« Incomes and inequality: what the numbers don't tell us | Main | Markets in everything, deficit spending edition »
1000 Word Rebuttal
The tax code, [Bush] said, “unwisely encourages workers to choose overly expensive, gold-plated plans. The result is that insurance premiums rise, and many Americans cannot afford the coverage they need.”
Again, wow. No economic analysis I’m aware of says that when Peter chooses a good health plan, he raises Paul’s premiums.
Rebuttal:
Posted by Alex Tabarrok on January 25, 2007 at 07:25 AM in Economics, Medicine | Permalink
Comments
BUUURRRRN!
Tou-frickin-che'
Posted by: Aaron at Jan 25, 2007 8:49:53 AM
Alex:
Can you please explain the assumption that the tax code "encourages" employees to select gold-plated health plans? Do employees, in fact, select "overly-expensive" plans? And by who's lights are the plans "overly expensive?" Thanks.
Dan
Posted by: Dan Cole at Jan 25, 2007 9:25:04 AM
I don't think this counts as a burn. Krugman already clarified that he was assuming that the long-run supply curve would be flat (therefore, no externality). In fact, Krugman's clarification is included on THE EXACT SAME blog that Alex is quoting (economist's view).
http://economistsview.typepad.com/economistsview/2007/01/what_pecuniary_.html
I seriously hope Alex had just over looked this post and isn't just looking for a new way to burn Krugman for holding slightly less libertarian views than him.
Posted by: Student at Jan 25, 2007 9:29:04 AM
Wow. Maybe Krugman thinks health insurance supply is extremely elastic? (shrug)
He goes on: "Will all those who think they have 'gold plated' health coverage please raise their hands?" I guess that's me. My health plan covers regular visits to the chiropractor, the kids' regular visits to the pediatrician, visits to the dentist for cleanings. It's not that these are unnecessary. But to the degree that they're utterly routine, predictable costs, this plan isn't insurance. I don't know what to call it. The insurance company calls it "profit".
Posted by: Jason at Jan 25, 2007 9:30:22 AM
I'm glad I wasn't the only one who caught this glaring misrepresentation. Unfortunately Krugman continues down the path towards disingenuous political hack.
Posted by: Taggert J. Brooks at Jan 25, 2007 9:34:00 AM
Dan,
I think the term gold-plated is slightly misleading. My reading of things is that gold-plated doesn't mean necessarily that my insurance covers things like pedicures (aside: you should see how many economists I ran into at the AEA meetings that were getting manicures and pedicures) but rather that they regress to offering first-dollar coverage, low-premiums, low copays and the like.
I also think it is misleading to suggest that the tax code encourages employEES to select these types of plans - I do not think that employees have these options. The right way to frame it is that the tax code encourages employERS to offer these types of plans. At the last three places I've worked, it seems that I had a choice between two plans that only marginally differed, but that I think Bush would consider gold-plated.
Why does the tax code encourage employers to offer these sorts of plans? Two reasons I suggest. The obvious one is the asymmetry in the tax code - which makes a $1 of employer-provided insurance coverage worth more than $1 increase in income. That in and of itself should not be sufficient to cause the "problem" of gold-plated coverage. Second is that most of the costs of the employer provided system are hidden to the employee - and I do not mean insulation here. I have never seen a W-2 or paystub where it shows me the dollar contribution to the total premium that my company is making.
-Mike
Posted by: Mike at Jan 25, 2007 9:38:06 AM
I just can't get past the fact that a government official is deciding that some plans are "overly expensive". Especially when state and federal governments mandate minimum requirements for plans including generous mental health benefits, chiropractors, etc., and the requirement to practice defensive medicine that the courts effectively impose. Do the tax incentives make a bigger difference than government mandates?
Posted by: DK at Jan 25, 2007 9:39:13 AM
Dan,
The standard account of a subsidy is that it encourages over-consumption - over-consumption in the sense that the marginal units consumed are valued at less than their cost.
Do note that the issue addressed in the post is not "overly expensive" but Krguman's statement that "No economic analysis I’m aware of says that when Peter chooses a good health plan, he raises Paul’s premiums." That statement is doubly false in that a) economic analysis does show that subsidizing Peter can raise Paul's premiums and b) Krugman is surely aware of this analysis!
Best
Alex
Posted by: Alex Tabarrok at Jan 25, 2007 9:49:53 AM
Krugman clarifies: "Aha - I was wondering if anyone would raise that. I was taking it as true to a pretty good approximation that the long-run supply curve for medical services is horizontal. Unless you think that there's permanently limited supply of medical education, or something, why should we think otherwise?"
Hmm. I thought the usual assumption was that supply curves slope up somewhat. I'm probably just confused.
Krugman is saying that in the long run, there's no scarcity of doctors, nurses, hospital rooms, etc. That's counterintuitive to me. I thought potential doctors were rare.
Posted by: Jason at Jan 25, 2007 10:15:40 AM
Jason,
"Wow. Maybe Krugman thinks health insurance supply is extremely elastic?"
Yes, he does. And he said so in an e-mail that is posted on Mark Thoma's blog.
http://economistsview.typepad.com/economistsview/2007/01/what_pecuniary_.html
Posted by: Student at Jan 25, 2007 10:17:02 AM
I was thinking a bit more about this. If Krugman belies the long-run supply of health care is perfectly elastic, then isn't he therefore predicting a pretty steep decline (or at least leveling off) of health care costs in the medium term, given the steep recent rise in health care costs we've seen recently?
After all, if the long-run supply curve is perfectly elastic, then prices shouldn't rise over the long run...
Posted by: Keith at Jan 25, 2007 10:28:18 AM
Yep, Krugman's really wrong on this one. You can't talk about the problems of rising health care costs one day and then say you believe that the long-run supply curve of health care is elastic. At least not if you want to be credible.
Posted by: Keith at Jan 25, 2007 10:37:46 AM
I think 'gold plated' here simply means expensive as in more than $15,000 a year. But how do you know how expensive your plan is unless your employer shares with you his cost as well as yours or if you're buying it yourself?
Anyway, here's a point for Krugman.... If insurance companies have two products...gold plated and standard why wouldn't the subsidy for 'gold plated' simply serve to push up the price of gold plated plans? If the gov't subsidizes Paul's shopping at Nordstrom I'm not sure why that would cause Wal-Mart's prices to go up....they may in fact go down!
Posted by: Boonton at Jan 25, 2007 10:40:49 AM
I would have said "gold-plated" means Congress' plan - maybe even the government labor unions' plans, but on reflection, at least Congress' is "platinum-plated."
Once again the stupid party doesn't do anything, let's compare the working man's policy against Congress'.
I see an opportunity for government cost-cutting.
Posted by: Sandy P at Jan 25, 2007 10:47:56 AM
Boonton says:
If the gov't subsidizes Paul's shopping at Nordstrom I'm not sure why that would cause Wal-Mart's prices to go up....they may in fact go down!
Ah, but don't many of the people who believe in a crisis in health care think that it's much more difficult for consumers to know the difference in price and quality among different health care choices than it is to know the difference between Nordstrom and Wal-Mart?
Sure it's possible that hospitals will somehow charge higher prices to only people with good insurance plans. I think, however, that in reality we actually see the opposite-- insurers seem to use their market power to negotiate lower prices than those charged for people without insurance, by using the threat of diverting the patients' business elsewhere.
There are more complicated arguments that overconsumption of expensive experimental medicine will lead to greater research and in the long run benefit. I'm not sure why we necessarily want to subsidize the wealthy in doing that, though.
I'm always a little surprised when people seem to argue that subsidies cause higher prices and overconsumption in one area, say Medicare prescription drug benefits, and deny it that it's even possible in another, such as health insurance or college tuition costs. I can understand arguments that the long term supply curve is flat and all, though that would seem to argue against a long term crisis.
Posted by: John Thacker at Jan 25, 2007 11:09:54 AM
I do not agre with your chart. Your representtion is that the ENTIRE demand curve is mved outward by the subsidy. That seems misleading as the subsidy applies only to certain people. Also, I am not entirely sure what market you are graphing? Are you graphing the marlet for health insurace coverage? Are you graphing the total demand for ALL health insurance?
Posted by: GeorgeNYC at Jan 25, 2007 11:32:15 AM
People who have insurance through their employer pay smaller premiums due to the nature of underwriting a policy that spreads the risk over more people. Also, health insurance premiums paid through an employer, regardless of the cost, are paid prior to any payroll or income taxes. People with private policies can only deduct insurance premiums in extreme circumstances on their taxes at the end of the year.
He views the best route to resolving this discrepancy as subsidizing everyone with a standard deduction from our taxable income.
Is this the advantage that he wants to level between the those who pay for insurance through an employer and those who buy private policies?
Thanks for the clarification.
Posted by: Josh at Jan 25, 2007 11:37:03 AM
I thought hat do to AMA pressures the supplyof doctors is not extremely elastic.
Posted by: mike at Jan 25, 2007 11:41:19 AM
Perhaps Krugman believes that rising health costs are due to an upward shift of the horizontal supply curve rather than a movement along the curve. Keith, would this explain Krugman's reasoning better? I'm pretty new to the economics of health-care and to this particular argument but the curve shift makes sense to me.
Posted by: eriks at Jan 25, 2007 12:28:29 PM
pwned
Posted by: asg at Jan 25, 2007 12:29:14 PM
"Perhaps Krugman believes that rising health costs are due to an upward shift of the horizontal supply curve rather than a movement along the curve. Keith, would this explain Krugman's reasoning better? I'm pretty new to the economics of health-care and to this particular argument but the curve shift makes sense to me."
And these upward shifts (a decrease in health care supply) would be from...? You actually need an event that decreases the overall supply of health care to actually shift the supply curve like that...does Krugman believe that a sudden illness will disproprtionately kill doctors, that computer viruses or electrical failures will destroy medical equipment?
Nope, it looks like Krugman has either predicted a future drop (or at least leveling off) of future health care costs thanks to shifting short-run supply curves along the horizontal long-run supply curve, or he made an erroneous argument in order to defend a flaw in his initial reasoning.
Krugman's only argument left: He really believe health care outsourcing will take off and that we will pass legislation that will allow health insurance companies and organizations to do business across state lines, which will then lower (level off) heal;th care costs. In this model, short-term political pressures are endogenous, so that when health care costs get high enough, politicians then undo restrictions that made the supply curve inelastic.
Posted by: Keith at Jan 25, 2007 12:54:26 PM
I suspect the difference has to do with disagreement over what economic analysis means.In Tabarrok's world, proposing a model and drawing a chart illustrating it is sufficient. Krugman, on the other hand, is more stringent - I'm guessing he would like to see some real data validating the model.
Posted by: alex at Jan 25, 2007 1:01:18 PM
Many problems with Alex's post. First, the graph goes too far in assuming that the supply of medical care is perfectly inelastic (though i think it's also a stretch to assume that it's perfectly elastic over time). It's way over-simplistic to represent either the supply or the demand of 'medical services' as a unitary thing - so from a functional standpoint of representing something about the real world, it's meaningless.
It's also missing the point to depict a tax subsidy to the purchase of group insurance as if its a subsidy to the purchase of medical care per se. Insurance 'subsidizes' the purchase of medical care in two fundamental ways: by annuitizing or amortizing the cost of medical care that the insured population will consume over their lifetimes; and by making it possible for insurers, representing groups of consumers, to negotiate for lower prices. Both of those advantages are subject to very strong economies of scale and scope, with the result that for any given population, group insurance is always more efficient (lower total cost per person) than individual insurance, and is always more efficient for larger groups than for smaller groups. (There are also, in theory, some economies of scale in payment processing and so forth; and managed care advocates would say that insurance can improve the quality of information and care to boot. But I think those are speculative and marginal benefits, so ignoring them for this purpose.) So before we join Alex in laughing off Krugman's argument, let's remember what it is that the President has suggested, that Krugman is rightly critiquing: reducing the tax subsidy for employer-provided group insurance, and using the proceeds to provide a new subsidy to individual insurance. Right off the top, that implies using our tax system to subsidize a less efficient type of health insurance, at the expense of a more efficient type. That's a deadweight loss, one that Alex might want to consider representing somewhere on his tidy little graph.
But i understand Alex is making a theoretical point, not a quantitative one, so let's look at it that way. If we subsidize one person's consumption of a scarce good, we can end up driving up the price of that good for everyone else - fair enough. If nobody had comprehensive insurance - i.e. if everyone had to pay for medical care out of pocket - then we could expect the prices of a lot of medical services to come down. The mechanism by which that would happen would be a reduction in consumption - probably a pretty drastic one. Over time, medical care in general being a high-fixed-cost type of good, the supply curve would most likely shift to the left, which would depress consumption even further, up to the point that some new equilibrium would be achieved. In the meantime, we as a society would have to deal with all of the externalities to a fall-off in the use of medical care that don't make it into Alex's graph: more undiagnosed chronic conditions that spiral into acute health crises, more people courting bankruptcy based on medical bills, more people foregoing even necessary care because they don't have the money, fewer medical providers and fewer choices of treatment (because there won't be sufficient buying power out there to support very many specialists), fewer medical advances (because why invest in a market where demand is drying up) ... all of the things that were characteristic of health care in the US before the dawn of group health care in the 40s - 50s & Medicare/medicaid in the 60s-70s.
Alex seems to be a smart guy, but this is a dumb, cheap-shot post. It's the kind of argument that somebody would make if they dropped Econ 101 halfway through the semester: taking the most rudimentary of economic analysis tools and treating it like it holds the answers to everything.
Posted by: Tom at Jan 25, 2007 1:16:55 PM
A large number of people with plans that do not discourage economically unreasonable defensive medicine tests such as MRIs for simple headaches [an exaggeration, perhaps, but maybe not as much as you might think] can make such tests arguably the "standard of care", which can make treatment of simple headaches more expensive for everyone.
There's no long term / short term issue here, although as for any other product technology can change the supply curve -- either way.
-dk
Posted by: Dick King at Jan 25, 2007 1:38:30 PM
"First, the graph goes too far in assuming that the supply of medical care is perfectly inelastic"
No it doesn't. Perfectly inelastic would be if the curve is vertical. Since the curve is linear, I'd need to know the price and quantity to tell you if that's the elastic or inelastic part of the curve, but even a rudimentary familiarity with Econ 101 would tell you that the curve is not perfectly inelastic.
And Krugman did not make the defensible claim that Econ 101 may not be a complete paradigm for health care economics. Krugman made the indefensible claim that no such analysis of subsidy increasing price existed. In addition, by arguing that he believes the long-run supply curve is horizontal, Krugman himself used Econ 101 tools to defend his argument.
Let me say, I believe, contra Econ 101, that subsidies COULD decrease long-run prices, if we subsidized certain preventive care. I believe that because I believe people get disutility from visiting the doctor, and so don't visit until something's a bigger problem, which causes negative pecuniary externalities. But we do not subsidize preventive care now; we subsidize other types of care.
Posted by: Keith at Jan 25, 2007 1:39:41 PM
