December 2000 Fraser Forum:
Questioning Medicare's "Advantages":
Does Risk-Pooling Matter?
When it comes to justifying medicare, Canadians offer a number of reasons. Some salute its equity, while others praise its 'efficiency.' These might be good reasons to champion medicare—if they were supported by the evidence. They are not, unfortunately.
Its claimed equity has been refuted by a number of recent studies finding queue-jumping for heart surgery by politicians, celebrities, and doctors' friends (Alter, et al., 1998), and lesser access to specialists by poor people (Dunlop, et al., 2000), resulting in lower cardiac and cancer survival rates for the poor (Alter, et al., 1999; Mackillop, 1997). The claim of efficiency is at least as dubious, as my recent research found that provinces which spend more government money per person on health care have no shorter waiting times and no higher rate of service provision by specialist physicians (Zelder, 2000b).
One supposed virtue of medicare still appears intact, however—the efficiency of centralized provincial insurance plans. One aspect of this purported advantage is lower administrative costs. Perhaps medicare's defenders are right in this case, as OECD data indicate that insurance administration consumes far less of health spending in Canada (2.4 percent) than in the US (4.6 percent). Of course, these figures may just reflect the cost of running an effective insurance program in the US, in which non-beneficial uses of the system (e.g., patients with mild influenza going to the emergency room) are deterred by modest user fees.
This was demonstrated by the RAND Health Insurance Experiment, which implied that if Canada reformed medicare by imposing a 25 percent user fee on the non-poor (up to a modest annual expenditure limit), health spending on physicians and hospitals would be reduced by 19 percent and the health of the non-poor would be unaffected (Zelder, 2000a). Public spending on these two categories amounted to $36.8 billion in 1999; a 19 percent saving would constitute $7 billion. Because the poor would not be subject to user fees, the net saving would be somewhat less than $7 billion. This would easily exceed the additional administrative cost (assuming that Canada would spend the US share of 4.6 percent) of $1.9 billion.
But there is another advantage commonly attributed to 'national' or provincial insurance coverage: better risk pooling. Leading work on the economics of insurance from the 1960s and '70s identified a crucial theoretical problem for insurance markets: asymmetric information. This means that those seeking to buy health insurance have different (obviously more complete) information regarding their prospects of getting sick than do those selling insurance. Because the insurance companies cannot fully discern some of these intrinsic characteristics which make one person more illness-prone than another, and thus more likely to make insurance claims than another, they are forced to charge some sort of average premium rate which reflects the proportions of the more illness-prone and the less illness-prone in the population.
Charging this average premium creates a problem: insurance is underpriced for those who are more illness-prone, and they consequently buy more insurance than the less illness-prone. Insurance companies, of course, are aware of this tendency, and must thus charge higher premiums to break even. In particular, economic theory predicts that they will charge higher per-unit prices as individuals increase their insurance holdings, the opposite of a quantity discount for large-volume purchases. They are predicted to do this because it discourages high-risk individuals from loading up on underpriced insurance.
Because of these predicted tendencies, compulsory insurance at the provincial or national level has been widely viewed by economists as economically efficient, as it prevents insurance pools from being dominated by illness-prone individuals, and better spreads the risk of insuring those people among the whole population. The only problem is that economists have assumed that there are serious information problems of this nature with- out trying to measure whether there were or not.
That has all changed with a recent ingenious article by two economists from the University of Chicago, John Cawley and Tomas Philipson. Cawley and Philipson (1999) decided to examine whether these sorts of information problems existed in US life insurance markets. Specifically, they carefully assessed whether important predictions from the economic theory of insurance were confirmed in actual insurance contract data.
First, they tested whether individual purchasers of larger amounts of insurance faced increasing unit prices, as would be necessary in a world where insurers were concerned that high-risk individuals would purchase more insurance than low-risk ones. Amazingly, taking account of age, gender, smoking behaviour (yes or no), measured health status, income, wealth, and demographic factors, they found that larger insurance purchasers are charged decreasing unit prices as they purchase more insurance, the exact reverse of what the theory predicts!
They also directly examine whether the evidence supports a second important prediction of the standard model, that high-risk individuals purchase more insurance than low-risk individuals do. Again taking account of many factors, they discover that, in fact, low-risk individuals purchase more insurance than high-risk ones.
The implication of their intriguing results is to cast serious doubt on the argument that compulsory national/ provincial insurance is necessary to overcome information problems. This argument should, consequently, be tested on data from health insurance markets. Such tests may show that it is yet another justification which the defenders of medicare can no longer hide behind.
References
Alter, David A., Antoni S.H. Basinski, and C. David Naylor (1998). "A Survey of Provider Experiences and Perceptions of Preferential Access to Cardiovascular Care in Ontario, Canada." Annals of Internal Medicine 129: 567–72.
Alter, David A., C. David Naylor, Peter Austin, and Jack V. Tu (1999). "Effects of Socioeconomic Status on Access to Invasive Cardiac Procedures and on Mortality after Acute Myocardial Infarction." New England Journal of Medicine 341:1359-67.
Cawley, John, and Tomas Philipson (1999). "An Empirical Examination of Information Barriers to Trade in Insurance." American Economic Review 89:827-846.
Dunlop, Sheryl, Peter C. Coyte, and Warren McIsaac (2000). "Socio-Economic Status and the Utilisation of Physicians' Services: Results from the Canadian National Population Health Survey." Social Science and Medicine: 1-11.
Mackillop, W.J., J. Zhang-Salomons, P.A. Groome, L. Paszat, and E. Holowaty (1997). "Socioeconomic Status and Cancer Survival in Ontario." Journal of Clinical Oncology 15: 1680-9.
Zelder, Martin (2000a). "Canadian Health Reformers Should Understand RAND." Fraser Forum (February), pp. 8-10.
________ (2000b). "Spend More, Wait Less? The Myth of Underfunded Medicare in Canada." Fraser Forum (August).
Martin Zelder (martinz@fraserinstitute.ca) is Director of Health Policy Research at The Fraser Institute. He has a Ph.D. in economics from the University of Chicago.
info@fraserinstitute.ca
4th Floor, 1770 Burrard Street, Vancouver, BC, Canada, V6J 3G7
Tel: (604) 688-0221 Fax: (604) 688-8539 Book Orders: 1-800-665-3558 ext. 580
You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.