nancy Pelosi, D-Calif., and Senate Majority Leader Harry Reid, D-Nev.

"The momentum truly is building," said Crist, who was in Washington for the National Governors Association meeting, where state leaders discussed topics ranging from healthcare to energy. At that meeting, Crist drew support from a group of Southern governors, and he said he also has enlisted California Gov. Arnold Schwarzenegger's backing for a national insurance fund.

It undoubtedly is building among states which would expect to draw on the fund on a frequent basis, but let's not forget that federal disaster relief is part of the reason Californians seem to keep building houses in mudslide-prone areas.

Reid and Pelosi heard Crist out on Monday, but they didn't make any promises.

Reid said the proposal was "something I'm happy to take a look at." Pelosi said she was "interested in hearing the governor's views."

Is it just me, or does that have the ring of "don't call us, we'll call you" about it? That InsuranceNewsNet piece (which is dated 5 March /07) says this:

Senate Banking Committee Chairman Chris Dodd, D-Conn, is planning to schedule a hearing to discuss the development of a national catastrophe fund, referring to the matter as extremely important one and deserving of his committee's consideration.

This came as welcomed news to Senator Mel Martinez, R-Fla., who earlier voiced his pessimism on the chances of the bill’s success. Speaking before Dodd’s committee Martinez expressed fears over a looming insurance crisis affecting Florida families and businesses.

“The economic distress wrought by disasters affects us all…we’ve got to find a better way to spread and finance risk,” said Martinez. “I support the concept of a national catastrophe fund in order to stabilize and strengthen the insurance market and encourage proper disaster mitigation.”

Martinez added that the time has come for Congress to step up and move the CAT Fund legislation forward. “It's the right thing to do,” said Martinez in an official statement following Dodd’s announcement.
...................................................................................................
While the plan received a lot of support from disaster-prone states, Martinez and members of Florida's congressional delegation feared resistance coming from Montana, Colorado and other states where natural catastrophes are not key concerns.

Republican Congressman from Orlando and chairman of House Financial Services Committee Tom Feeney said he didn’t think persuasive speeches would be enough to convince the senators from Montana or Colorado to tax their constituents to help pay for the hurricane risk in Florida. Previous efforts to create a national CAT Fund have also been thwarted by insurance companies operating in areas that aren't frequently hit by catastrophic disasters.

In a story dated 13 March/07 the South Florida Sun-Sentinel tells us:

Eager to help their constituents overcome the exploding costs of homeowner insurance, South Florida members are nudging Congress toward creation of a national catastrophic insurance fund to limit liability and keep premiums affordable.

Newcomer Reps. Tim Mahoney and Ron Klein will deliver a progress report on the emerging insurance plan later this week, backed by House leaders who have shown renewed interest in helping Florida and other high-risk states absorb the cost of disaster. The legislation taking shape would allow states to opt in or out of a national insurance pool and pay premiums based on their levels of risk

Uncle Sam, in effect, would become the financial backstop for claims that reach catastrophic proportions beyond a designated amount. If insurance companies knew their losses were capped, they would be able to charge lower premiums. Floridians would pay far more than homeowners in Iowa, but a limit on liabilities faced by insurance companies would keep consumer costs relatively low.

Unfortunately, these hon. gents reveal a typical legislator's attitude towards insurance companies:

"But we must mandate that the cost savings are passed down to the consumer. This can't be a windfall for the insurance companies," said Rep. Robert Wexler, D-Boca Raton, who, along with Mahoney, D- Palm Beach Gardens, and Klein, is pushing for the national catastrophic fund.

Not to worry, at least so long as the market is competitive. However, when you see the steps disaster prone states are taking to drive insurers away from their markets (they don't put it that way when they're passing laws, but that's what it amounts to) you become a tad less sanguine - less competitive markets mean greater profits for the remaining suppliers.

Property owners in South Florida, eager for any sign of relief, say they could use some help from the federal government by sharing their risk with the rest of the nation.

"It sounds like an excellent idea if they can make it work," said Nettie Ershowsky, a condominium owner in Coconut Creek, whose premium almost doubled after hurricanes in 2004 and 2005. A vast nationwide pool of homeowners "would alleviate the burden on just the few," she said.

Some Floridians acknowledge the need to pay more to live in risky areas but resent sending premiums to companies for years only to be dropped from coverage or face sudden high costs after being struck by a hurricane. "I don't feel sorry for people who build those big monstrosities right on the water. The insurance alone has got to eat them alive," said Blanche Delman, of Delray Beach, who is helping her neighbors negotiate the insurance market. "But there are people who cannot afford to live here anymore, and children can't afford to help their parents. We should have a nationwide type of thing."

I'm inclined to share Ms. Delman's attitude, especially with regard to certain new beach-front highrises. My bet, which I've expressed before, is that originally the costs of hurricane damage were capitalized into the price of housing in South Florida - either through lower housing prices or through absorbing the cost of putting up a building which could withstand a major storm. And I'd be interested in a study on whether the new beach-front highrises had the cost of acquiring heavy hurricane insurance capitalized into their prices.

Prodded by Florida Sens. Bill Nelson, a Democrat, and Mel Martinez, a Republican, the Senate Banking Committee has scheduled a hearing for April 11 on the high costs of homeowners' insurance in disaster-prone areas.

The South Floridians say they want to take their time to devise a workable plan likely to pass Congress. Along the way, they want to show their constituents they are making progress.

"I don't think we can face the voters again without having proposed a national catastrophic property insurance fund," Wexler said.

Frankly, I think Florida's going to have to go it alone. And in that regard, RiskProf reminds us that

.............the Chicago Mercantile Exchange will be trading CME-Carvill Hurricane futures and options. The benefit is that anyone could buy a contract and payment is based on objective triggers (no wind versus water issue). However, the cost is that there is basis risk, i.e. the damage one experience may not be enough (in conjunction with other actual damages) to trigger the payments under the contract as damages are based upon an overall index.

Sate funds could buy contracts to spread the risk away from the taxpayers.

I believe this is the second foray into these types of contracts. After Hurricane Andrew I think the CBOT had a similar contract but it died from lack of interest. I suspect that people are slightly more interested today.

Perhaps a better use of state funds than subsidizing baseball stadiums? Here's a bit more on those Hurricane futures, from InvestmentNews:

The Chicago Mercantile Exchange today began selling hurricane futures contracts and options on futures contracts.

“This is an example of the exchange listening to customer needs,” said John Harangody, director of commodity products for the exchange. “Insurers approached us to look at this area of risk so it could be better managed.”

After Hurricanes Katrina and Rita caused about $79 billion in damages in 2005, it became clear that insurers have limits on insuring all claims. Hurricane contracts represent one way for insurers to transfer their risk to the capital markets, Mr. Harangody said.

For instance, an insurer with $100 million in exposure might usually sell the last $80 million to other firms to take away part of that risk. Using hurricane contracts, insurers can protect themselves and give non-traditional participants a way to put their capital to work in these markets.
....................................................................................................
The Chicago Merc contracts will cover five areas: the Gulf Coast, Florida, the southern Atlantic Coast, the northern Atlantic Coast and the eastern United States. The hurricane season in those areas lasts from June 1 through Nov. 30.

London-based Carvill Group, a reinsurance firm, will calculate the underlying index, which will be known as the CME-Carvill Hurricane Index.

Using maximum wind velocity and the radius of each hurricane, Carvill will calculate the damage potential of the hurricane, Mr. Harangody said. Once the hurricane hits land, the contract for that area expires, and the transaction will be settled in two business days.

Carvill created an index that is easy to calculate and based on publicly verifiable data, said Steve Smith, senior vice president of ReAdvisory, the analytical arm of Carvill. It was a challenge to create an index that can be calculated and settled within hours of an event, but that was most important to the trading community, he said.

The market opens with contracts available on Hurricane 1, Hurricane 2 and Hurricane 3 for the year. As these first three storms either make landfall or dissipate, the exchange will list subsequent contracts.

Forecasters at Colorado State University in Fort Collins are predicting seven hurricanes for the 2007 season, compared with an average of 5.9 hurricanes between 1950 and 2000. There were 14 hurricanes during 2005 and five last year.

As a hurricane gets closer to land, and as more is known about the effect it is expected to have, the price of the contracts is expected to rise or fall along with the worry about its impact, Mr. Harangody said.

And, from the same story, a note of caution:

Investing in weather is a growing business. The notional value of weather-risk-management contracts between April 2005 and March 2006 increased nearly fivefold to $45.2 billion, from $9.7 billion for the comparable period a year earlier, according to the Weather Risk Management Association in Washington.

Trading of weather products on the CME also has undergone a boom since weather derivatives were launched in 1999. Last year, the CME’s weather derivatives had a notional value of $22 billion, according to the exchange. Some say the hurricane contracts aren’t particularly novel and may run into rough seas in the marketplace.

“At least two or three carriers have been offering wind derivatives for two years now for operations in the Gulf and onshore within a described radius of defined wind speeds,” said Marshall Nadel, senior vice president of the Aon Natural Resources group, a unit of Chicago-based Aon Corp., which provides risk management services.

There haven’t been many takers, because of their expense, he said.

Alternatively, the state could issue its own Catastrophe Bonds. Oversimplifying, CAT bonds are ones on which the state guarantees that it will default, when a certain degree of damage is exceeded.

Basically, the state issue bonds with the proviso that if hurricane damage reaches some certain minimum level it won't repay what its borrowed. which raises the obvious question of why anybody would buy them - the answer is that as high risk bonds they'd carry a high interest rate, so you'd basically get back all of your investment plus a hefty interest rate or you'd get nothing back, and the interest rate on the bonds would have to rise until people were willing to buy them.

The state would issue the bonds, collect the principal and put it away in nice safe assets - federal government bonds, perhaps. The interest which the state earned from having put the CAT bond funds into safe assets would pay part of the state's cost of borrowing, the rest would come from something like an increase in the state sales tax. At the end of the year, if the level of hurricane damage hadn't been bad enough to trigger the default clause, principal plus interest would be repaid, and the state would put out a new issue next year. If the year was catastrophic, the state would keep the money and use it for hurricane relief.

Definitely a high risk investment, but the interest rate would compensate for the risk, and, as a bonus for researchers, the risk premium would signal the market's view on the risk that Florida would experience a catastrophic hurricane year.

CAT bonds could be sold, or re-sold, in relatively small units, so that smaller investors who wanted to put a modest part of their portfolio into high risk, high return assets could do so without risk of being wiped out completely. There's also the nice thought that, while folks in Colorado or Montana might object to being forced to pay into a federal catastrophe fund, they might be perfectly willing to buy Florida CAT bonds and take on part of Florida's hurricane risk, for a suitable risk premium.

(Here, from the St. Petersburg Times Online, is a piece dated January 20, 2006 on how the Florida Hurricane Catastrophe Fund works and on its funding.)

As I said above, I think Florida's likely to have to go it alone, and do it on a sound basis, working with the insurance industry and financial markets. With, for preference, as little insurance company bashing by state legislators as possible.

Labour Supply, Labour Demand and Wages

Perhaps it takes a guy who's actually involved in the labour market - in this case on the labour demand (meaning job offering, or hiring) side - to explain how markets work.

A piece from the CBC New Brunswick website on 12 March/07:

Lower unemployment a challenge for small businesses

New Brunswick's unemployment rate is noticeably lower than usual these days (which in NB terms means that it's down to about 7%) because so many young New Brunswickers have gone off to Alberta looking for work. There's nothing usual about that - it's been the pattern for years - but of late the outflow's been heavier than usual because of the increased demand for labour and therefore significantly higher wages to be found in Alberta. David Campbell of It's The Economy, Stupid has comments on the numbers here and here, including some caustic remarks on the issue of whether the declining unemployment rate's a good thing or not.

The CBC story whose header's above sets it up this way:

The record low unemployment rate of 6.9 per cent rate in New Brunswick is raising labour costs for small businesses.

The three-decade low is the result of a higher number of jobs combined with a shrinking workforce. There were 25,000 fewer people looking for work in the province in February.

And goes on to discuss the problems one Fredericton businessman's having hiring labour:

Jeff Magnussen, who owns Read's Newsstand in Fredericton, said he knows all too well that fewer people chasing more jobs can drive wages up. He said he's working harder than ever to attract employees.

"The choice that they have now is 'I can work at place X, or I can work at place Y,' " Magnussen said Sunday. " 'One offers considerably more per hour, and I think I'll try that one first.' So they're going directly after the wage per hour."

Read's is a good place to pick up a paper or a magazine and read it over a very large cup of coffee, by the way.

But note how well the story Mr. Magnussen tells fits as an example for use in an intro micro lecture on labour markets. Increases in the price of oil have caused the demand for labour to rise in Alberta, causing wages there to rise. New Brunswick's a perpetually economically depressed (we're actually a pretty cheerful bunch), high unemployment province. (Unemployment has remained higher in the Atlantic region than in other parts of Canada in part because the region's favourable treatment under the federal Unemployment Insurance program has raised the wage differential which would be necessary to lure unemployed low skilled people from the east into jobs further west - the boom in Alberta has raised wages there to the point where migrating is attractive for lower skilled labour.)

Anyway, because of the out-migration the labour supply curve in New Brunswick has shifted to the left, and the equilibrium wage has risen. (You'll often heard it said that wages are sticky, meaning that they don't adjust rapidly, so the labour market is slow to clear. I buy that story to a degree when the equilibrium wage falls below the current market wage, but when the equilibrium wage rises above the current market wage, the current wage rises very quickly.) As a result of the labour shortage (which is what we call it when the wage hasn't risen as far as is necessary for the market to clear, so there's still a surplus of jobs over workers), workers can pick and choose among jobs, and Mr. Magnussen's in agreement with the standard micro story when he says they're picking very much on the basis of wages. A bit more from the story:

Magnussen said it's also getting harder to find the right type of employee, and when he does, training is expensive, especially when employees are likely to leave for another job.

"The whole hiring process over the past five years, especially in the last two to three years, has become extremely challenging and extremely demanding on our business," he said.

The government of New Brunswick has responded to the situation:

The recently elected government of Premier Shawn Graham announced Friday the establishment of a Population Growth Secretariat to retain and repatriate New Brunswickers and recruit immigrants to the Maritime province.

Business New Brunswick Minister Greg Byrne said the secretariat, which will be under his department, will co-ordinate efforts to create more outreach programs and public dialogue on population issues.

I don't think I'll comment on the likelihood of that idea working. I will, however, offer one thought. The Alberta boom doesn't have to have a one-way effect on labour markets. According to this CBC report doctors are starting to leave Alberta because the boom, by increasing the demand for office space, has caused a significant increase in its price.

The Alberta Medical Association says overhead costs for family doctors have risen by between 12 and 15 per cent recently, adding that doctors have not had a corresponding raise in fees.

(That may be changing even as I post.)

So how about this. Instead of spending money on brochures to try and lure ex-pat New Brunswickers back east, maybe the provincial government should use it to attract a few of those MDs who are being driven away from Calgary.

After all, in a few years, they'll have become New Brunswickers themselves.

Financing Health Care in New Brunswick

New Brunswick Auditor General Mike Ferguson has just released the second volume of his latest annual report, and from the news coverage I learned something I hadn't known about the way NB funds its health care system.

Let's pull a headline from the Moncton Times and Transcript:

Health levy complicated: auditor

and take snips from this story in the Times and Transcript and this one from the New Brunswick Telegraph Journal. They're both via the Canadaeast website, so odds are that both are behind a subscription wall.

The thing I learned was that New Brunswick imposes a tax on auto insurance to cover at least part of the health care costs attributable to auto accidents.

Well, why not? Aren't smoking related costs of health care used to justify cigarette tax increases, and aren't they the putative reason for various jurisdictions jumping on the "sue the tobacco companies" bandwagon? Don't health activists want taxes imposed on fast food? So why shouldn't we be using taxes - are these Pigovian? - to cover automobile accident related costs?

Of course, those other examples are easier to sell, because activists have managed to make it sound as if the taxes in question are pure profits taxes, which wouldn't be passed on to consumers in the form of higher prices (although they also want to discourage people from indulging, which requires that the taxes be passed on).

In principle what we're talking about here is a pretty straight forward bit of insurance theory - your auto insurance includes an element to cover your accident-related health care costs, the price of which is included in your auto insurance premium. It's just that, since health care is paid for out of the public purse, and not out of private insurance, that bit of your premium is collected by the government, not the auto insurers.

Let's look at some bits from the two papers' stories. first, from the TJ:

Auditor General Michael Ferguson's annual report cited the health levy as a program that is confusing and doesn't even properly calculate the amount needed to cover the costs for treating motor vehicle accident victims.

"The health levy generates a lot of revenue for the province so obviously they can't simply scrap the health levy," Ferguson told reporters Thursday. "The issue for us is that it is probably a more complex process that they go through to generate that amount of revenue than they need to."

And a bit of background on the levy:

The health levy first appeared in 1993 and is designed to recoup specific health costs covered by the province resulting from motor vehicle accidents. The levy is charged to insurance companies based on the amount of premiums they bill. Drivers also pay a separate Insurance Premium Tax when they insure their vehicles.
..................................................................................................

The province budgeted to collect $16.9 million from the health levy in 2007 down from its highest of $26.7 million in 2002. The main reason for the steady decrease in the amount collected from the health levy is fewer automobile accidents occurring on the province's roads.

But presumably fewer accidents means lower health care costs, so the drop in revenue's not a problem for, say, the provincial department of finance, right?
The T&T has a bit more historical background:

Prior to 1993 the government had to go to court to fight for the money. However this process proved to be cumbersome and costly in terms of administration and court fees.

And:

Last year insurers paid about $40 per vehicle in levies. It's likely that cost was passed on to drivers, said Ferguson.
..................................................................................................

In fact the way the levy is calculated and applied to insurance companies is so convoluted it should be scrapped altogether, Ferguson recommended.

Ideally, you'd like to tie the premium to some observable characteristic of drivers which was closely correlated with the likelihood of their being involved in an accident which would impose costs on the health care system. Age, perhaps, if younger drivers are more likely to be risky drivers, or perhaps region of the province, if the probability of costly (to the health care system) accidents is higher in some regions that others? But New Brunswick's gone through some pretty acrimonious debate in the past few years about the cost of auto insurance, and various governments have promised to eliminate (what some call) discrimination in premiums based on things like age. Tricky to turn around and impose an age-based premium tax. And if you don't have a risk factor you can tie the premium to, then you might as well tie it to the fact of driving - the AG suggests replacing the levy with a fee on drivers' licenses or on motor vehicle registrations.

It'll be interesting to see how politicians react to that idea. Remember that at present the levy is nominally a tax imposed on insurers, with the amount of money each company hands over to the provincial taxman tied to the company's premium revenue. That apparent tax on insurance companies would be replaced with a tax which is collected from drivers, through their license renewal fees. I look forward to hearing the provincial minister of finance explain tax incidence theory in a province where a whole lot of people are, to put it mildly, not well-disposed toward their automobile insurers.

Here's another interesting bit from the T&T:

Legislation states insurers who are late paying the levy must lose their licence to sell policies.

However, that rule isn't enforced as the government is also missing the annual deadlines in setting the levy.

Picky, picky, picky.

Speaking of the incidence of the levy, let's go back to the TJ:

Ronald Godin, the province's independent consumer advocate for insurance, said the province should increase the levy to accurately account for the health costs associated with vehicle accidents. Otherwise taxpayers are in a way subsidizing the insurance industry.

"We should have a true reflection of what the costs are so it is not the taxpayers that pick it up," Godin said.

The taxpayer not being a driver, one presumes. Or an insurer. Let's see. If we assume that the demand for insurance is pretty inelastic, and the supply pretty elastic, and all suppliers are taxed at the same proportional rate on their revenues, and then we raise the levy, who pays the increase?

Maybe I'll set that one as an intro microeconomics exam question.

Middlemen Are Good

A lot of people take for granted the claim that "eliminating the middleman" is a good thing. they work on the assumption that middlemen make their profit from buying low and selling high - what has been called profit from alienation - without adding any value themselves. In fact, of course, every retailer is a middleman, doing what middlemen do - bringing buyers and sellers together.

The precise form a middleman takes depends on the nature of the market in question. Sometimes it's a matter of collecting a whole lot of goods together in one place, making it easy for consumers to find what they want to buy. In that case the middleman's profit is based on the value to consumers and suppliers of not having to go out in search of each other. Other times, the middleman function converges on that of the agent - travel agents, perhaps, or real estate agents, who charge a fee for their services.

Now the New York Times reports ($$$$ subscription required) on the appearance of a form of middleman in the US health care market. The article, by Michael Mason, is dated 27 Feb. /07 and is headed

Bargaining Down That CT Scan Is Suddenly Possible

and what it has to say will come as absolutely no surprise to market-oriented health economists.

First, a few excerpts:

Patrick Fontana twisted his left knee last spring while hitting a drive down the fairway on a golf course in Columbus, Ohio. But what really pained him was the $900 bill for diagnostic imaging ordered by his doctor.

Mr. Fontana, a 42-year-old salesman, has a high-deductible health plan coupled to a health savings account. Since he was nowhere near meeting his deductible, he was on the hook for the entire bill.

So he did something that insurance companies routinely do: he forwarded the bill to a claims adjuster, in this case My Medical Control, a Web-based company that reviews doctor and hospital bills for consumers.

After concluding that Mr. Fontana was not getting the best possible price, the company’s representatives called the imaging facility and demanded a lower one, promptly saving him $200 — minus a 35 percent collection fee.
..................................................................................................

The savings are possible for one reason: medical care is often priced with the same maddening, arbitrary opacity as airline seats and hotel rooms.

“The average provider — doctors or hospitals — has between 5 and 100 reimbursement rates for the exact same procedure,” said Timothy Cahill, president of My Medical Control (mymedicalcontrol.com). “A hospital chain with multiple locations may have 150 rates for the same procedure. Consumers don’t know this.”

The varying reimbursement schedules, negotiated between the nation’s 850,000 providers and more than 6,000 health plans, have been kept all but secret. Consumers almost never get information on prices before treatment. Even insurers do not know what other health plans are paying.

That lack of information is one of the primary sources of inefficiency in the US health care market. In a properly functioning competitive market, buyers and sellers are well-informed about prices, giving buyers the opportunity to seek out the best deal, and sellers an incentive to compete on price.

“Traditionally, when you went to the doctor, price was never mentioned,” said Paul B. Ginsburg, president of the [Center for Studying Health System Change, a research group in Washington]. “That’s changing, and I think it’s a healthy change.”

Damn right it's a good thing.

The reason for the imperfect information flows is historical. Back in the 1960s, when the US government introduced Medicare, its health care program for the elderly, Medicare's designers were worried that if they came up with a national fee schedule, Medicare patients would be disadvantaged in places where local fees were higher than Medicare fees, and that Medicare would be overpaying in other areas, where local fees were less than Medicare's fees. On the other hand, coming up with regional fee schedules would, at the very least, be an extremely difficult task. So instead of coming up with a list of the fees Medicare would pay for different procedures, they decided to adopt local market fees.

To do that, they came up with the concept of Usual, Customary and Reasonable, or UCR fees (sometimes known as Customary, Reasonable and Prevailing - CPR - fees). Basically, they announced that the Medicare fee would be set at the eightieth percentile of the range of fees charged in a particular geographic area for a particular procedure or service. It sounded like a perfectly reasonable idea at the time, but as this 1979 article (the link is to its abstract) by Delbanco, Meyers and Segal shows, it didn't work out quite as planned. From the abstract of the Delbanco et al. piece:

We review how Medicare's "reasonable-charge" formulas fostered Blue Shield "usual, customary, and reasonable" (UCR) contracts. In a three-year period in the Washington, DC, area, Blue Shield UCR protocols permitted "customary" allowances for selected surgical procedures to rise 29 to 75 per cent; charges by two physicians increased allowances for coronary-artery bypass from $2000 to $3500.

Note that the article refers to Blue Shield fees - at that time Blue Shield was a non-profit, doctor-sponsored insurer.

Basically what happened was that medical fees came to be like list prices for cars: remember the old saying, that nobody actually pay list? Doctors quickly realized that there was profit to be made from gaming Medicare's UCR approach, and began raising their posted fees. Medicare responded by raising what it paid accordingly. Instead of Medicare patients being at a disadvantage when it came to access to care, it wasn't long before Medicare fees were higher than the fees other insurers were actually paying. At that point, insurers like Blue Shield decided that they had no choice but to adopt the UCR approach themselves.

But wait, there's more. Doctors had traditionally engaged in price discrimination, charging different fees to different patients for the same service - charging a lower fee to low income, uninsured patients than they did for higher income, or well-insured patients, for example. But under the UCR approach they couldn't admit this - at least, they couldn't publish lists of fees they'd charge different patients, because acknowledging the low fees they charged their less generously-insured patients would pull the UCR fee down (eightieth percentile, remember?). So what they did was produce a single list of fees they'd charge, and send every patient who received a particular procedure a bill for exactly the same amount, but accept a lower payment from poorer patients. Had they done it any other way, the insurers who were covering their well-insured patients would cut the fee they paid. So doctors had a single list of fees, their discounting, or price discrimination was hidden in their reported bad debts, and nobody knew exactly how much anybody else was paying for an appendectomy.

Medicare did twig to this, and introduced a national fee schedule with different geographic allowances, but being a government operation it took them a couple of decades to pull that off.

So that's why suppliers have "between 5 and 100 reimbursement rates for the exact same procedure”. The NYT article goes on to mention that

Rudimentary information is increasingly available to consumers. Thirty-two states now require that hospitals provide pricing information to the public. Just this month, the Georgia Hospital Association started a Web site listing fees of common medical procedures at each of the state’s 141 acute-care hospitals.
...................................................................................................
Moreover, the available data are not always useful to consumers, because prices disclosed by hospitals are not the steeply discounted rates negotiated by insurers.

Nobody pays list, remember?

So why has it taken all this time for pricing information to start to appear? Because in the past, nobody cared.

“Consumers don’t care so much what the hospital charges,” said Carmela Coyle, senior vice present for policy at the American Hospital Association. “What they want to know how is much it’s going to cost them out of pocket.”

And in the past, when companies like GM negotiated health insurance plans for their employees, one attractive feature (from the employee's perspective) was a low out of pocket payment. What's changing? That mea