I won’t pretend to have a comprehensive answer to that question; it’s a pretty big one. But I do think the best evidence we have indicates that when health insurance comes with high deductibles and co-pays, consumers do on average make rational health care choices by reducing their health care expenses with almost no adverse health effects. This is the result of the famous $50 million RAND Health Insurance Experiment, a study which is old news to people who read about health care policy (it’s so famous it even has it’s own Wikipedia entry!).
Or so I thought anyway. Economist Paul Ginsburg seems to disagree, and finds the state of the literature ambiguous:
…many of those companies will rely on what she described as “the tried-and-true method” — passing along more of the costs to employees, in the form of higher deductibles and co-payments, in order to reduce overall premiums.
The public policy goal of the tax, in theory, is to have everyone spend less on medical care, even if it means using it less.
“We know people will use less care under such plans,” said Paul Ginsburg, president of the Center for Studying Health System Change, a nonpartisan group.
What is not so clear, Mr. Ginsburg said, is whether people will make — or be able to make — rational choices between treatments that are not particularly effective and treatments that may help them from becoming sicker later.
Mr.Ginsburg is a frequent author of papers in Health Affairs and a noted expert on the health care policy, so when he makes this claim part of me suspects my disagreement may be based on evidence I am unaware of. I skimmed some of his past articles looking for clues as to what studies he may be referencing but have found none yet.
The CBO at least seems to agree with me. In a paper on price transparency in health care (H.T. Tyler Cowen) they say
The RAND Health Insurance Experiment, conducted in the 1970s and 1980s, showed that individuals purchase less health care when faced with greater cost sharing and that the reduction in health services had little impact on their health. Although the study did not consider the question of price transparency, its results suggest that individuals who pay for health care on their own are responsive to out-of-pocket payments.
Clearly, the evidence shows consumers can make rational decisions about seeking effective versus ineffective medical care. Perhaps Ginsburg is referring to selecting among treatments once a patient has decided to seek some medical care, which the next line in the CBO report refers to as an unsettled question:
At the same time, the evidence indicated that the primary effect was on whether individuals sought care for a medical condition; once they went to a doctor, subsequent spending per person was similar in the different cost-sharing arrangements.
If this is what he meant, his statement seems consistent with the RAND HIE. However, that is a very selective representation of the literature by the authors of the article Ginsburg was quoted in. The article is addressing the “public policy goal of… hav[ing] everyone spend less on medical care, even if it means using it less”. The answer to whether consumers can do this without adverse health effects, at least to the best of my knowledge, should be “yes, the evidence suggests they can”, not a technically-correct-but-missing-the-bigger-point statement about selecting amongst treatments. Then again, perhaps Dr.Ginsburg and the article’s author, Reed Abelson, are aware of evidence that I am not. Dr.Ginsburg is the expert, so this is certainly a possibility.
12 comments
Comments feed for this article
Tuesday ~ December 29th, 2009 at 12:04 pm
Tom Hickey
The idea that “the representative consumer” makes “rational” (fully informed) health care choices, or could make them, is absurd on the face of it, and anyone who has been faced with any kind of serious medical decision knows this. Unless one is a physician or someone equally educated and up to date in the field, one is totally at the mercy of the system, including the bean-counters at one’s insurance company. And there is often some disagreement regarding expert advice, e.g., between first and second opinions. This isn’t about “tear and compare.” How can a “representative consumer” possibly know about relative effectiveness of different treatments based on evidence. Oh, “the invisible hand.” Right. Count me skeptical. And just why are consumers being presented with choices that are relatively ineffective in the first place? This is best practice?
Health care is a good instance of a sector in which rational expectations really does work. If the goal is for everyone to spend less on health care, then more must be spent lower down in the system, making sure that everyone is properly examined on a regular basis, instructed in preventive measures, and can contact a qualified practitioner immediately when a possibly serious problem arises independently of the emergency room.
The objective of health care reform should not be to “spend less” or “bend the curve,” but to provide maximally effective care at optimal efficiency. From a purely economic standpoint, disregarding the human values involved, we are dealing with our human capital here, and capitalism is about capital formation, preservation, and effective/efficient deployment. A great party of the cost involved in health care is loss of potential due to degradation of human capital. That is almost never plugged into the model.
Tuesday ~ December 29th, 2009 at 12:05 pm
Tom Hickey
Oops, I see that I should have written in the first sentence of the second para, “Health care is a good instance of a sector in which rational expectations really does NOT work.”
Tuesday ~ December 29th, 2009 at 4:54 pm
teageegeepea
Did you actually read the above post? I see little evidence that you are responding to it. For example, you could have mentioned the RAND study, if only to dismiss it.
Tuesday ~ December 29th, 2009 at 5:18 pm
Tom Hickey
Uh, I am saying that the RAND study is biased by rational expectations theory, a favorite tool of researchers there.
Soldiers of Reason: The RAND Corporation and the Rise of the American Empire by Alex Abella
http://www.amazon.com/Soldiers-Reason-Corporation-American-Empire/dp/0151010811
Wednesday ~ December 30th, 2009 at 9:28 am
Adam Ozimek
Tom,
Your criticisms seem completely unrelated to the RAND HIE study. For one example, what is “maximally effective care at optimal efficiency” if it’s not decreasing healthcare costs without affecting health outcomes? I’d be very interested to hear specifically how rational expectations assumptions have lead the HIE to false conclusions I discuss above, what your alternative assumptions are, and how they lead you to interpret the results of this randomized expirement differently.
The only thing you’ve shown thus far is that your skepticism of rational expectations has biased your ability to interpret sound empirical work.
Wednesday ~ December 30th, 2009 at 1:43 pm
Tom Hickey
Adam, thanks for getting back on this important issue instead of dismissing my comment as irrelevant.
My apologies. I apparently failed to make the reasoning clear. Let me try again.
You state: “But I do think the best evidence we have indicates that when health insurance comes with high deductibles and co-pays, consumers do on average make rational health care choices by reducing their health care expenses with almost no adverse health effects. This is the result of the famous $50 million RAND Health Insurance Experiment, a study which is old news to people who read about health care policy (it’s so famous it even has it’s own Wikipedia entry!).”
This is about one study. The most one study can used to justify is that the conclusion may suggest something important (and it may not). In most debates about such matters, a number of studies are required to justify a conclusion involving policy that will impact millions of people, involves trillions of dollars affects a key sector of the economy, and has a fundamental impact on national quality of life.
Hanging all this on a single study? I would regard the RAND study as suggestive rather than indicative, and contributory rather than definitive.
I also suggested that the RAND study is based on the controversial assumptions of rational expectations. There is a significant literature discussing this, as I’m sure you know. I really don’t have anything to add to what has already been objected, other than to say that I find many of the objections well taken. Strike one against the RAND study.
In addition, the fact that the study did not consider price transparency is telling, I think, because prices in health care are notoriously opaque. That, in my view, is significant for the conclusion. What the study seems to show, as far as I can see, is that when care is more costly, people choose less of it. Duh. As far as the further conclusion, that this is immaterial to their health, is, I think, a stretch for one study to show definitively.
The other issues you mention is the empirical evidence. Studies must be methodologically sound and also based on good evidence. There are two ways that evidence is generally checked, when there is a agreement the method used was appropriate and properly applied. The first is checking the data. I am not in a position to evaluate the data of the RAND study, so I cannot comment on that aspect.
“The article is addressing the “public policy goal of… hav[ing] everyone spend less on medical care, even if it means using it less”. The answer to whether consumers can do this without adverse health effects, at least to the best of my knowledge, should be “yes, the evidence suggests they can”, not a technically-correct-but-missing-the-bigger-point statement about selecting amongst treatments.”
I would say that if “the best evidence” is the RAND study, then more research needs to be conducted before this conclusion is used to establish public policy. To my mind, Dr. Ginsburg is suggesting that the study is too simplistic and not nuanced enough to present a true picture.
This is a problem with models based on non-empirical assumptions like rational expectations. They provide a thumbnail heuristic that is suggestive but needs to be fleshed out. Engineers use rules of thumb and simple heuristics to think about problems, but they know that they cannot build bridges and buildings with them, let alone put a man on the moon. Health care is on that level of engineering complexity.
That brings me to points two and three. The second is replication using a different data set but the same methodology. A single study is not considered more than suggestive until it is replicated. I would want to see replication through other research. Strike two.
The third is replication of the conclusion of the study using not only different data but also different methodology. I am not aware that this has been done.
As I indicated, I don’t automatically take RAND as an authority. While I don’t have recent experience with their work, I am familiar with them from the Vietnam era, when all kinds of so-called authoritative research based on impeccable methodology and supposedly good data was used by the whiz kids in the WH and generals to justify policy, strategy, and tactics that led to an ignominious US military defeat and untold losses. Here we are dealing with health care, the impact of which is comparable on the domestic front. Sorry, I don’t trust one study, even by the prestigious RAND Corp. Sounds to me like an argument from authority. Strike three.
You ask what is “maximally effective care at optimal efficiency” if it’s not decreasing healthcare costs without affecting health outcomes? When it is put in terms of cost-cutting, then the conclusions like increased consumer cost-sharing without adversely affecting health sounds perfectly reasonable. (I am involved in an debate in another thread with a neoclassical economist who thinks that a 20% cut in wages across the board is a way out of the present crisis,)
However, I think that the issue is better stated in terms for what is the real output potential of the health care system now and going forward, in light of a lot of expensive technology coming online that promises to revolutionize health care in the 21st century How can the US provide the economic environment that utilizes the real output capacity of the health care system now and in the future? This not only maximizes health care but it also utilizes the full potential of an important sector of the economy, adds to human capital, and improves the quality of life of citizens.
Explaining how this is possible through the potential of fiscal policy properly conceived, without raising taxes or borrowing to fund it, goes beyond the scope of this comment. I’ll try to make time to address it in another comment.
Suffice it say here that I think “fiscal responsibility” arguments are adversely influencing the debate and biasing it toward cost-cutting. For example, the conclusion that increasing consumer cost-sharing misses the obvious point of removing obvious waste from the system. But there is a whole lot more that can be done to make health care (and education and other public utilities) function better while contributing to the economic advancement of the country, other than through subtraction.
Wednesday ~ December 30th, 2009 at 2:37 pm
Tom Hickey
I left off the previous comment with the observation that cost-cutting was not the best way to approach the questions of efficiency/effectiveness/equity in the health care debate. Let me state right way, that I am am not dismissing cost-cutting. Consumer cost-sharing, improvement in communications and record-keeping to prevent duplication of services, reduction of waste through elimination of less effective treatments, and so forth are, of course, integral to improving the efficiency of the system and should be vigorously pursued. However, they should not be regarded as any more than a part of the solution, and a minor part at that.
My contention is that subtraction is not as important as addition, if it is possible. What I mean is this. Health care needs to be looked at not only as such but as an important aspect of the economy going forward. Some of the most significant innovation is expected in the field of health care in this century. The challenge is not only to provide the best health care available to all but also to see health care an an important sector contributing to economic potential and to maximize this opportunity.
The largest challenge that most people see in health care is economic, namely, cost. How does the country fund the basic health care that is required, not to mention afford the best possible system. As a result, the best possible system is never even considered. This is putting the US at an economic disadvantage. If we could develop a new weapon that would make the US significantly safer, would we fail to provide it “because the country can’t afford it?” Of course, not. It would not even be seriously discussed.
The problem lies in failing to understand the economics. The problem is not one of funding goods and services that the government can purchase as part of the real output potential of the country as long as nominal aggregate demand does not exceed real output potential and lead to inflation. The thinking that the sovereign government is financially constrained is a hold-over of the convertible currency and fixed rate regime that characterized the gold standard. The US went off the gold standard in 1934, and President Nixon finally closed the gold window for international convertibility on August 15, 1971. Since then the US has been the monopoly issuer of a non-convertible floating fx currency of issue. Under these conditions, the US government does not have to “fund” its deficit spending with taxes or borrowing. It just credits and debits different accounts with a stroke of the keyboard.
There is a real constraint on currency creation, however, and that is inflation. Monetary inflation occurs only when nominal aggregate demand exceeds the real output potential of the economy. This means that the government can fund anything it chooses politically as long as it manages nominal aggregate demand so that AD doesn’t exceeds real capacity. This means that the government has wide leeway in funding public utilities such as education and healthcare, which are economic investments in the future of the national undertaken for the good of all citizens. So, yes, the US government can fund health care directly, as long as it manages nominal AD in relation to real output potential. The government also has the tools to do this.
Sound to good to be true? Not so. There is abundant research on this, and it is not theoretical. No assumptions necessary. This analysis is based on how the monetary system actually operates in terms of stock-flow consistency (initially developed independently of each other by James Tobin and Wynne Godwin). A good introduction to this subject can be found in Understanding Modern Money by L. Randall Wray (1998). It is available to read online at Google Books. Wray provides abundant references for follow-up. This is also discussed on in blogs by Wray, Bill Mitchell, Scott Fulwiler, Marshall Auerbach, and Warren Mosler. Numerous working papers by a variety of people working in this field are freely available at The Levy Economics Institute of Baird College.
My point is that along with cost saving, which conserves funds for more appropriate use, the ability of the government to fund public utility directly must be marshaled. In a fiat monetary system, money (net financial assets of non-government) is itself a public utility created by the sovereign government through its monopoly provision of the currency of issue. The government has not only the ability to provide money to facilitate commerce and advance public purpose, it has the corresponding responsibility to do so. The government alone can do this, because all transactions of commercial banking and finance net to zero as a matter of accounting. Failure to meet this responsibility appropriately damages the economy and nation through price instability and under-utilization of real capacity.
The US can afford to operate at its full potential, and conversely, it cannot afford not to operate at its full potential — this is a competitive world, and others are catching up fast. We should not not let fiscal scolds get away with erroneous arguments to the contrary, nor an intellectual elite supporting the self-serving agenda of a plutocratic oligarchy, whose objective is rent-seeking, as is presently happening. Let’s put the health care debate on proper legs instead of falling for the myths that are being propagated through shibboleths like “fiscal responsibility,” which are based on obsolete conditions and no longer apply.
Wednesday ~ December 30th, 2009 at 6:23 pm
Adam Ozimek
Tom,
I am not aware of the rational expectations critique of the RAND HIE. If you can pointt to such literature, please do, but I’m skeptical since the results of the expiriment- costs went down, health outcomes did not- do not depend whatsoever on rational expectations assumptions. They are the result of a large randomized trial, the gold standard in the social sciences. This is the best evidence we have on this issue, and as much as I can tell the only counterevidence you have is an ad hominem against RAND and an irrelevant theoretical objection.
If you want the government to stimulate economic activity, then shouldn’t it be in areas where we have evidence that the money is buying what we think it is buying? I doubt you would suggest we should building power plants if the best evidence we had suggested they wouldn’t generate any power.
If there is better evidence than the HIE that cost sharing does not decrease costs or decreases health, then by all means, present it, until then, the best you can declare is “we know nothing”, in which case your setting a very high empirical hurdle, which would condemn you to much more ambivalence than you seem to have about the world.
Wednesday ~ December 30th, 2009 at 8:21 pm
Tom Hickey
Adam, I am not asserting that there are criticisms of the RAND study based on rational expectations specially. I am simply stating that REH is controversial as a methodological assumption. For example, cognitive scientists have shown that this just not the way decision-making actually works because there is no such thing as “rationality” in the brain. Neural pathways inextricably intertwine logic processes and emotional response, as well as linking present conditions and past experience on many levels. Some levels are conscious and others are not so much, yet they are subliminally influential in our “rational” decisions.
So I do question your claim, “Clearly, the evidence shows consumers can make rational decisions about seeking effective versus ineffective medical care.” Rational?
Notice also that I am not denying that cost-sharing may be useful as one tool for cost containment if carefully designed. I suspect that there is a way to use cost-sharing effectively to reduce frivolous use of services, just as there is way to reduce waste, e.g., by denying coverage for anything but best practices IAW evidence-based medicine. We already do this in the alternative field, which is considered to involve solely private goods. The research indicates that billions of dollars are spent on treatments that are not based on evidence or are even disconfirmed by evidence. “Rational” decision-making?
However, the claim that cost sharing reduces costs without adversely affecting health outcomes is a broad one that needs to be carefully specified in determining policy. How much cost-sharing is appropriate for different income groups and family size? Is the cost-sharing socially equitable. Does cost-sharing exclude people from basic services or reduce use of relatively inexpensive services, and result in increased use of emergency rooms due to complications? We know, for instance, that early diagnosis of hypertension and diabetes greatly reduces subsequent costs. Does cost-sharing reduce timely diagnosis and inexpensive treatment so that we are treating more heart disease and kidney failure later on?
What I am saying is that taking the RAND study as “the gold standard” and building policy around it alone appears to me to be rushing to judgment on matters of extreme importance. Count me suspicious. My position is that, instead, further clarification of the issues and more research on specifics are needed in order to determine policy recommendations regarding cost-sharing as a tool, and only one tool among many.
Cost-sharing has been around long enough for me to suspect on anecdotal evidence that a lot of people with significant co-pays and deductibles are not seeking care in a timely way because of their perception of the arrangements they under. They tend to put off care until it is an emergency and they feel forced to seek aid. In my experience, a lot of people are psychologically disposed to postpone care even without cost consideration. If that is the case, then it is undermining human capital while increasing costs later. Such people need to be encouraged to seek care, not further discouraged.
Incentive a basic principle of motivation, and incentives underlie free market economic thinking. Just exactly how does the incentive and disincentive that cost-sharing involves work to produce a balance between economy and equity? Is this a hidden way of rationing what many see as a public utility on the basis of wealth?
I freely admit my bias toward providing more and better designed care, rather than restricting care in the name of cost containment. I think that there is an affordable way to do this. Existing research has been done on the present system, which I believe to be seriously broken. It is heavily weighted toward specialization and getting more so all the time. The US needs to redesign its approach to primary care by increasing access and encouraging use, so that early diagnosis and treatment can reduce much more expensive treatment down the line, not mention all the rest of the downside that goes with it.
Finally, I want to comment on the following statement, because I think it lies at the heart of this discussion. “The RAND Health Insurance Experiment, conducted in the 1970s and 1980s, showed that individuals purchase less health care when faced with greater cost sharing and that the reduction in health services had little impact on their health.” The logical conclusion is that the optimally efficient system would be one in which individuals bear all costs. Prices would decrease to what the market can bear, and the market would be the mechanism for price discovery through the equilibrium of supply and demand. This is Econ 101, but it would be a very different system from the one that is now in place.
The question it suggests is how much do we want to deviate from efficiency for broad effectiveness and social equity, instead of letting wealth alone ration care. this means that we have to decide whether health care is a private (discretionary) goods or a public (necessary) utility. What are the criteria, and what justifies their selection. These questions involve value judgments and are therefore questions for political economy. Policy is a combination of fact and value.
I have no trouble admitting that the RAND study is one contribution to the body of research that it relevant. I do have a problem with elevating it to the status of an idol without a serious examination for feet of clay.
BTW, I referred to “Wynne Godwin” in the previous post. It should be Wynne Godley.
Wednesday ~ December 30th, 2009 at 10:16 pm
Tom Hickey
I see that RAND came out with a 2006 reassessment of the HIE in which it mentions several points I pointed out above and concludes that another study is needed.
A Classic RAND Study Speaks to the Current Health Care Reform Debate
“Testing the effects of cost sharing in today’s environment and determining its usefulness as a tool for health system reform would require another large-scale demonstration.”
http://www.rand.org/pubs/research_briefs/RB9174/index1.html
Thursday ~ December 31st, 2009 at 2:17 pm
Adam Ozimek
Tom,
Another HIE would be great, and should be a priority for policy makers.
“I also suggested that the RAND study is based on the controversial assumptions of rational expectations. There is a significant literature discussing this, as I’m sure you know. I really don’t have anything to add to what has already been objected, other than to say that I find many of the objections well taken. Strike one against the RAND study.”
“Adam, I am not asserting that there are criticisms of the RAND study based on rational expectations specially. I am simply stating that REH is controversial as a methodological assumption.”
I’m not sure how to reconcile these statements. I take this to mean you are backing off your initial claim.
The question of the extent to which the government should provide health insurance, and how, is like you say more complicated a question than can be answered based on the RAND HIE results alone. None of this changes the fact that the best evidence we have suggests more cost sharing reduces health expenses without reducing health outcomes. More studies? Yes, great, we should have them. Until then, this is the best evidence we have, it’s good evidence, and it’s results should be our tentative assumptions.
Thursday ~ December 31st, 2009 at 3:45 pm
Tom Hickey
Adam, I don’t want to give the impression that I am “hard core.” I believe that science is tentative and, of course, we should proceed on the best evidence. However, what the evidence indicates is often controversial. I was primarily objecting to your statement, “Clearly, the evidence shows consumers can make rational decisions about seeking effective versus ineffective medical care.” You apparently were presuming rationality HIE, or at least I read it that way.
I still don’t think that it is clear that consumers necessarily make rational decisions that reduce costs without adversely affecting over the scope of decision-making that is involved in health care. The HIE study doesn’t assert that, at least as I understand it. This needs to be emphasized because what generally happens is that conclusions get extended, e.g, in the media, beyond the scope of the studies that produced these conclusions.
Should the original HIE be canned and we wait until it is redone? Of course not. We need to evaluate its findings along with everything else available and proceed to gather what seems to be needed to round out the picture. But I don’t think that we should be cowed by its being the putative “gold standard” either. Science is an incremental enterprise, in which no cow is sacred.
What about cost sharing? On a general level it is obvious that some cost sharing is required in the mix because “that which is underpriced is overused.” But the devil is always in the details.
Furthermore, I question whether consumer “rationality” is a sufficient basis for thinking about cost sharing a a principle, or justifying it as policy. I doubt that it is for reasons stated above. There is a lot of irrationality involved in health care decisions because it involves emotions, special knowledge, a hit to the pocketbook that many people cal ill afford. There is a lot of uncertainty as well, even with the best of evidence. I just had a vivid demonstration of this in recent discussions with a friend about options, evidence, and conflicting recommendations involved in treating his prostate cancer. Experts do disagree, and he is left with a personal decision he is not really equipped for, and he is a very smart guy.
I greatly appreciate your taking the time to read my comments and respond to them. My intent was not to challenge your position as much as inject another perspective into the discussion. I follow your blog because I think you are making valuable contributions to appreciating the wide spectrum of issues you treat. Thanks for freely sharing your knowledge with the world.
Happy New Year,
tom