Pro and Contra-ception

Published

on February 10, 2012

. 146 Comments

I’d been planning a blog post on the birth control mandate, but it turns out that John Cochrane has already said everything worth saying.

Click here to comment or read others’ comments.

146 Responses to “Pro and Contra-ception”

Feed for this Entry Trackback Address

1 1 Bennett Haselton
February 10, 2012 at 3:35 am

The article says:

‘The minute pills are “free,” under insurance, the incentive for
drug companies to come up with cheaper versions vanishes. So
does their incentive to develop safer, more convenient, male-centered or nonprescription birth control.’

Why? If a competing drug company finds a way to make a cheaper birth-control pill, couldn’t some health insurance companies decide to cover only *that* pill, to save money, and hence charge less to the employers buying the insurance for their employees? (And hence be more likely to win the employer’s business, if insurance companies compete on price.) So the drug companies would still be competing on price, indirectly, via the insurance companies passing the savings on to employers.

Similarly, why would drug companies lose their incentive to make more “convenient” birth-control? If employees can choose what to spend their birth-control benefits on, they’d be more likely to choose a more convenient method (all other things being equal), and the drug company still makes money from the sale.
2 2 Robert Lewis
February 10, 2012 at 8:44 am

I can see the argue that, by making birth control free to the end user, it is cutting down on incentive for suppliers to compete on price, but wouldn’t that increase the incentive to compete on everything else? As so well evidenced by the regulation of the airlines in the middle of last century, when you can’t compete on price you compete on everything else instead.
3 3 Joe
February 10, 2012 at 9:18 am

So many things wrong with this.

“I put “insurance” in quotes for a reason. Insurance is supposed to mean a contract, by which a company pays for large, unanticipated expenses in return for a premium”

Health insurance hasn’t been “insurance” for a very long time. Health “insurance” is the way most of the world funds medical care. That includes preventative care and routine checkups that wouldn’t fall under his definition of insurance. Cochrane seems to argue that “insurance” is bad because it isn’t insurance. That’s just nonsense. It’s either good or bad independent of what we call it.

There is really no interesting economics here, other than the trivial observations that this will increase access to and utilization of birth control, and someone has to pay for it. It’s simply a political decision, and Cochrane’s article is simply a political rant.
4 4 Eric Nilsson
February 10, 2012 at 9:24 am

The article, addressing employee insurance, states: “If your employer pays you $100 less in salary and buys $100 of group insurance for you, you don’t pay taxes on that amount. Hence, the more insurance costs and covers, the less in taxes you seem to pay. (Even that savings is an illusion: The government still needs money and raises overall tax rates to make up the difference.)”

I think that insurers cover items based on cost, not “‘that’ pill”, implying specific items. So there are incentives for companies to reduce costs to gain insurance coverage; this will not necessarily make products better, just cheaper.

I think people overlook a very important, yet wholly unstated point regarding contraception. Contraception reduces the possibility of a pregnancy and thus reduces replenishment. That’s not a religious argument, that’s a statistical fact. Since most “safety net” features of the government are paid through workers’ earnings and taxes, reducing the worker supply must reduce earnings and taxes, leading to the problem John Cochrane cites in his article (a snippet of which begins this post).
5 5 Robert Lewis
February 10, 2012 at 10:04 am

Eric:

While replenishment may be an issue in some countries, it is not in the US. The US currently has a birth rate above the replacement rate. Additionally, birth is not the only means of replenishing the work force, immigration works as well. Currently, we are not facing any shortage of people who want to immigrate to the United States.
6 6 Ken B
February 10, 2012 at 10:08 am

@Bennet Haselton:
The mandate covers the pill for women. It does not cover the pill for men, so there is a (much) smaller market for a pill for men and so (much) less incentive.

Cover only “*that* pill”? and restrict women’s choices??? (That’s the argument you will hear. I don’t endorse it.)
7 7 Floccina
February 10, 2012 at 10:44 am

The only thing that I can figure is that someone was looking for a fight for some political reason.
8 8 Todd
February 10, 2012 at 10:47 am

Joe: “Cochrane seems to argue that “insurance” is bad because it isn’t insurance. That’s just nonsense. It’s either good or bad independent of what we call it.”

The reasons that this is bad are precisely what this article deals with; his pointing out that it is not “insurance” is true, but not central to his point.
9 9 nobody.really
February 10, 2012 at 10:51 am

Salting mandated health insurance with birth control is exactly the same as a tax—on employers, on Catholics, on gay men and women, on couples trying to have children and on the elderly—to subsidize one form of birth control.

Whereas socializing the cost of unplanned pregnancies does NOT impose a tax on employers, on Catholics, on gay men and women, on couples trying to have children, on the elderly…?

The author has some fine points about the nature of insurance in general. But he fails to wrestle with the idea of substitutes.

First (and I know this is a challenge for Cato people to acknowledge), externalities exist. At least, they exist as much as rocks and natural numbers do.

Evidence suggests that unplanned pregnancies contribute mightily to these externalities. If you buy one of the theses from Freakonomics, abortion rights lead to reduced crime; the implication is that people who are born into homes that did not want them are more likely to become criminals. This strongly suggests that we all have an incentive to discourage unwanted births. Ensuring abortion rights is one vehicle, but subsidizing birth control might well be a more efficient one.

But even if you deny the broad externalities issue, you can’t ignore the narrow one: CHILD BIRTH is expensive and, thanks to Obamacare, the cost of it will be socialized. Extraordinary child births – with lots of neonatal and post-natal interventions – get really pricy. And don’t forget the liability that health care providers bear for screwing up a delivery! We could “squander” a lot of money on birth control before we’d exceed the cost of even a single child birth.

Thus, even if I were persuaded that mandating birth control insurance is bad public policy standing alone, it would not follow that it would be bad public policy in a world in which we already socialize other health care costs – especially the cost of child birth. Maybe the various concerns raised by the author actually do exceed the substitution effects, but I’d want to see more analysis than he offers.

There is also the issue of religious freedom. Our nation is divided on social issues. The natural compromise is simple: Birth control, abortion and other contentious practices are permitted. But those who object don’t have to pay for them. The federal takeover of medicine prevents us from reaching these natural compromises and needlessly divides our society.

Eh. I subscribe to checkbook libertarianism: To the maximum extent feasible, I want government to maximize autonomy – except for autonomy about whether to pay taxes. Where government imposes burdens, I prefer having those burdens reduced to the form of cash.

I opposed the Iraq war. Consequently I chose not to enlist, and government respected my choice. But as far as government was concerned, that did not relieve me of the duty to pay my taxes that would fund the war.

Some employers don’t like paying for insurance that includes birth control. Government respects that choice; I am not aware that ANY employer has a duty to provide health insurance to employees. But employers that do not provide insurance for WHATEVER reason then have to pay into the fund that underwrites insurance for others, and that fund will pay for the birth control.

Does that offend people? Sure – in the same ol’ way that any collective decision will offend dissenters. I’m not seeing anything new, or especially sub-optimal, here.
10 10 nobody.really
February 10, 2012 at 11:33 am

[B]irth is not the only means of replenishing the work force, immigration works as well.

One more example of the US relying on the unskilled labor of foreign countries….
11 11 Ken B
February 10, 2012 at 11:43 am

nobody.really makes an amusing remark:

“First (and I know this is a challenge for Cato people to acknowledge), externalities exist. …But even if you deny the broad externalities issue, you can’t ignore the narrow one: CHILD BIRTH …”

This is amusing as he is ignoring an externality much discussed on this board: other peoples’ kids are a positive externality.
12 12 Harold
February 10, 2012 at 11:50 am

He does have a point about employer provided policies – these are a terrible idea, and must do great harm to the USA in discouraging movement of labour and entrepreneurs from setting up on their own. This, or something like it, is a predictable outcome of insurance-based health provision, as insurers try to avoid taking on those with high risk. Employed people are low risk, so there will be attempts to cherry pick them.

I think he falls down on some of his other arguments. Since when did preventative measures not lower overall costs? The medical cost of a contraceptive – a few cents. The medical cost of a pregnancy many, many times more. This is true of the other preventative health measures mentioned. The direct medical costs will probably be lower on their own, but when we factor in the wider costs of sickness to individuals and society, prevention clearly saves money. Sickness (and unwanted pregnancy) prevention has positive externalities, probably occurs too little, and so should be encouraged.

Bennett has a point about competition above – insurance does not remove competition, it only makes the decision maker a different person. Unless the decision maker has no interest in the cost or effectiveness of the product there will still be competition. So there goes another argument.

He uses oil changes and toothpaste as examples of why this sort of thing is a bad idea. Economists often use this sort of example to illustrate their points. Clearly we are to conclude that if it is good for toothpaste, it is good for medical insurance also. Are there any differences between contraception and preventative healthcare, and toothpaste? For example, do we have good evidence that toothpaste is used less than 50% of the recommended amount? Are people as aware of the consequences of their healthcare plans as they are about cleaning teeth? Is there an effective cosmetic and odour incentive to encourage healthcare as there is for teeth cleaning? Can we get caught in an unexpected uncleaned-teeth incident that will have life changing consequencies?

The real nub of it is here:
“It’s not about “access” and it’s not about “insurance.” It’s because Americans, when paying even modest copayments, choose to spend their money on other things.” This is the real point – should we get out of the way and let people make their own choices?

It depends what we want. If we use the end point of many free-market economic arguments, we want to maximise utility. A free market is the only way to ensure that we end up in position that we cannot make anyone better off without making someone else worse off, according to their own preferences. There are arguments against desiring this, but for now I am happy to say that this is the end we wish to achieve.

A free market is the only way we can achieve this end. Any interference in the working of this market *must* cause us to end up in a worse position.

So what is a free market? It requires certain things – including competition, no barriers to entry, and informed, rational decision makers. If we do not have these, then we *must* end up at a less than optimum position.

If we are at a less than optimum position, it is possible in principle to improve things with further interference. Of course, it is also possible to make things much worse.

Steve has often posted about situations where people make “mistakes” in their assessments – often about probabilities. The fact is we know that people make these mistakes, and if we can predict these mistakes, it is possible to correct for them to improve things, and get closer to the optimum position a properly free market would provide.

There are few areas of life where choices are less rational and less informed than pregnancy and healthcare. We can safely assume that we are far from the optimum position we desire. To suggest that just to leave things as they are will produce the best outcome is clearly nonsense. There is a very clear case that intervention could make things better. I am not sure that the specific interventions outlined here are sure to do this.
13 13 Ken B
February 10, 2012 at 11:58 am

“One more example of the US relying on the unskilled labor of foreign countries….”

Not entirely fair. We rely on you for our unskilled comments.
14 14 Joe
February 10, 2012 at 12:28 pm

Todd:”The reasons that this is bad are precisely what this article deals with”

But it never does. He’s writing as an economist, but the only time he come close to an economic argument is “The minute pills are “free,” under insurance, the incentive for drug companies to come up with cheaper versions vanishes.” And that’s just wrong, as pointed out in the first response by Bennett Haselton.

Beyond that, the only argument he has is based on religious freedom. Well that hardly supports the claim that “Critics are missing the larger point”. That is itself the critics point.

He and everyone else is entitled to his opinion, but what he wrote is nothing but an economics-free screed.
15 15 Ken B
February 10, 2012 at 1:02 pm

Harold: “Since when did preventative measures not lower overall costs?” Since the year 1. Frequently, and probably mostly, ‘preventative’ is on net more costly. Do men have PSA tests starting at age 21? Do we give 18 year old women bone desnity pills? No — too costly. Do you boil your tap water, irradiate your carton of milk, disinfect your paperbacks before you read them?

if contraceptives are as cheap as you say, why can’t people pay for their own?
16 16 nobody.really
February 10, 2012 at 1:19 pm

…But even if you deny the broad externalities issue, you can’t ignore the narrow one: CHILD BIRTH …”

[O]ther peoples’ kids are a positive externality.

I concede the point.

I suspect we could get a lot of agreement that kids produce positive externalities in aggregate or on average. But if we disaggregate the data, I suspect we could get even more agreement that some kids produce more positive externalities than others. Freakonomics suggests that unwanted kids tend to be in the “others” category. I’m suggesting that we design public policy to accentuate the positive.

But as Eric Nilsson and Ken B suggest, to say that some kids produce more positive externalities than others is not the same thing as saying that unwanted kids in general, or any unwanted kid in particular, actually generate net NEGATIVE externalities. I’m inclined to think that some do, but it’s a more controversial proposition.

Anyway: If we think that some circumstances will produce people that generate net negative externalities, then we might want to design public policies to discourage people being born into those circumstances. That might involve improving the circumstances. And it might involve discouraging births.

But if we merely identify circumstances that will produce people that generate a less-than-average amount of positive externalities, then we may want to design policies to improve those circumstances, but may not want to design policies to discourage births.
17 17 KS
February 10, 2012 at 1:20 pm

Is this the same John Cochrane who recently wrote any government stimulus during a recession is necessarily offset by loss of private spending becaus GDP = C + I + G and therefore if G goes up C must go down — making a mistake that high school economics students get routinely penalized for — then stayed silent for a while, then went on this childish “Krugman is so mean” rant and entirely recanted what he said?

Excuse me if I’m not holding my breath over the depth of his arguments here.
18 18 Ken B
February 10, 2012 at 1:45 pm

Is the this the same KS who made a mistake in his high school logic course? Because if so I guess I can ignore what he actually said this time. Such a time-saver.
19 19 Todd
February 10, 2012 at 1:51 pm

Joe,

1. Mandated coverage forces those who do not want that coverage to subsidize those that do, and distorts incentives such that these services are overused relative to alternatives

2. Mandating that insurance cover routine expenses creates unnecessary, burdensome red tape

3. Incentives to create alternatives to such things as “the pill” are significantly reduced (‘vanishes’ is an overstatement, but I reject Bennet’s argument that incentives are unaffected)

4. Distorting incentives to increase the number of wellness visits increases healthcare costs

You may not agree with them, but I’m not sure how you can say that they aren’t economic arguments.
20 20 KS
February 10, 2012 at 1:52 pm

For a distinguished Chicago economist to make elementary and boneheaded mistakes does reduce the probability that he should be taken seriously in the future. In what other field can a tenured professional make high school level mistakes, and still be listened to?
21 21 KS
February 10, 2012 at 1:56 pm

Economics is stupid, because everyone argues on purely theoretical grounds with no emphasis on evidence. The people are smart, the field is stupid. Those of us in physics and medicine are equally smart, yet continually humbled by the complexity of the real world. Economists are trained to be arrogant.

Good example — Cochrane writes “‘The minute pills are “free,” under insurance, the incentive for drug companies to come up with cheaper versions vanishes. So does their incentive to develop safer, more convenient, male-centered or nonprescription birth control.’”

What the hell is this based on, exactly, apart from his theoretical and imagined relationships? Couldn’t it be the case pills become “free” — yet the insurance company re-evaluates what pill to provide each year — so the incentive remains? This is the problem with the “dismal science.”
22 22 Ken B
February 10, 2012 at 1:58 pm

@KS: I refuse to dismiss Paul Krugman so airily.
23 23 KS
February 10, 2012 at 2:24 pm

Please provide a time Krugman made a mistake that elementary. Cochrane asserting government spending is always non-stimulative because of accounting is as stupid as me, a medical professional, saying DNA is made of proteins.
24 24 Ken B
February 10, 2012 at 2:26 pm

Robert Lewis:
“I can see the argue that, by making birth control free to the end user, it is cutting down on incentive for suppliers to compete on price, but wouldn’t that increase the incentive to compete on everything else? As so well evidenced by the regulation of the airlines in the middle of last century, when you can’t compete on price you compete on everything else instead.”

This is exactly right. But this ‘non-price’ competition [better description: competition at fixed prices] has lower social utility that when price competition is part of the mix. The incentive to produce cheaper versions does not vanish, but it will be sharply reduced, especially if the mandate allows patients to pick their preferred medicine.
25 25 Todd
February 10, 2012 at 3:16 pm

Bennett (echoed by others): “If a competing drug company finds a way to make a cheaper birth-control pill, couldn’t some health insurance companies decide to cover only *that* pill, to save money…”

In theory, this could happen, but in practice it does not. No health insurance company that I am aware of dictates which brands are covered and which are not for the same type of drug.

I feel confident that if a health insurer did attempt to do this, their savings would be very small relative to the amount of business they would lose for restricting their clients’ choices. Nobody wants health insurance that tells you (and your doctor) which drugs you can and cannot have.

It may be an overstatement to say that the incentive to innovate “vanishes,” but it isn’t much of one.
26 26 KS
February 10, 2012 at 4:02 pm

Correct me if I’m wrong. Cochrane argues that an insurance plan that offers contraception for free decreases the incentive for research to develop new and better forms of contraception. Why exactly is this the case?

Let’s make an agreement regarding incentives — no-one is smart enough to absolutely predict the change in behavior of a large population of people by a flowchart they make in their head.
27 27 AndrewG
February 10, 2012 at 4:25 pm

KS, My stab at an answer is that health insurance plans typically only cover drugs that are part of “standard medical practice”. New drugs by definition are outside this scope. This adds another hurdle/cost for a new drug to become profitable. This should lower the incentive to innovate.
28 28 Martin
February 10, 2012 at 5:17 pm

KS: No-one is smart enough to “absolutely predict” anything in the real world. Didn’t you say you were a doctor?
29 29 Al V.
February 10, 2012 at 5:21 pm

I think Mr. Cochrane somewhat misses the point. He is right that covering repetitive prescriptions doesn’t make sense. If I take a daily contraceptive, blood pressure medication, or antidepressant, someone else is subsidizing my medication. However, once we assume that these types of prescriptions in the coverage, it is unconstitutional to discriminate against one class of employee (non-Catholic women), by not covering their contraceptives while covering my blood pressure medication.
30 30 KS
February 10, 2012 at 5:56 pm

“KS: No-one is smart enough to “absolutely predict” anything in the real world. Didn’t you say you were a doctor?”

Fair enough, I should add a caveat. Science requires something to be true more than 95% of the time to be accepted (with a few not-relevant exceptions). In medicine, we only apply treatments if they have been shown to be effective at doing what they’re intended to more than 95% of the time.

So, nobody is smart enough to predict changes in behavior/impact on incentives accurately MORE THAN 95% OF THE TIME based on a flowchart in their head. That includes me, you, SL, and Cochrane.

The difference? I don’t write entire WSJ editorials and give speeches assuming I can do that. Humility is the key to science.
31 31 Harold
February 10, 2012 at 6:17 pm

Ken B: Good point about prevention- indescriminate and useless attempts to prevent things would be more costly. The word “effective” was lost somewhere in editing. Preventative measures must be evidence based- these measures will be cost effective, and only these should be covered by insurance.
32 32 Patrick R. Sullivan
February 10, 2012 at 6:34 pm

‘Please provide a time Krugman made a mistake that elementary.’

Several months ago he attempted to denigrate the governor of Wisconsin’s policy toward his public employee unions by comparing the performance of Texas public schools (where the unions had less power) to those of Wisconsin. Without bothering to control for the very different demographics of the two states (Texas borders Mexico and Wisconsin doesn’t).

When controlled for different ethnic make-ups the results were the opposite of what Krugman claimed. That’s about as elementary an error as a supposed professional could make.
33 33 KS
February 10, 2012 at 6:59 pm

“When controlled for different ethnic make-ups the results were the opposite of what Krugman claimed. That’s about as elementary an error as a supposed professional could make.”

That certainly would be an error — not controlling (or at least trying to) for known confounders. Do you have a link?

As elementary an error as possible? No. Controlling for confounders via multivariate regression, randomization, or some other process is fairly advanced; maybe graduate level. It is certainly NOT a high school level mistake.

Arguing that A = B + C and therefore if B goes up C must go down is a high-school level economics mistake. I know, because my high school economics teacher used to instill that point to us all the time. To say that fiscal stimulus doesn’t do anything because if G goes up C must go down is high school econ 101 mistake. And sure enough, go to Cochrane’s blog, go back long enough, here is what you find:

“Every dollar of increased government spending must correspond to one less dollar of private spending. Jobs created by stimulus spending are offset by jobs lost from the decline in private spending. We can build roads instead of factories, but fiscal stimulus can’t help us to build more of both. This is just accounting, and does not need a complex argument about ‘crowding out.'”

To think about how stupid this is really, ask yourself, is government spending inherently different from private spending? If me and 49 friends get together and each contribute $10 so we spend $500 total, would Cochrane argue our spending has no net stimulatory effect? And if not, is that fundamentally different from government spending???
34 34 Will A
February 10, 2012 at 7:55 pm

The article says:
You’d have to fill out mountains of paperwork,
the oil-change and toothpaste markets would become much less competitive, and you’d end up spending more.

And yet he provides no reference to a study/or data to back up his claim.

Maybe insurance company would end up paying more with the cost of unwanted pregnancies (hospitalization, the percentage of pregnancies that go bad, etc.) if they didn’t provide this coverage.

How can economist make such statement without the data (or link to data) to back up his claim.

If his claim is that insurance companies shouldn’t be mandated to provide coverage for anything regardless of whether the coverage saves money, he should have said this.

What sloppy work.
35 35 Ken B
February 10, 2012 at 8:33 pm

Will A:
“What sloppy work.”

Will, I think you should keep the self-referential stuff on the Godel posts where it fits in naturally. Steve makes about 15 of those a month; the wait won’t be long.

“… toothpaste markets would become much less competitive, and you’d end up spending more. And yet he provides no reference to a study/or data to back up his claim.” You are quite right, he does not. Perhaps toothpaste markets are special, and less competitive ones have lower prices. Perhaps toothpaste is the holy grail of pro-monopoly advocates, the case where it works best. It merits attention. Still for now I’m willing to go with the odds on this one: competition lowers prices. Even for toothpaste.
36 36 Ken B
February 10, 2012 at 9:30 pm

KS:
“Cochrane argues that an insurance plan that offers contraception for free decreases the incentive for research to develop new and better forms of contraception. Why exactly is this the case?”

Actually he doesn’t argue quite that. He says ‘free’. The quotes indicate irony. The pills are not FREE it is just that the cost of choosing a more expensive one is not borne by the person making the choice. Or not fully borne anyway. That means that the choosers — women on the pill — will choose more of the expensive ones and will not react much to a rise or fall in the actual cost. So the providers have less incentive as new competitors will not tempt the choosers with cheaper offerings. If I offer to pay for your coffe in the morning I bet you’d order more lattes. And the coffee shop down the street would have less incentive to provide a cheaper alternative that is 90% as good and 10% as expensive since (not paying the price) you won’t choose the 90%. Same principle.
37 37 Joe
February 10, 2012 at 10:25 pm

Todd,

1. Mandated coverage forces those who do not want that coverage to subsidize those that do, and distorts incentives…

–So what? My taxes pay for things that I don’t like. I pointed out the distorted incentives, and that it’s a trivial point.

2. Mandating that insurance cover routine expenses creates unnecessary, burdensome red tape

– only asserted through analogy. He doesn’t try to make the case that this applies directly to the current case because he can’t

3. …‘vanishes’ is an overstatement

– point made

4. Distorting incentives to increase the number of wellness visits increases healthcare costs

– again, acknowledged as trivial economics, and probably beneficial

You may not agree with them, but I’m not sure how you can say that they aren’t economic arguments.

– again, to the extent they are correct, and “economic” they remain trivial.
38 38 KS
February 10, 2012 at 11:09 pm

Ken B:

Thanks, this is a classic example of the economist’s fallacy, also known as “I make a flowchart in my head ergo I’m right where’s my paycheck?” Let’s deconstruct, shall we?

1. “The pills are not FREE it is just that the cost of choosing a more expensive one is not borne by the person making the choice. Or not fully borne anyway.”

– agreed

2. “That means that the choosers — women on the pill — will choose more of the expensive ones and will not react much to a rise or fall in the actual cost.”

– Wtf? How does this follow from #1? Why would the women necessarily choose a more expensive pill? If a woman has been using the same form of contraception and it works great, she’s not going to switch just because a more expensive one comes out. Also, why would she become less sensitive to changes in price, just because she bears less of the cost? An increase in price of the contraceptive could affect her insurance premium, making her sensitive. Neither of your two points necessarily follow from the premise, except of course in your head.

3. “So the providers have less incentive as new competitors will not tempt the choosers with cheaper offerings.”

– This might be even more of a leap from 1->2 was. First, I’m not sure what you mean by ‘providers’, as providers are doctors, and not a part of this discussion (we’re talking about companies who make contraceptives, insurance companies, and patients). Second, why will new competitors not attempt to develop superior and/or cheaper offerings? You seem to say the women are less price sensitive, ergo they’re less likely to choose cheaper options. Even if that’s true (which it might not be), it’s irrelevant. Keep in mind the relevant relationship here is between the insurance company and the drug company. If a drug company makes a cheaper and superior contraceptive , insurance companies will contract that in their formulary instead. The incentive to produce cheaper and/or better drugs has not “vanished”. Remember, the patient does not contract directly with the drug company — that’s why it’s an insurance market!

4. “If I offer to pay for your coffee in the morning I bet you’d order more lattes. And the coffee shop down the street would have less incentive to provide a cheaper alternative that is 90% as good and 10% as expensive since (not paying the price) you won’t choose the 90%. Same principle.”

– Incorrect times two. First, I don’t like coffee, so my increase in coffee consumption will be zero even though I would bear less of the coffee cost if I chose to consume it. (Hey, it’s almost like you can’t predict my demand curve based on your theoretical flowchart!) Second, say I do like coffee, and you are the insurance company who will pay for 80% of my coffee (ie, my co-pay is 20%). The coffee store will STILL have an incentive to make cheaper and/or better coffee, as you will then contract with them to get coffee for cheaper. I don’t buy from the coffee store directly, if this was an insurance market.

I appreciate your comment though, it is a CLASSIC example of what’s wrong with economics. Let’s paint a narrative, ignore all confounders, assume linearity, and try to construct meaningful policy!
39 39 Ken B
February 10, 2012 at 11:19 pm

@Joe: You keep using this word ‘trivial’. I do not think it means what you think it means.

Todd’s points are all correct, and easily seen to be so. That makes their demonstration trivial. It does not mean their significance is trivial.

And your comment under point 4 shows you do NOT understand the point, trivial or not. Distorted incentives produce poor results. Poor results are not ‘probably beneficial’.

Under 3, no point not made. A rhetorical flourish — ‘vanish’ — is used to emphasize an important point, that you miss.
[“Hitler was totally evil.” “Not totally, the Autobahn was a good thing.” “OK not totally.” “Point made.”]
40 40 KS
February 10, 2012 at 11:23 pm

“Distorted incentives produce poor results. Poor results are not ‘probably beneficial’.”

Uh huh. Right. This is always the case, that distorted incentives produce poor results? So if you’re a murderer and I pass a law saying those who kill go to jail for life, I have effectively distorted your incentive to kill. The overall result for society is poor?

Where do you get the idea that distorting incentives ALWAYS produces poor results? It takes a vast well of intellectual hubris to assume something like that.
41 41 Todd
February 10, 2012 at 11:31 pm

Joe: “I pointed out the distorted incentives, and that it’s a trivial point.”

This statement trivializes the entire field of economics, which is all about incentives and distortions thereof. No wonder you didn’t see any relevant economics in his article–you don’t seem to believe such a thing exists.
42 42 Ken B
February 11, 2012 at 12:39 am

KS proves he misses the point. He wrote
“Wtf? How does this follow from #1? Why would the women necessarily choose a more expensive pill? If a woman has been using the same form of contraception and it works great, she’s not going to switch just because a more expensive one comes out.”

But the point at issue is the lack of incentive to create new cheaper pills. So the comparison is the hypothetical NEW and CHEAPER pill vs the existing and COSTLIER one BY HYPOTHESIS.

You misunderstand the coffee analogy too. Did you notice the competing coffeeshop? Do you see the relevance of competition? And 20% of a price reduction might not motivate a change that 100% of the same reduction would.
43 43 Ken B
February 11, 2012 at 1:11 am

I think I found my quote of the year:

“Also, why would she become less sensitive to changes in price, just because she bears less of the cost?”

This is almost too beautiful to burst, and I should just leave it in its pristine perfection, a treasure for the ages. But I won’t. My price sensitivty to nonsense is just too high.

KS seems to think we are discussing some ONE woman somewhere. We are not. Talk about ‘the chooser’ or ‘the customer’ is shorthand. We have a large number N of women. Their price sensitivity will vary. It is indeed not the case that any particular woman may or may not be swayed by a price change. For one woman at random we cannot tell. But when N is large some will. And that some make a difference. They change their behaviour and that provides incentives, which changes the behaviour of sellers and producers. These single person examples are ex hypothesi persons on the margin, to allow us to use the short hand. (KS clearly does not get using hypotheticals and margins, as I noted earlier.)

And KS, thank you!
44 44 AMTbuff
February 11, 2012 at 3:58 am

Illegal immigrants: Having the babies Americans won’t have!
45 45 KS
February 11, 2012 at 8:45 am

“But the point at issue is the lack of incentive to create new cheaper pills. So the comparison is the hypothetical NEW and CHEAPER pill vs the existing and COSTLIER one BY HYPOTHESIS.

You misunderstand the coffee analogy too. Did you notice the competing coffeeshop? Do you see the relevance of competition? And 20% of a price reduction might not motivate a change that 100% of the same reduction would.”

Ken, I feel like you’re a lost cause; the number of incorrect things you have said is just alarming. Question: what’s a provider? Second question: why do distorted incentives always produce poor results?

Anyway, insurance companies get to choose which products they cover, or which drugs they provide as part of their formulary. Usually they re-evaluate these on an annual basis. So say an existing contraceptive is out on the market, and now a drug company produces one that is cheaper and as effective. Insurance companies, when they re-evaluate their formularies, will choose the new one and no longer offer the older, more expensive one. The drug company who makes the new contraceptive has now won out.

The incentive to produce new better/cheaper contraceptives has NOT “vanished”. It is clear from your writings that your understanding of the basics of the health insurance market are, um, lacking. (Question: who’s the provider again? I’m still confused as to why you mentioned providers in this context).

After all, if it is the case that providing health insurance for a good X makes the incentive to develop better X vanish, wouldn’t this be true for ALL of health care? And let me tell you my ignoramus friend, as a medical professional, one thing you can’t say about the health care market is that companies don’t spend billions performing research to produce even marginal improvements on the existing standard-of-care — because they now, even a marginally-improved X will quickly dominate over X. (Contemporary example: argatroban vs. warfarin).
46 46 Sonic Charmer
February 11, 2012 at 9:24 am

It seems to me the people defending (in principle) the mandate on the public-policy grounds that doing so could be a cost-effective way to mitigate externalities need, in order to be saying something, to postulate the existence of (a significant number of) women on the relevant margin.

What margin? Namely this one: women who are sexually-active, and are (absent “insurance”) somehow unable/unwilling (?) to just absorb a modest ongoing $35/month cost (or whatever; it’s not like these things cost $10k is it??) of going and getting pills themselves from, like, Planned Parenthood (probably comparable to what they spend on hair products, and a fraction of their iPhone bill), and so – purely because of the cost – are not on the pill, but would be if it were covered via such a mandate.

I suggest that for all intents and purposes of this discussion, approximately zero such women exist. So, what are you talking about?
47 47 Ken B
February 11, 2012 at 10:11 am

KS: “(Question: who’s the provider again? I’m still confused as to why you mentioned providers in this context)” And the context of my paragraph makles plain what I mean bt provider. You seem to think it’s a synmym for doctor. It’s not; it’s a proper English word. Try ‘producer’ or ‘supplier’ and see if that helps.
48 48 Ken B
February 11, 2012 at 10:13 am

Let me summarize the arguments of the plan’s defenders:

JOE: It is trivial to discredit the plan.

KS: The plan is fine if people never react to incentives.

I agree with both statements!
49 49 KS
February 11, 2012 at 10:57 am

In a health insurance context, provider means physician. Only an idiot would use provider to mean supplier in such an exchange.

I have never said people don’t respond to incentives. I have said you cannot predict how people respond to incentives a priori.

You are accusing me of saying something I haven’t said simply because you have nothing left to say. The best thing about blogs is you can go back and see what people have said before they try to change the issue once it comes out how wrong they are.
50 50 Steve Landsburg
February 11, 2012 at 12:14 pm

Sonic Charmer:

women who are sexually-active, and are (absent “insurance”) somehow unable/unwilling (?) to just absorb a modest ongoing $35/month cost (or whatever; it’s not like these things cost $10k is it??) of going and getting pills themselves from, like, Planned Parenthood (probably comparable to what they spend on hair products, and a fraction of their iPhone bill), and so – purely because of the cost – are not on the pill, but would be if it were covered via such a mandate.

I suggest that for all intents and purposes of this discussion, approximately zero such women exist. So, what are you talking about?

This paper suggests that zero is in fact a rather poor approximation.
51 51 Sonic Charmer
February 11, 2012 at 1:16 pm

Steven –

I saw that linked at MR too. I’m skeptical of the ramifications of a natural experiment on college girls for a policy that, basically by definition, affects full-time employees.
52 52 Will A
February 11, 2012 at 2:22 pm

@ Ken B:

Perhaps toothpaste markets are special, and less competitive ones have lower prices.

And perhaps if dental insurers had to buy toothpaste for their customers, dental insurers would become middle men and look for the best price toothpaste they could find so that they could make a profit on the $2 they got from their customers.

What do the models/data show? Do government imposed mandates in certain industries lead to innovation and competition?

I’m inclined to go with the odds that at least 1 government mandate lead to innovation and competition.

I’m also inclined to go with the odds that Cochrane was being sloppy.
53 53 KS
February 11, 2012 at 2:50 pm

@ Will A:

“And perhaps if dental insurers had to buy toothpaste for their customers, dental insurers would become middle men and look for the best price toothpaste they could find so that they could make a profit on the $2 they got from their customers.”

Exactly! Ken B is a lost cause. He clearly doesn’t understand the basics of the insurance market. Ie, insurance companies get to choose which products they insure, so their drive to choose cheaper and/or better products is itself a form of competition, even if the patient (insured) bears a smaller burden of the cost.

Cochrane isn’t being sloppy, he’s just an idiot. Read my quote from him about how government spending is necessarily non-stimulative because of “accounting.”

You may ask, how can an idiot be a distinguished professor in a field? Well he can’t. Unless that field is economics.
54 54 AndrewG
February 11, 2012 at 4:24 pm

Let’s assume insurance companies are successful at only incorporating the cheapest toothpaste/drug into their coverage. This is effectively a price ceiling on the supplier. Price ceilings are not good for competitive markets. Also insurance companies cannot simultaneously only cover cheaper options while leaving room for better/more innovative products as higher quality and novel options are inherently more expensive.
55 55 KS
February 11, 2012 at 5:36 pm

“Let’s assume insurance companies are successful at only incorporating the cheapest toothpaste/drug into their coverage.”

– Not true, insurance companies cover a host of similar medications, some of which may be cheaper (generic) than others.

“This is effectively a price ceiling on the supplier.”

– Ummm, no. A price ceiling would be if an insurance company were to mandate a maximum price in the entire market. Insurance companies don’t have the power to do that in the US (single-payer schemes do). An insurance company bargaining with a supplier (pharm company) about pricing of drugs is NOT a price ceiling. This comment made my head hurt.

“Price ceilings are not good for competitive markets.”

– Correct. Once again, not relevant.

“Also insurance companies cannot simultaneously only cover cheaper options while leaving room for better/more innovative products as higher quality and novel options are inherently more expensive.”

– Incorrect. Insurance companies offer coverage of many similar medications, some of which may be cheaper, like generic alternatives. This comment made my head hurt x2.

Look, if the argument you + Ken B + Cochrane are trying to make is that by offering insurance for a particular good X, the incentive for other companies to develop better and/or cheaper equivalents to X vanishes or is significantly reduced, you are DEAD wrong. Fact.
56 56 AndrewG
February 11, 2012 at 5:46 pm

If insurance company covers multiple drugs with different prices (as you yourself say) with no copay, there is no incentive for a consumer to choose the cheaper option and thus there is no incentive to develop cheaper equivalents. This is a fact.
57 57 KS
February 11, 2012 at 5:54 pm

“with no copay”

Sigh. All plans have copays, or deductibles, or some mechanism whereby the consumer faces a fraction of the cost.

“there is no incentive for a consumer to choose the cheaper option and thus there is no incentive to develop cheaper equivalents.”

Sigh times two. No, because when the insurance company re-evaluates its formulary, it will kick out expensive drugs that are making it face a loss and choose cheaper drugs where it would turn a profit. And if a drug company happened to produce cheaper equivalents, it would reap those rewards.

Think of the insurance company as the “consumer” and the drug company as the “producer” in this relationship, as the insurance company is deciding what drugs it will cover (pay for). Would the insurance company not prefer a cheaper alternative?
58 58 Ken B
February 11, 2012 at 5:55 pm

KS: “I have never said people don’t respond to incentives.”

KS: “Also, why would [a marginal consumer] become less sensitive to changes in price, just because she bears less of the cost?”
59 59 KS
February 11, 2012 at 5:58 pm

LOL. I was saying, a consumer doesn’t necessarily become less price sensitive even if she bears a smaller proportion of the cost. That is equivalent to me saying one cannot predict changes in incentive a priori. Ken B, you are a huge source of amusement
60 60 Ken B
February 11, 2012 at 6:01 pm

KS (again): “Look, if the argument you + Ken B + Cochrane are trying to make is that by offering insurance for a particular good X, the incentive for other companies to develop better and/or cheaper equivalents to X vanishes or is significantly reduced, you are DEAD wrong. Fact.”

It depends,as we have said repeatedly, on WHO decides what unit of X gets purchased. I consume more blood tests than I would if I bore the whole cost. Fact.

There are two effects here. Overconsumption of X and consumption of the costlier versions of X. Both are effected when the consumer of X does not pay the price for X.

KS and crew argue price incentives do not matter. Maybe that is true for toothpaste and the pill, or a contraceptive you brush your teeth with, but it’s false for everyhting else in life. Or do you own 17 Ferraris?
61 61 AndrewG
February 11, 2012 at 6:04 pm

It was my understanding that it was part of the HHS plan to mandate insurers not to charge co-payment for contraception. That is why Cochrane labels it as “free”
62 62 KS
February 11, 2012 at 6:06 pm

“KS and crew argue price incentives do not matter.”

Nope, no-one has said that. Clearly, your understanding is very very limited, so I’m done trying to educate you. It’s not my fault the public school system failed you so badly.

I will leave you with an example, for your brain to process. Make sure you aren’t driving at the same time as you read this, as clearly there are some bandwidth limitations here to be concerned about.

Insurance companies have been providing coverage for medications for decades. And the medical standard of care when it comes to medications gets updated every few years, if not sooner. Warfarin relaced heparin. Now argotobran is replacing warfarin. Ace inhibitors replaced diuretics; now ARB antagonists are replacing ACE inhibitors. So the incentive to research newer and better drugs is not reduced by the presence of coverage.
63 63 Todd
February 11, 2012 at 6:47 pm

KS,

I know at least 6 people who were lifelong chain smokers and never got lung cancer. So lung cancer susceptibility is not increased by chain smoking.

Ken B,

I wouldn’t put so much effort into going back and forth with someone who clearly has too much disdain for the study of economics to put any real effort into understanding it.
64 64 Ken B
February 11, 2012 at 10:20 pm

@Todd: You’re right of course. As his comment about me committing ‘the economist’s fallacy’ shows, his main accusation seems to be that I have correctly explained the economics of the situation. I plead guilty! (So should you.) Of course I don’t serve much credit because, as has been noted, it was ‘trivial’.
65 65 Ken B
February 11, 2012 at 11:26 pm

One last try. KS wrote:
“Think of the insurance company as the “consumer” and the drug company as the “producer” in this relationship, as the insurance company is deciding what drugs it will cover (pay for). Would the insurance company not prefer a cheaper alternative?”

As I and others have noted, the strength and direction of the incentives depends on who chooses the drug. Who makes the consumption choice. Not who pays for it. It is the disjunction of choosing from paying that we point to. That is why I used the inelegant term ‘chooser’. [And the apparently incredibly difficult term ‘provider’ for the counter party.] That distinction has been made so many times here it seems obtuse of KS not to notice it. But I repeat myself.
66 66 KS
February 11, 2012 at 11:46 pm

I give up, ignorance wins again. Clearly somethings are beyond the understanding ability of some people. Peace.
67 67 Richard
February 12, 2012 at 4:09 am

As I and others have noted, the strength and direction of the incentives depends on who chooses the drug. Who makes the consumption choice. Not who pays for it.

Ken B

If the particular drug isn’t covered under the consumer’s insurance plan, they can’t choose it. Insurance companies will choose the cheapest drug that will ensure the most satisfied cliental because it affects their profits. There is no reason for innovation to be stifled.

It depends,as we have said repeatedly, on WHO decides what unit of X gets purchased. I consume more blood tests than I would if I bore the whole cost. Fact.

No, first it depends on who decides what X can be purchased before the number of units.
68 68 Ken B
February 12, 2012 at 10:16 am

@Richard: It simply IS a fact that I consume more blood tests when I don’t pay for them — or pay only a fraction of the cost. Price affects demand curves. If you deny this you simply are not worth talking to.
69 69 Ken B
February 12, 2012 at 10:17 am

“ignorance wins again”

Of course. This is a struggle between the Obami and the Roman Catholic church. Ignorance has to win.
70 70 Ken B
February 12, 2012 at 10:23 am

Reuters:

“An Obama administration official said the new policy would not allow health insurers to increase their premiums, charge co-payments or deductibles to make up for the cost of contraceptives.”

http://ca.news.yahoo.com/insurers-stuck-cost-obama-birth-control-plan-003257037.html
71 71 AndrewG
February 12, 2012 at 10:38 am

Richard,

It is simply a reality that insurance companies do not just cover the cheapest drug. As KS noted, they often cover several(both generic and name brand) that have multiple price levels. Insurance companies limit people everyone choosing the most expensive drug by charging a co-pay. This is not the case with contraception where charging co-pay is disallowed (see KenB link above). Also there are good reasons not to simply cover the cheapest drug. Doctors and patients would be in an uproar. Usually there isn’t a one size fits all pill for each and every person. Finally, as I may have said before, an insurer cannot simultaneously cover only the cheapest drug and not stifle innovation. Innovative/novel drugs are often initially more expensive and therefore would not be eligible for a cheapest only insurer.
72 72 Ken B
February 12, 2012 at 10:43 am

@AndrewG: They are not listening. The argument seems to be, if the disincentive effect is not total and complete then it doesn’t exist at all. 100% or nothing. That is why KS can argue the presence of some innovation shows that insurance distortions have no effect (elegantly mocked by Todd).
73 73 KS
February 12, 2012 at 12:17 pm

@Todd-

“I know at least 6 people who were lifelong chain smokers and never got lung cancer. So lung cancer susceptibility is not increased by chain smoking.”

I am not sure what your occupation is, but I am guessing it is not within the scientific realm. Anyway, there is a fundamental asymmetry in science, whereby an anecdotal piece of evidence can never prove a hypothesis to be true, but one anecdotal piece of evidence can disprove a hypothesis.

I brought up those ancedotes — warfarin replacing heparin, for example — to DISPROVE the hypothesis, stated by Cochrane and garbled by Ken B (is there anything Ken B doesn’t garble?) — that providing insurance coverage for a good X makes the incentive to develop cheaper and/or better X vanish. My anecdotes can DISPROVE a hypothesis.

I would refer you to the writings of Popper for a better understanding of how science works, but I am guessing you would require a pop-up book version to even begin to understand him, and I’m not sure they’ve made that yet.

@Ken B-

“@Richard: It simply IS a fact that I consume more blood tests when I don’t pay for them — or pay only a fraction of the cost. Price affects demand curves. If you deny this you simply are not worth talking to.”

Hahaha, you chose the WORST example, because that is factually incorrect. The price of a blood test may drop from $1000 to $995, that does not necessarily mean I will choose to purchase more (translation: demand curves may be nonlinear). Furthermore, even if the blood test dropped from $1000 to $5, and I may want to order one everyday now, I will not be able to, as blood tests require prescriptions from providers (a provider is a doctor, I forgot I was talking to you and basic terms must be defined). FURTHERMORE, even if I and the provider want to purchase a blood test every day, the insurance company can deny coverage, as they constantly evaluate their in-house provider’s practices to limit costs.

Do you see why it’s better to have a deeper understanding of what’s going on, rather than simply saying “price affects demand”?

Finally — and I don’t know how many times this can be repeated — I have never said that insurance distortions have no effect. I have said you cannot predict what the effect is a priori, and you certainly cannot make the claim that presence of insurance distorts the incentive to develop new drugs SO MUCH that new drugs aren’t developed. I have provided anecdotal counter-examples of such, as well. There is really nothing more I can do to try educate you on this matter, other than a brain transplant, and the field of neurosurgery isn’t there yet.

(But then you would argue, since most neurosurgical procedures are covered by insurance, there is no incentive to develop newer or better neurosurgical procedures ever, wouldn’t you?)
74 74 KS
February 12, 2012 at 12:33 pm

After all, I can just-as-easily make the case that the presence of insurance creates a market for drugs that simply would not be there otherwise, and the creation of this market spurs the need for research into better drugs.

Not many people could afford Avastin out-of-pocket, as it costs something like $200,000 a year for treatment. With insurance though, they can. So insurance creates a market for very expensive drugs, and while most people would never need Avastin, most people would buy insurance. Hence the drug company has an incentive to spend billions to develop expensive and advanced drugs, because they know there is now a market for them where patients can actually afford them (through an insurance intermediary), as opposed to otherwise.

Ken B, I am guessing this never occurred to you. I am sure you will respond with a knee jerk, blanket “competition lowers prices” statement. I am writing this for anyone else reading. The point is simple — to state, as Cochrane did (and Ken B garbled), that the presence of insurance for a good X makes the incentive to develop better alternatives to X “vanish”, is naive and wrong.
75 75 AndrewG
February 12, 2012 at 12:39 pm

@KS

I think one problem is that your examples do not apply to the market that Cochrane is analyzing. In your example, warfarin/heparin are not completely free to the consumer at the time of purchase. If I’m not mistaken, insurance companies structure their co-pay to entice patients to choose cheaper options. In Cochrane’s example this is definitely not the case as there is no co-pay at all and the incentive to develop cheaper drugs truly does vanish. I would say that incentive to develop better drugs is reduced but does not vanish in Cochrane’s scenario and would agree with you here that Cochrane overstepped.
76 76 Richard
February 12, 2012 at 12:45 pm

It simply IS a fact that I consume more blood tests when I don’t pay for them — or pay only a fraction of the cost. Price affects demand curves. If you deny this you simply are not worth talking to.

Actually, Ken B, you’ve showed your true colors with this statement; you have little understanding of basic microeconomics other than what you may have read on this blog or Steve’s popular books. Price doesn’t affect demand curves at all, it affects quantity demanded. Shifts in demand curves are caused by something other than the price, this is taught in the very beginning of every economics 101 course. So don’t get snide with me because you made an elementary error (or simply don’t know what you’re talking about) on par with Cochrane.
77 77 KS
February 12, 2012 at 12:58 pm

@Andrew G-

“In Cochrane’s example this is definitely not the case as there is no co-pay at all and the incentive to develop cheaper drugs truly does vanish. I would say that incentive to develop better drugs is reduced but does not vanish in Cochrane’s scenario and would agree with you here that Cochrane overstepped.”

Sure, the examples I gave did not include drugs which have no copay, and the contraceptives would. I would maintain that the incentive to develop better contraceptives doesn’t vanish even in this scenario, as (1) the price of contraceptive is still reflected in the insurance company’s premium which the consumer pays anyway, and (2) the important relationship regardless is insurance company to drug company. If the insurance company can make a higher profit by marketing a cheaper drug, it will, even if the price the patient sees is zero.

@Richard — agreed! Price is NOT a demand shifter. For Ken B to make an elementary and frankly-stupid error is okay. For Cochrane to, however, is unacceptable, as his thoughts do have policy consequences.
78 78 Richard
February 12, 2012 at 1:07 pm

It is simply a reality that insurance companies do not just cover the cheapest drug. As KS noted, they often cover several(both generic and name brand) that have multiple price levels.

Andrew

Simply asserting “it is simply a reality that…” is intellectually lazy, provide proof. Furthermore, insurance companies may cover multiple drugs with different costs, but that doesn’t necessarily mean that it couldn’t be the cheapest bundle given consumer preferences. Remember, profits are at stake here.

Insurance companies limit people everyone choosing the most expensive drug by charging a co-pay. This is not the case with contraception where charging co-pay is disallowed (see KenB link above).

Yeah, Ken B actually has a reading comprehension disability. When you look further into the article…

Thomas Carroll, an analyst who covers health insurance companies for Stifel Nicolaus, said that, “in the grand scheme of things, it doesn’t seem like a material cost to be added to the managed care company or the employer.”
“Any services that are mandated are ultimately covered in the premium, either to the specific group or to the system in general,” Carroll said.

Also there are good reasons not to simply cover the cheapest drug. Doctors and patients would be in an uproar. Usually there isn’t a one size fits all pill for each and every person.

And this precludes cheapest ‘bundle’ because…?

Finally, as I may have said before, an insurer cannot simultaneously cover only the cheapest drug and not stifle innovation. Innovative/novel drugs are often initially more expensive and therefore would not be eligible for a cheapest only insurer.

If the pill is too expensive to make, even if it works better, then that’s the choice of the insurance company to make the financial call to not purchase it. Simple trade-offs exist. However, if the pill is THAT effective, what’s to say the company whose R&D developed it couldn’t sell it privately. If it did well in the market, insurance companies would pick up on this and may offer to add it to their coverage in order to attract more clients.
79 79 KS
February 12, 2012 at 1:09 pm

Finally, I must add, if Krugman have ever made an error as retarded as Cochrane did (namely, that government spending is necessarily non-stimulative because of “accounting”), SL would be all over him so hard and so often like butter on toast.

The fact is, SL repeatedly points out Krugman’s mistakes, and yes, Krugman does make mistakes. This single-minded focus on Krugman largely stems from a difference in political beliefs, however, as Krugman is nowhere near as bad as his conservative economist counterparts. The last few years have vindicated Krugman above his peers in ways I didn’t think were possible — including the predictions that the stimulus was too small and unemployment would hover above 8%, or that extreme deficit spending wouldn’t raise interest rates as we’re in a liquidity trap (it hasn’t). I think it’s time for SL to show some objectivity and give Krugman his due.
80 80 AndrewG
February 12, 2012 at 1:33 pm

@KS

I am not married to the word vanish. Do you think the incentive to develop cheaper drugs is at least reduced under a no co-pay scheme? If by how much?
81 81 KS
February 12, 2012 at 1:41 pm

@AndrewG-

No, I do not feel confident to make the claim that the incentive is even reduced under a no co-pay scheme. Certainly, one can theoretically imagine a relationship whereby incentives are reduced. Conversely, one can theoretically imagine a relationship where incentives are increased — as I did. Furthermore, there are other unknown or hidden variables at play here.

The only way to assert that insurance coverage of a good X under a no co-pay scheme reduces the incentive to develop better alternatives to X is empirically. No-one is smart enough to predict this theoretically. This is what separates me from most economists.
82 82 AndrewG
February 12, 2012 at 1:48 pm

@Richard

-Here is an example of structured co-pay covering multiple drugs by an insurance company. http://www.bcbsnc.com/content/services/formulary/rxdef.htm

-I’m not sure how you quote changes the fact that the plan would mandate that co-payment for contraception be prohibited

– My response was to your initial argument that insurance companies would only cover the cheapest drug. Now you have shifted to cheapest bundle (i.e. multiple drugs). I am glad we are getting somewhere.

– I would imagine that since most people are covered by insurance, the market is essentially dominated by that. I doubt a drug would be successful outside this market for sheer lack of prospective customers.
83 83 KS
February 12, 2012 at 1:55 pm

@Andrew G–

“- I would imagine that since most people are covered by insurance, the market is essentially dominated by that. I doubt a drug would be successful outside this market for sheer lack of prospective customers.”

Exactly! So maybe the mere presence of an insurance company intermediary produces a market for drugs. And production of this market is an incentive to develop newer drugs as well which would not exist otherwise.
84 84 AndrewG
February 12, 2012 at 2:03 pm

KS

You are right for treatments that entail high and unpredictable costs. In something like the toothpaste example, the existence of insurance would basically just fracture the existing market.
85 85 KS
February 12, 2012 at 2:12 pm

AndrewG–

Agreed. Development of a new contraceptive? That requires vast amounts of investment, and would fall into my category, def NOT a toothpaste category.
86 86 AndrewG
February 12, 2012 at 2:19 pm

Sorry I mean high and unpredictable costs to the patient
87 87 Richard
February 12, 2012 at 2:38 pm

AndrewG

-Here is an example of structured co-pay covering multiple drugs by an insurance company.

I didn’t deny that an insurance company would cover multiple drugs. I was asking for proof that it wouldn’t aim towards covering the cheapest and effective drug, or bundle of drugs.

-I’m not sure how you quote changes the fact that the plan would mandate that co-payment for contraception be prohibited

The burden of the payment would still be on the insured, not the insurer. It would be a disguised discount.

– My response was to your initial argument that insurance companies would only cover the cheapest drug. Now you have shifted to cheapest bundle (i.e. multiple drugs). I am glad we are getting somewhere.

I never denied that it would cover multiple drugs, though I did use drug in the singular sense (maybe that was a bad idea?). The logic doesn’t change, and you bringing up the fact that insurance companies cover multiple drugs doesn’t add anymore weight to your side that it stifles innovation. Maybe I should’ve used the word ‘bundle’ from the get-go?

– I would imagine that since most people are covered by insurance, the market is essentially dominated by that. I doubt a drug would be successful outside this market for sheer lack of prospective customers.

Are you denying that consumers pay out-of-pocket expenses for medical services/goods that aren’t covered by insurance?
88 88 KS
February 12, 2012 at 2:38 pm

Hmm not necessarily, I disagree. If the cost to develop a drug is high, then the price of that drug to consumers without insurance will be exorbitant. In effect, the quantity demanded then will be minimal. (Ken B, notice how I said quantity demanded, NOT demand). But with insurance, the price to consumers is a lot less, the quantity demanded a lot more, and overall revenue may increase. Thus, the incentive to spend a lot to develop newer drugs exists where it may not before.

I’m not saying this is NECESSARILY the case. I’m saying it’s a possibility, and the effect of insurance on incentives is unknowable a priori.
89 89 AndrewG
February 12, 2012 at 2:59 pm

Richard,

We can agree that insurance companies offer multiple drugs at different price levels. The reason they do this is because they can offered a structured co-payment that entices the consumer to choose the lower priced option but still have the freedom to pick the more expensive one. The mandate explicitly disallows co-payment. Yes premiums can change but there is a huge difference between premiums and co-payments. One affects the buyer’s choice at the time of purchase and the other does not. Without co-payment there is nothing limiting the buyer from just choosing the most expensive option in the “bundle”. So now the bundle has been effectively reduced to the most expensive drug that was originally in the bundle.

I definitely do not deny that people pay for items that aren’t covered by insurance..that would be nonsensical. What I was saying is that if treatments for disease A are covered by insurance then insured people will buy an insurance-covered drug for disease A purely for cost reasons. Since most people are insured, more people will buy an insurance covered drug. Therefore, the potential market for a non-insurance drug for disease A is very small. It would have to be a tremendously innovative/effective drug to overcome these cost and market size hurdles and bypass the insurance market.
90 90 AndrewG
February 12, 2012 at 3:14 pm

KS,

The price of contraception is not exorbitant. Also if usage of the drug is predictable, then insurers will just pass the cost to the premium and it will not be cheaper.

I think I agree with much of what you say, but with the caveat that it apply only to certain classes of medical treatments (chemotherapy for example). I don’t want to put words in anyone’s mouth but I suspect that Cochrane and KenB would also concur.
91 91 AndrewG
February 12, 2012 at 3:18 pm

To be clear, when I say they’d concur I mean specifically that they would agree that insurance can help to create markets for certain classes of medical treatments
92 92 Richard
February 12, 2012 at 3:46 pm

Andrew

We can agree that insurance companies offer multiple drugs at different price levels. The reason they do this is because they can offered a structured co-payment that entices the consumer to choose the lower priced option but still have the freedom to pick the more expensive one. The mandate explicitly disallows co-payment.

Then what’s to stop the insurance company from then choosing a different bundle, or even narrow the choices further? If the co-payment structure has been junked, then a profit seeking insurance company will choose a different or limited bundle of contraceptives that are cost-effective. I still don’t see why you object to this, it’s merely trying to change its variable costs.

Without co-payment there is nothing limiting the buyer from just choosing the most expensive option in the “bundle”. So now the bundle has been effectively reduced to the most expensive drug that was originally in the bundle.

That doesn’t follow. Insurance company A has to allow for the drug to be purchased under their plan by consumer X, and if this is not the case then how is it a consumer can choose the most expensive good?

Take Plan B for example. There are multiple alternative emergency contraceptive pills that use the same level of progesterone and are just as effective as Plan B, but slightly cheaper (though in my experience they’ve been pretty similarly priced for the most part.) Insurance companies will go for the latter, the only reason they would go for the former is if it somehow worked better and/or attracted more clients to make up for the cost differential.

What I was saying is that if treatments for disease A are covered by insurance then insured people will buy an insurance-covered drug for disease A purely for cost reasons.

Ah, but don’t you see what I’ve been trying to get across to you? Consumers face trade-offs, and if drug A is really that good, yet expensive, compared to its alternative drug B that is cheaper and covered by insurance, consumers will prefer that drug to the alternative as opposed to saving money (disease makes you price inelastic anyway. Not sure about contraceptive pills.) If it cannot survive on this measure, its not worth the trade-off. The only exception to this rule is if the drug cannot be acquired unless people pool resources (ie insurance), yet the demand for the drug is already there. And unless that drug is patented, price competition between drug companies for insurance deals will ensue.
93 93 KS
February 12, 2012 at 3:54 pm

@Andrew G-

“The price of contraception is not exorbitant. Also if usage of the drug is predictable, then insurers will just pass the cost to the premium and it will not be cheaper.

I think I agree with much of what you say, but with the caveat that it apply only to certain classes of medical treatments (chemotherapy for example). I don’t want to put words in anyone’s mouth but I suspect that Cochrane and KenB would also concur.”

The price of existing contraceptives is not exorbitant, yes, because it has already been developed, and a lot of it is insured.

The price of NEW contraceptives were they to be developed, in a market without insurance, would be exorbitant. I remember reading a quote that the average drug to hit the market requires about $800 million in R&D, not counting marketing.

My argument was regarding development of NEW contraceptives. Those require vast amounts of R&D. Without insurance, the price of these would be exorbitant to consumers. Insurance may make a market for these new contraceptives which wouldn’t exist otherwise.
94 94 Ken B
February 13, 2012 at 9:27 am

RIchard:
“Price doesn’t affect demand curves at all, it affects quantity demanded. Shifts in demand curves are caused by something other than the price, ”

Yes, I misspoke. I should have said that the price affects the amount *I* demand rather than *my* demand. Which is of course exactly what was happening in my hypothetical. If the price gopes up I get fewer blood tests. if the price goes down I get more blood tests.
But is does the inurer’s damnd function, which is what matters.

What I should have said to be correct is that the presence or absence and size of copays for me — the person making the choice — will shift the demand curve for the insurance company, the one actually paying for the purchase. Example. At $1 copay 10 of us want the test which is priced from the maker at $20. The insurer’s demand at $20 is 10. At $5 copay only 3 of us want the test. Now at $20 the insurer’s demand is 3. That is a change in the insurer’s demand function.

This effect of copays on the aggregate demand function is what you are denying.
95 95 Harold
February 13, 2012 at 9:36 am

“You may ask, how can an idiot be a distinguished professor in a field? Well he can’t. Unless that field is economics.”
Not true – we all display idiocy sometimes.

There is much discussion above about competition and insurance which in part perhaps gets things mixed up.

I don’t think anyone feels that individuals paying for medical treatment at the point of use is viable, so we are not talking about insurance vs, no insurance (state or otherwise). The discussion is about the effect of mandated coverage for certain treatments, not the existence or effects of insurance per se. There will be costs and benefits – these must be evaluated.

I think the factors argued for by Cochrane, Ken B et al. exist, and will reduce competition to some extent. By mandating this cover for all policies, it must remove some element of competition. This might be insignificant compared to other factors. There will clearly still be competition to get listed and/or promoted by the insurers, which may be much greater than the competition the individual exerts by selecting policies. This reduction in competition should be counted as a cost of the policy – although it may be a very small one.

Given a perfect market, there is a good argument that there is no justification to interfere. Almost by definition, the outcome of intervention will be worse – at least in the short term.

The toothpaste market is not too far away from this ideal. There is not too much justification for interfering, and we probably get about as good an outcome as we could get. (Ignoring safety and employment laws etc.)

The health provision market is about as far away from this as it is possible to get and still call it a market. Even if we had individual rather than employee policies, it is impossible for the individual to be fully informed, and we know that people make health and treatment choices on an irrational basis (for example the framing effect). Throw in the fact that most people get their insurance from employers, and that there are doctors in between patients and suppliers. Even these doctors are not fully informed, and susceptible to sales pressure (- why else would pharma spend so much on sales? You don’t really think it is just to inform the doctors do you?). We have something that bears almost no relationship to a free market. There is no good reason to think that this will get us to an ideal position.

That is not to say that we can necessarily do better, although from the principles of economics it is clearly possible to do so, if only we knew which interventions would correct some of the myriad failures in the market.

Employers choosing policies based on irrational religious grounds may be one such. Evidence that people are reluctant to do things that they are later glad they did could be another. Mandatort coverage for preventative health checks, anyone?

I do not have the evidence to hand, but it is possible that this policy will improve the general good, even given the normal laws of economics. To fall back on the economic principles of competition as though we were talking about toothpaste is to miss 9/10ths of the point.
96 96 Ken B
February 13, 2012 at 9:48 am

KS: “The point is simple — to state, as Cochrane did (and Ken B garbled), that the presence of insurance for a good X makes the incentive to develop better alternatives to X “vanish”, is naive and wrong.”

Oh for pity’s sake. NO POSTER HERE HAS ENDORSED ‘VANISH’. We have all repeatedly identified that an an overstatement.

The issue in the case at hand is much narrower than some of the fantastic hyp[otheticals floating around. Will requiring insurers to provide coverage for a common class of existing drugs, with no copays, affect the demand for those drugs, and will it affect the incentives to create new drugs with a similar purpose.

I have already given an example showing that the aggregate demand will be increased. Under any sane analysis sales of the drugs will increase more than they would with the copays or lack of coverage.

Now as for incentives, imagine a male contraceptive pill. I think it should be obvious that the deamnd for such a pill is decreased by presence of ‘free’ pills for women. I don’t say the demand will VANISH just that it will be decreased.
97 97 Ken B
February 13, 2012 at 9:59 am

@Harold: I do not agree with the last paert of your comment, but I’d like to thank you for making for the other side a coherent, thoughtful comment that does not rely on ad hominem or distortions of what I and others have actually said.

I think you are wrong because in the case at hand — the pill — we are close enough to a functioning market that the usual considerations of incentive effects do matter. I think in fact that is the PREMISE of the proposal: that more women will be on the pill more often when it costs them less.
98 98 Ken B
February 13, 2012 at 10:29 am

Harold noted one of KS’s tricks:
“You may ask, how can an idiot be a distinguished professor in a field? Well he can’t. Unless that field is economics.”
Not true – we all display idiocy sometimes.

Exactly right. We have here a pretexter. Rather than answer the argument actually made he looks for a pretext to dismiss it. Pretexters have several standard moves, all on display in this thread.

There’s the past blunder pretext. Never mind that the blunder has nothing to do with case at hand, was admitted, was corrected, or that we all blunder. How often has KS referred to a blunder Cochrane made long ago?

There is the latch onto small misstatements pretext. People often make small errors in summarizing their points, or drawing a conclusion form an example. Or they get hyperbolic. Rather thasn identify the misstatement or hyperbole as such but still grappling with the example or argument the pretexter latches on to the misstatemtn and never lets go. He ignores the substantial point. Need I say, vanish?

There is the all or nothing trick. It’s a speicies of straw man. A particular argument may apply to the case at hand but not generalize to all situations. The pretexter latches onto the generalized assertion, shows it has flaws, and then dismisses its current application as the ‘naive’ application of a bogus principle.

What all these have in common is that they are pretexts to say, I don’t have to take your argument seriously without actually addressing the argument.
99 99 KS
February 13, 2012 at 11:34 am

Ken B: “The incentive to produce cheaper versions does not vanish, but it will be sharply reduced, especially if the mandate allows patients to pick their preferred medicine.”

How exactly do you know how “sharp” the reduction is? How do you even know there is a reduction?

Anyway, Ken B, you have repeatedly made claims which others including myself have shown to be wrong (“distorting incentives always produced poor results”, “price affects demand”, “providers make drugs”, etc…) It is clear — and others have noticed this — that you have no understanding of economics, other than reading blogs and maybe a book or two. There is no further point in trying to educate you, as you are a lost cause.

This is a disclaimer for others out there — you cannot make claims a priori without evidence (Cochrane: the incentive vanishes; Ken B: the incentive is sharply reduced). To do so, you are essentially stating, “I am an idiot”. That is all I, as a scientist in a field that respects empiricism, hear.
100 100 Ken B
February 13, 2012 at 11:44 am

KS has resorted to fabricating quotes:
“distorting incentives always produced poor results”.
What I said was
“Distorted incentives produce poor results.” They do. Spilling coffee makes stains. General statements are often couched thus. Spilling coffee makes stains. Alcohol reacts with drugs. I did not assert that thety always without fail in every situation do. This is an example of KS pretexting again.

“providers make drugs” Another bogus quote. What i argued that was any exchange has a provider and a consumer. KS insists provider means doctor; my dictionary insists otherwise. Another example of pretexting.

“Price afffects demand” Actually what i said was stupider, I said price affects demand curves. In casual speech one can argue demand is a loose term for amount demanded, but with demand curves you cannot. Waht I meant — as I said earlier — was copays affect demand curves. As they do.

Oh, and my earlier querstion, how often KS has mentioned Cochrane’s old blunder? In at least 10 separate comments so far. Pretext.
101 101 KS
February 13, 2012 at 11:54 am

Let me reiterate. In the context of health insurance, nobody — and I mean NOBODY — let me repeat, NOBODY — would use the term provider to mean someone who makes drugs. Unless that person has no experience/familiarity with anything and is just garbling things they have read online, of course.

Cochrane (and you) make so many blunders it is simply not feasible for me to list and state them. In that article alone, Cochrane asserted a number of things based on his theoretical considerations, of which the fact incentive to develop new drugs would vanish is just one of them. You have attempted to support him, but you just don’t know anything (insurance companies and aggregate demand, for example). You are basically beyond hope. My concern is with people like Cochrane, whose opinions do affect policy.
102 102 KS
February 13, 2012 at 12:09 pm

@Harold–

“Not true – we all display idiocy sometimes.”

I don’t think you appreciate the magnitude of Cochrane’s blunder. It would be akin to a tenured physics professor saying Force = mass * velocity, and writing an entire blog post defending it.
103 103 Ken B
February 13, 2012 at 12:49 pm

11 and counting.

Pretext.
104 104 Ken B
February 13, 2012 at 1:03 pm

I confess I am curious about this ‘one blunder means I can ignore his arguments for the rest of time’ idea. Naturally only in economics can one make a blunder and still be taken seriously later.

I mean no Fields medal winner would suggest that the Peano Axioms are inconsistent, and no math professor would claim to have proved it.

Certainly no chemist’s reputation could survive claiming in print that DNA was a triple helix.

No-one could count 57 states and be seen as a serious contender for high office. The idea is just absurd.
105 105 AndrewG
February 13, 2012 at 1:16 pm

KenB,

Well if we are good Bayesians, we should probably reweight the amount of trust we place in someone’s analysis if they have truly made a blunder. Probably it should not go to zero especially if it is isolated. In Cochrane’s case, I am not really familiar with the “blunder” that KS is referring but the totality of his work should at least merit benefit of the doubt.
106 106 Ken B
February 13, 2012 at 1:34 pm

@AndrewG: Ahhh but you’re not looking for a pretext to ignore Cochrane. That’s not your style.

Plus of course Cochrane isn’t saying ‘trust me on this’. He’s presenting pretty routine mainstream economic arguments. Which have been explicitly endorsed by other economists like Landsburg and Boudreaux. And which I’m told are so obvious they are trivial.
107 107 KS
February 13, 2012 at 1:53 pm

“Certainly no chemist’s reputation could survive claiming in print that DNA was a triple helix.”

Actually the evidence suggested a triple helix before high resolution X-ray crystallography came out. Evidence changes paradigm, that is the key to science. In economics, evidence has a very hard time changing paradigm, as people mix political beliefs with economic evidence (as Cochrane did, because he believes the government ought not to be spending people’s money, he made the confusion that government spending isn’t stimulative).

“I confess I am curious about this ‘one blunder means I can ignore his arguments for the rest of time’ idea.”

Once again, you are mis-characterizing me. I have never said Cochrane made one stupid mistake ergo I can’t believe anything he says. In fact, I have provided a point-by-point refutation of another claim he made, that the incentive to develop new contraceptives will vanish if they are providing insurance in a no copay scheme. I have also refuted you, although that was easier, as you tend to say stuff which is logically-inconsistent and clearly ill-defined, so I really just need to repeat back to you what you’re saying.

However, I have pointed out just how stupid the mistake Cochrane made was. And it wasn’t a blunder. It was an entire blog post arguing against deficit spending in a recession.
108 108 Ken B
February 13, 2012 at 2:00 pm

Again with the ‘vanish’.
109 109 KS
February 13, 2012 at 2:03 pm

Once again, Cochrane’s exact words were “vanish”. You garbled something else
110 110 Ken B
February 13, 2012 at 2:20 pm

Oh I get it now. I agree with KS that Cochrane’s claim — which I read as hyperbole btw — is too strong, so I make a weaker claim — that it will diminish not vanish. That is ‘garbling’.

While we’re on that, KS demanded evidence on the issue that the price affects the consumption for the pill. Never mind that that is a premise of the proposal. (Accepting your adversarie’s assumption; deuced unfair.) Never mind it’s common sense. Never mind the entire history of contraceptive sales. Never mind that Steve provided on this thread a link to a study showing just exactly that effect for college women.
111 111 Ken B
February 13, 2012 at 2:38 pm

David Henderson weighs in http://econlog.econlib.org/archives/2012/02/how_to_increase_1.html#comments

I didn’t check. Did David ever make a blunder? That would disqualify him from commenting of course. Especially if the blunder was in a blog post.
112 112 KS
February 13, 2012 at 2:57 pm

“While we’re on that, KS demanded evidence on the issue that the price affects the consumption for the pill. ”

Nope, I contested your argument that a decrease in price necessarily leads to an increase in price sensitivity. That is not the same as saying “price affects consumption”.

Yet again, all you do is make blanket statements with vague terms you clearly have an imprecise understanding of.
113 113 CO
February 13, 2012 at 3:00 pm

@Ken B–

I don’t think KS ever said that if contraceptives get cheaper women won’t consume less. I wouldn’t be as harsh as KS is being to you, but you are misrepresenting what he’s saying.

Also, I think all KS is trying to say is that you can’t make the claim that incentives will even reduce without some evidence.
114 114 KS
February 13, 2012 at 3:03 pm

“blunder?”

It depends what you mean by ‘blunder’. Usually a blunder is a slip-up, a quick and easy forgettable mistake. A physics professor in class saying “F = m * v” and then correcting himself would be a blunder. A physics professor writing an entire blog post arguing how F = m * v? That’s not a blunder. That’s idiocy. It’s something you + Cochrane seem to share.
115 115 Ken B
February 13, 2012 at 3:33 pm

Someone I imagine is a doctor said this:
“In medicine, we only apply treatments if they have been shown to be effective at doing what they’re intended to more than 95% of the time.”

Now this is not true. Doctors presciber prescriptions or treatments that are effective with less than 95% of patients. This is so obvioulsy incorrect, so clearly confuses success rate and statistical significance that some might call it idiotic. I think though it was just a goof in a blog, and that the actual sensible assertion underlying this false claim can be recovered. Not being a pretexter, I did that.
116 116 KS
February 13, 2012 at 3:41 pm

@Ken B–

Once again, you are confusing two statements.

Me: In medicine, we only apply treatments if they have been shown to be effective at doing what they’re intended to do more than 95% of the time.

You: Doctors prescribe prescriptions or treatments that are effective with less than 95% of patients.

My statement had nothing to do with the percentage of patients who are successfully treated by medicine. It had to do with the percentage of times the treatment exhibits intended efficacy. The true are different. It is you who in fact confused success rate (in terms of proportion of patients) with statistical significance.

Once again, you exhibit a poor understanding of a basic scientific relationship, then make a confused and garbled statement with confidence, and all I have to do is repeat it back to you and highlight the obvious idiocy…
117 117 Ken B
February 13, 2012 at 4:09 pm

KS, “It had to do with the percentage of times the treatment exhibits intended efficacy.” One can prove, with 95% confidence, that X has an effect without proving X has an effect 95% of the time. A low probability effect can be established with high confidence.
118 118 KS
February 13, 2012 at 4:18 pm

“One can prove, with 95% confidence, that X has an effect without proving X has an effect 95% of the time.”

I could go into a long explanation involving probabilistic schools of thought, but I figure that first you would need to learn addition. I’m sorry the public education system failed you so badly. At this point you have confused/mistaken so many statements all hope is lost.
119 119 AndrewG
February 13, 2012 at 4:54 pm

KS

I assume this 95% discussion comes from studies comparing a drug and a placebo and having a p-value of less than 0.05 for rejecting the null hypothesis that drug = placebo. This is different from both a success rate of 95% and saying “shown to be effective at doing what they’re intended to more than 95% of the time”. First of all you have to define what they are intended to do. Are they intended to be different than placebo? If that is the case, then (1-p) is not the probability of this.
Economists chart the outcomes of policies based on laws of supply and demand. These do not come from nowhere..there have been numerous lab and field studies that say with p<0.05 that they accurately describe human behavior in the marketplace. So I would say that in this sense, both economists and doctors use tools that have been statistically validated to predict outcomes.
120 120 KS
February 13, 2012 at 5:22 pm

@Andrew G–

“This is different from both a success rate of 95% and saying “shown to be effective at doing what they’re intended to more than 95% of the time”. First of all you have to define what they are intended to do.”

Medical studies examining the effectiveness of treatments first clearly define an outcome variable. Say the treatment is a new blood pressure-lowering medication. Then the outcome variable might be “mean BP”.

Once the outcome is identified, the intervention (new medication) is applied to a group. In addition, a control group is made, ideally of similar characteristics to the first group, but with no intervention. (Sometimes this can mean doing nothing, sometimes this can mean applying the standard of care only).

Now the intervention (medication) is applied, and the outcome is measured. Let’s say in this case the data returns a p-value of 0.03.

This is equivalent to saying, there is a 97% chance that the observed reduction in mean BP is because of the new medication, or a 3% chance it is due to random chance. This is also equivalent to saying, if we apply the new medication over and over and over, 97% of the time we will see a reduction in mean BP of this magnitude.

This is what I meant when I said, “In medicine, we only apply treatments if they have been shown to be effective at doing what they’re intended to do more than 95% of the time.” It’s the same statement! To be considered effective, treatments have to exhibit a p-value less than 0.05, which is the same as saying there is at least a 95% chance that they produce the intended outcome, which is the same as saying in 100 trials they will produce the intended outcome at least 95 times.

NOTHING in this discussion as anything to do with, as Ken B put it, “Doctors presciber prescriptions or treatments that are effective with less than 95% of patients. ” Unless you happen to define the outcome variable as the percentage of patients in which the treatment works, which of course nobody does, as you still need to define what “works” is.

Trust me, I run clinical trials, I know what I’m talking about. Unlike Ken B.
121 121 KS
February 13, 2012 at 5:28 pm

Sorry that post is long, here is a brief synopsis. The following statements are equivalent:

(1) In this clinical trial, there is a 95% chance that this medication X resulted in Y.
(2) If we were to apply X repeatedly, 95% of the time it will result in Y.

My phrasing was similar to #2; you may hear it as #1. The two are the same.

– KS

P.S. Note that the following statement:

(3) The medication X works in 95% of patients,

is neither equivalent to 1 nor 2. Only an idiot would think so.
122 122 AndrewG
February 13, 2012 at 5:47 pm

KS,

Based on my understanding I would think if you did repeat experiments at p = 0.03 97% of the time you will see a reduction in BP of that magnitude OR SMALLER.
123 123 KS
February 13, 2012 at 5:52 pm

AndrewG, that is correct; my statement (“of this magnitude”) was vague. The chance of producing a reduction of the same exact magnitude would be very small.
124 124 AndrewG
February 13, 2012 at 8:07 pm

KS,

Well the problem is that (1-p) and smaller is also very vague and has little inferential meaning. I think that’s what KenB was getting at.
125 125 KS
February 13, 2012 at 8:31 pm

Not sure what you mean?

And I have no idea what KenB gets at. I’m not sure he does either. His arguments are regurgitated blanket statement vomit, basically
126 126 Richard
February 13, 2012 at 8:57 pm

Ken B

This effect of copays on the aggregate demand function is what you are denying.

No I didn’t. I said these ‘exact’ words…

No, first it depends on who decides what X can be purchased before the number of units.

Meaning if there are two hypothetical blood tests, A & B, and insurance company X decides it will only cover blood test A, then quantity demanded for blood test A for those consumers with insurance will go up. B will not. In order for clinics that provide only blood test B to get business, they must either create an innovative new bloodtest or reduce their price in order to get deals from insurance companies.
127 127 Richard
February 13, 2012 at 8:58 pm

VANISH!
128 128 KS
February 13, 2012 at 9:00 pm

@Richard–

There’s no point. Ken B misinterprets statements, regurgitates them back as something they’re not, and then attacks that on suspect grounds as well. He’d make a great spokesperson for the tobacco industry!
129 129 AndrewG
February 13, 2012 at 9:33 pm

KS,
From the previous example, we expect in 97% of the replicated studies you will have a magnitude change smaller than or equal to the original one. What does this mean? You could have an overall average effect from the replicated studies that is large, small, zero who knows. The strength of evidence of a p = 0.03 study is less than the number 0.03 would have you believe.
Richard,
Sorry to get in this whole ‘demand’ question again, but doesn’t the demand (not quantity demanded) increase for blood test A?
130 130 iceman
February 14, 2012 at 12:23 pm

This exchange has been hilarious…thanks guys.

A late observation — it seems part of the disconnect is over where the burden of proof should lie: with the party making a seemingly unorthodox claim, or the party critiquing it as such. Cochrane is basically saying, “HHS’ claims are at odds with some pretty (*empirically*) well-established, general principles of economics, and they offer no evidence or even consistent arguments for why this should be.” People who support the policy (+/or don’t like Cochrane) respond, “since those principles don’t apply in every conceivable situation, Cochrane must specifically disprove the claims or he is being arrogant and sloppy.” To use Todd’s analogy, I bet most physicians would dispute outright an unsubstantiated claim that smoking has no linkage to cancer, without feeling obligated to conduct an(other) exhaustive study on the matter.

At a minimum, the uncertainty argument should cut both ways; if we can’t predict the future, why give the HHS proposed *policy change* the benefit of the doubt (or even speculate on their behalf)?

This may all flow from another disconnect over the appropriate standard of proof. Economics simply does not have the same ability to do controlled experiments as a field like medicine. [This is why Krugman says in the preface to Peddling Prosperity, “economics is *harder* than physics; luckily it is not as hard as sociology”.] Obviously we’re dealing with patterns or tendencies in human behavior, not physical laws. This doesn’t mean that it isn’t necessary (or that it’s “arrogant”) to try to make predictions about important issues as best we can, even if some ‘absolute’ or 95% standard isn’t feasible. (Would Keynes have met such an a priori test?). And again, whatever standard we pick has to apply symmetrically to HHS’ proposal as well. In particular, the ‘one anecdotal piece of evidence can disprove a hypothesis’ is less applicable. E.g. the real issue is how many drugs are NOT developed when we impact incentives? That’s unknowable in a rigorous sense, but critically important.
131 131 KS
February 14, 2012 at 3:10 pm

@iceman–
“People who support the policy (+/or don’t like Cochrane) respond, “since those principles don’t apply in every conceivable situation, Cochrane must specifically disprove the claims or he is being arrogant and sloppy.”

Nope, this is wrong. You’re assuming all of Cochrane’s claims are empirically-supported, when in fact they’re not.

“Economics simply does not have the same ability to do controlled experiments as a field like medicine.”

Agreed. Nonetheless, in ANY science, one cannot repeatedly assert claims based on purely theoretical grounds without any evidence. If you can’t find a way to get evidence for your claim, then shut up and stop asserting it. This is the fool’s defense of economics.

“This doesn’t mean that it isn’t necessary (or that it’s “arrogant”) to try to make predictions about important issues as best we can, even if some ‘absolute’ or 95% standard isn’t feasible. (Would Keynes have met such an a priori test?).”

I refer you to The Black Swan by NNT.

“In particular, the ‘one anecdotal piece of evidence can disprove a hypothesis’ is less applicable.”

Nope, that’s wrong x2. Cochrane’s claim (paraphrased): there will be no incentive to develop new contraceptives if existing ones are covered for ‘free’ by insurance. A single counter-example automatically disproves this hypothesis.

” E.g. the real issue is how many drugs are NOT developed when we impact incentives? That’s unknowable in a rigorous sense, but critically important.”

You reveal your narrow-mindedness. You have already assumed some drugs will NOT be developed once incentives are impacted. You are assuming your hypothesis from the getgo; that is simply bad science.
132 132 iceman
February 14, 2012 at 6:00 pm

@KS: I guess the keyword here is going to be “symmetry”. Again, why do your demands for empirical evidence not apply with (at least) equal force to HHS’ proposed policy changes? Do they get a pass because they don’t even attempt, for example, to reconcile ‘free’ mandates with their purported over-arching reform goal of reducing costs? I.e. is one only a “fool” if one actually tries to make a defense? (I note, as someone else here mentioned, one thing that is clearly an empirical issue but is frequently just asserted is the notion that increased preventive care lowers overall costs.)

I also thought we had moved past the hyperbole of “no” incentives or “vanishing” entirely; a single counter-example does not disprove a concern that there will be *less* of an incentive. And the idea that if there is a reduced incentive to do something, people will respond *in kind* (not just randomly) and vice-versa, is perhaps the most basic and important principle of economics (far beyond the purely ivory-tower “flowchart” stage). To be sure, sometimes there are anomalies +/or counter-intuitive truths; we can reasonably require a compelling reason to believe this is such a case. And again, how are you not simply assuming there will be *no* negative impact on drug development? It’s not an issue of narrow-mindedness; we’re forced to speculate on both sides about something – a critically important issue — that can be framed as ‘proving a negative’ either way. So there is a role for educated logic. Different “sciences” have different tools at their disposal.

Look, the bottom line is that if your basic point about acting under uncertainty is that economists and policymakers should abide by some sort of Hippocratic oath, I agree; let’s be amply cautious about messing around with things. But you seem to be holding the HHS proposal (and maybe health care reform in general) to a different standard, and I don’t understand why.

P.S. On the standard of proof, I’m talking about estimating expected values or central tendencies, for which (in my view) it’s simply not possible to hold macroeconomic forecasts etc. to the same ‘confidence interval’ standard of, say, a clinical trial of a new drug. References to black swans have to do with risk analysis i.e. the distribution around those expected values, which is important as well.
133 133 KS
February 14, 2012 at 7:22 pm

“Again, why do your demands for empirical evidence not apply with (at least) equal force to HHS’ proposed policy changes? Do they get a pass because they don’t even attempt, for example, to reconcile ‘free’ mandates with their purported over-arching reform goal of reducing costs? I.e. is one only a “fool” if one actually tries to make a defense?”

I never said I agreed with the HHS policy decisions. I am critiquing Cochrane’s critique of it. No-one here actually knows my stance on these policies. Yes, HHS’ proposed policy changes would require empirical evidence as support as well.

“I also thought we had moved past the hyperbole of “no” incentives or “vanishing” entirely; a single counter-example does not disprove a concern that there will be *less* of an incentive.”

Please look up the definition of the word “vanish”. I repeat: I am critiquing Cochrane’s editorial. I repeat again: I am critiquing Cochrane’s editorial. You can certainly claim he shouldn’t have been so melodramatic, and I would agree.

“And the idea that if there is a reduced incentive to do something, people will respond *in kind* (not just randomly) and vice-versa, is perhaps the most basic and important principle of economics (far beyond the purely ivory-tower “flowchart” stage).”

Once again, you have already assumed the incentive is reduced. Perhaps the existing of an insurance intermediary creates a market that increases the incentive? Perhaps none of us is smart enough to know what the effect on incentives really is a priori?

“And again, how are you not simply assuming there will be *no* negative impact on drug development?”

Did I ever say that? All I said was me nor you nor Cochrane (and certainly not the moron Ken B) can make that claim a priori.

“But you seem to be holding the HHS proposal (and maybe health care reform in general) to a different standard, and I don’t understand why.”

Nope, I’m not. If HHS were to assert preventative medicine in the form of free contraception reduces costs, I would also retort, how do you know this? Please keep in mind this blog post is focusing solely on Cochrane’s editorial.

“References to black swans have to do with risk analysis i.e. the distribution around those expected values, which is important as well.”

An important lesson in NNT’s book is that some information is often worse than no information, as it provides the confidence to affect policy which wouldn’t be otherwise. Case example: financial models had assumed Gaussian distributions in their risk assessments, which means, a drop of 1,000 in the stock market may happen once every 100,000 years. They used these assessments to make speculative and risky investments (“some information”). The end result was disastrous. Would they have made such investments had they not been provided the false confidence by models which hadn’t been empirically verified? No.
134 134 Ken B
February 15, 2012 at 9:19 am

KS has spent 300+ words explaining his initial imprecise reamrks on statistical significance. Let me repeat my point. It was not to assert that KS does not understand the concept. It is was that his phrasing was not actually crisp, clear, and unambiguously correct.
As informal rhetoric in a quick blog comment though that is not unusual. One can read the comment and understand the point or instead one can obsess about the wording, and pretend it discredits every other point he makes. I said we should not do the latter, but is is exactly what KS does with VANISH (or provider).

Nice to see KS acknowledge he has not been discussing the merits of the arguments, just harping on Cochrane’s bit of obvious hyperbole ‘vanish’. Which is obvious.

Nice also to see KS has expanded on his vocabulary. He seemed stuck on various forms of ‘idiot’ for a while. Speaking of vocabulary … vanish actully means to become invisible. That is not quite the same as to cease to exist.
135 135 KS
February 15, 2012 at 2:18 pm

“KS has spent 300+ words explaining his initial imprecise reamrks on statistical significance. Let me repeat my point. It was not to assert that KS does not understand the concept. It is was that his phrasing was not actually crisp, clear, and unambiguously correct.”

That’s fair.

Now, a professional economist arguing that government spending is necessarily non-stimulative because of “accounting” is retarded. This is not an issue of “phrasing” or things not being “crisp, clear, and unambiguously correct”. This is an issue of saying something retarded, then defending something retarded, then continuing to be retarded.

It is something you + Cochrane share in common. A LOT.

– KS

P.S. And yes, you are an idiot.
136 136 Ken B
February 15, 2012 at 2:33 pm

“retarded”

Good, good. Keep up those vocabulary lessons.
137 137 KS
February 15, 2012 at 2:43 pm

” One can read the comment and understand the point or instead one can obsess about the wording, and pretend it discredits every other point he makes. I said we should not do the latter, but is is exactly what KS does with VANISH (or provider).”

Agreed.

But there are people out there — let’s call them “Kens” — who commit both mistakes: (1) improper and ambiguous wording, and (2) a complete misunderstanding of even basic principles resulting in statements which are wrong and stupid, no matter how you might word them.

Ken, I’m going to give you a hint. You’re one of the “Kens”.
138 138 iceman
February 15, 2012 at 4:07 pm

I’ll try to be as precise as I can:

To me the problem is that sometimes we have to make choices one way or another, under uncertainty — like when HHS announces a policy change, now, unencumbered by data. Do we support it (which includes doing nothing) or oppose it? How is one response (a priori) more enlightened or arrogant or narrow-minded than the other? What if this involves trying to prove a negative or imagine what might have been? E.g., how exactly would we test (practically and even ethically) for whether more / better drugs would be developed (for billions of dollars over long horizons) with or without a ‘free’ mandate? This is where we may come to see some value after all in appealing to well-established economic principles to consider the expected impact on incentives, at least as a starting point. And while we may ultimately conclude this is one of those interesting cases where price impacts supply or demand in an unexpected way, why shouldn’t the burden of proof tilt toward HHS to make that case? (Of course we may also conclude that their real goals are merely political.) I’ll try the analogy again: if someone proposed increasing access to free cigarettes based on an unsubstantiated claim that under certain circumstances smoking did not increase cancer risk (and perhaps even had some benefits like reducing stress), would physicians really feel obligated to conduct a new study before voicing skepticism or criticism?

BTW at this point I think everyone else is more interested in discussing this in more meaningful terms than the all-or-nothings of “vanish”. I repeat: everyone else is more interested in discussing this in more meaningful terms than the all-or-nothings of “vanish”. I think we all agree that choice of term has been a distraction from the issues of substance.
139 139 KS
February 15, 2012 at 5:15 pm

“BTW at this point I think everyone else is more interested in discussing this in more meaningful terms than the all-or-nothings of “vanish”. I repeat: everyone else is more interested in discussing this in more meaningful terms than the all-or-nothings of “vanish”. I think we all agree that choice of term has been a distraction from the issues of substance.”

Let me try make it simple for you.

(1) Cochrane used the term ‘vanish’. If you have a problem with that term, your comment should be ‘Cochrane is an idiot’, or something of that sort.

(2) Even if Cochrane were to assert the incentive is reduced, that would be wrong. Nobody can predict that a priori.

Either way, Cochrane is a partisan hack, while Ken B is a moron who tries to be a partisan hack but comes off as a partisan idiot.
140 140 Greg MN
February 15, 2012 at 8:37 pm

@iceman

I completely agree that making decisions under uncertainty is difficult and therefore it is difficult to decide whether one should support it or oppose it. I think each person has to do some research, some soul searching, and some civilized debate is helpful. Ultimately, each person decides whether to support, oppose, or abstain. With whom the burden of proof lies is also a wise element to consider. In this case, HHS is given the responsibility to implement legislation and they should do so in a manner consistent with their conscience and wisdom.

I’m willing to defer to well established economic principles, but I’m not certain which principles we’re talking about. My impression is very credentialed economists continue to disagree on what caused the depression, what made it better, and what made it worse and whether the current Obama policies are making things better or worse. The contraception issue is even more difficult because it involves emotional and social elements. Proponents likely feel the proposed policy will reduce the number of unwanted pregnancies and the number of abortions. Opponents likely feel the policy creates inefficiencies and disincentives. How does one compare the potential to reduce unwanted pregnancies and abortions to creating potential inefficiencies and disincentives in the contraceptive industry? I certainly don’t know.

No, I don’t think physicians would feel obligated to conduct another study due to an unsubstantiated claim disputing the relationship between smoking and cancer. I don’t think this applies quite right though because I don’t think the opponents are disputing the proposed policy won’t reduce unwanted pregnancies and abortions, are they?
141 141 KS
February 15, 2012 at 9:01 pm

“No, I don’t think physicians would feel obligated to conduct another study due to an unsubstantiated claim disputing the relationship between smoking and cancer.”

The evidence linking smoking to lung cancer is vast, and in fact, the golden example of epidemiology.

An unsubstantiated claim facing a body of high-quality evidence that massive deserves no peer review.
142 142 iceman
February 16, 2012 at 12:31 pm

@GregMN: I appreciate the civil tone, refreshing and much more productive.

“With whom the burden of proof lies is also a wise element to consider. In this case, HHS is given the responsibility to implement legislation and they should do so in a manner consistent with their conscience and wisdom.”

This statement seems to hold HHS to no standards whatsoever.

“I’m not certain which principles we’re talking about”

Things like the expected impacts of artificially lowering a price on consumers and suppliers, based on an accumulated body of knowledge through both theory and analysis. I’m not sure if this is quite as “vast” or “massive” – *in relative terms* — as the body of research on smoking-related cancer in medicine, but to me it certainly warrants a position as the starting point from which anomalies must be explained.

“economists continue to disagree on what caused the depression”

Agreed, and yet here we are 80 years later, faced with decisions about whether and how we should respond to a recession with stimulus. A good example of how the level of uncertainty does not preclude us from the necessity of making choices.

@SL: “Even if Cochrane were to assert the incentive is reduced, that would be wrong. Nobody can predict that a priori.”
Of course the assertion can’t be both “wrong” and unpredictable. But at least now we’re talking about something between 0 and 100% which is the relevant and interesting scenario (perhaps “vanish” was never intended to be taken entirely literally?). Anyway, again the level of uncertainty doesn’t preclude us from having to make a decision one way or the other. The fact that this is often simply *harder* (Krugman’s word) in economics than other physical sciences doesn’t make the field “stupider”. The point of the smoking analogy was that one’s starting point is usually based on the preponderance of evidence, from which anomalies require new evidence.

@All:

I note that this whole discussion has largely overlooked Cochrane’s main point (which was right in the headline) that insurance is generally not used to cover relatively “small, recurring and predictable” expenses. I’d sincerely like to understand the rationale for why this is (should be) so different for health care. (Submissions that use the word “vanish” will not be considered.)
143 143 AndrewG
February 16, 2012 at 1:08 pm

@GregMN
You bring up the fact that economists do not agree what caused the Depression. This is true but probably not directly applicable to analyzing the consequences of the contraception mandate. The depression was a complex macroeconomic event while the mandate is largely confined to microeconomics for which forecasting/understanding is easier. Also the issue does not necessarily need to be about weighing benefits of reduction in unwanted pregnancies vs disincentive costs of the mandate. Are there better ways to reduce unwanted pregnancies that avoid the disincentive costs? Subsidies are one possible option.
@iceman
This a good point and is not different for healthcare insurance in general (ie voluntary) as the huge adverse selection problems would still be there. The reason it is a bit different with regard to the HHS proposal is that it mandates coverage and thus basically taxes people who don’t normally use birth control to pay for contraceptives for those who do (as noted by Cochrane). The main rationale, though not necessarily the only one, is that this will reduce unwanted pregnancies and therefore limit externality costs.
144 144 iceman
February 16, 2012 at 5:27 pm

@AndrewG:

Sure. But we still might want some *empirical* evidence that 1) more kids do in fact impose net negative externalities (some on these pages have suggested it may be positive); and 2) the policy will actually reduce unwanted pregnancies. (Uh oh, how do we test for that?) As I recall, economists have debated (e.g. via Freakonomics) whether legalized abortion has reduced or increased (let’s call it) ‘initially unplanned’ births. The counter is that reducing the perceived cost / risk of sex leads to more pregnancies (along with other attendant risks), and sometimes people end up changing their minds. Here the analogue might be when birth control is free, people are less careful. Something like “How Seatbelts Kill.”
145 145 AndrewG
February 16, 2012 at 5:55 pm

iceman,

I agree. I was just starting the theoretical rationale for it. As for #2 SL posted a paper somewhere back up in this comment section that if my memory serves correctly supports your claim. Something like a price doubling of birth control pills did not increase unplanned pregnancies partly bc people chose to be more careful.
146 146 Anshu
February 19, 2012 at 7:55 pm

@KS,

I just wanted to point out that you cannot draw inferences about economics as a field from the way op-ed articles are written by one particular economist or by economists in general. If you were to look at an A journal empirical article in economics and finance then you will see that economists typically employ a battery of empirical tests to disprove any empirical statement they eventually make. (Theory articles are also super tight, careful and compact in their scope) A typical article contains a lot less certainty, a lot more qualification and a lot less content than an op-ed article. A typical introduction to a typical paper takes three to five pages to carefully describe the lay of the land, with sources, and then make one qualified and careful statement without proof or evidence(which of course is what the body of the paper contains) but with intuition underlying the statement and some summary statistics whereas a typical op-ed, even those written by non-economists, makes about 5-10 assertions, strings them together to make a coherent story explaining some real life phenomenon. If economists used professional rigor in their op-eds then it would take a hundred pages of qualified and undramatic statements that would lose most readers. An op-ed by design is a necessary oversimplification where the best you can do is make a series of assertions in 700 words. Cocharane could’ve written a more empirics based op-ed and cited relevant studies and natural experiments which have tried to quantify the price elasticities and other behavior in insurance markets. Not every op-ed on a topic has to cite empirical support. On the other hand if he were writing a theoretical paper on the matter for a peer reviewed journal then he would certainly have presented a thorough literature review of both the theoretical background literature and also of the empirical literature surrounding the theoretical/testable claims he is making.