A few days ago over at The Health Care Blog, Robert Laszewski posted a list of Sarah Palin's health care priorities while serving as governor of Alaska. Number one on her list was the repeal of certificate of need (CON) laws in the state. Such laws give state planning agencies the final say in approving the construction and development of a new hospital, nursing home, or medical service center. Simply put, in the 36 states currently regulated under CON laws, nobody can build a hospital or introduce a new hi-tech device such as magnetic resonance imaging (MRI) scanners without first getting government approval.

    This approval is based on "need" and "quality assurance." Basically, the planning board asks whether a given community could benefit from a service or facility and if those services can be delivered effectively over time. In theory, the main goal of such a vetting of facilities is to reduce health care costs: by regulating the supply of health care in a given region, CON regulations are meant to limit the proliferation of expensive, medically unnecessary services.

    On paper, CON regulations sound exactly like the sort of policy that we at Health Beat have been advocating for a while now: one that realizes an all-you-can-eat buffet of health care options drives up costs without improving quality. Yet CON regulations haven't been as successful as supporters hoped-not just by the ideological standards of free-marketeers, but also in terms of empirical impact. Why is this so?

    Critics like Palin contend that CON laws inhibit competition and perpetuate health care monopolies, thus keeping health care prices high. But the fact is that in communities where there are more hospitals, hospital prices are no lower. Often they are higher because competing hospitals all invest in the same high-tech equipment, creating excess capacity. If six hospitals in an eight mile radius all have MRI units, the only way to pay for the units is to use them. Excess capacity drives unnecessary treatments-showing once again how the health care market is different from other markets. In most sectors of the economy, more competition will lead to lower prices. But when it comes to health care, supply drives utilization. Build the beds, and somehow, they will be filled.

    Nevertheless, the always-conservative Bush Administration stands by the story that CON laws don't work because they get in the way of competition: a 2004 Federal Trade Commission/Department of Justice report concludes that CON laws "are not successful in containing health care costs."

    In fact, the FTC and DOJ are half-right. There are indeed a fair number of studies showing that CON laws have not reduced total health care expenditures in states that have them (like this one from Duke University and these two from the University of Alabama).

    Yet on the whole, evidence on CON regulations isn't definitively negative, but rather mixed. There is in fact research out there suggesting that CON regulations actually help to reduce health care spending. Some of the most oft-cited research that CON laws work comes from America's Big Three automakers, who, of course, have a deep interest in controlling runaway health care inflation.

    Instead of looking at the before and after of CON regulation in the same state, the Big Three research compared health care costs in different states: those with CON laws and those without. In 2000, the DaimlerChrysler corporation found that health care costs for employees were up to 164 percent lower in states with CON programs than in those without CON laws. As Joe Paduda recently observed on his blog, GM and Ford came to similar conclusions: GM spent almost one-third less on health care for its employees in CON states and Ford calculated that inpatient and outpatient hospital costs were 20 percent lower in CON states.

    Automakers, at least, became convinced that CON regulation is associated with lower health care bills. In its report, GM said that, while "some argue that deregulating health facility expansion will trigger free-market forces of supply and demand and lead to lower costs," the company "has not found that to be true based on our vast experience in states that have varying degrees of CON regulation."

    Moreover, for every paper that suggests that CON regulations don't contain costs there are case studies which reveal that when CONS law are repealed, the building boom that follows almost certainly exceeds the community's needs. For example, when Ohio eliminated its CON program in 1995, the next four years saw an increase of 19 new hospitals, a 137 percent surge in outpatient dialysis stations, and a 600 percent increase in ambulatory surgical centers. After Pennsylvania's CON law expired in 1996, the state saw a dramatic growth in the number of open heart surgery programs, which increased from 35 to 62.

    Did the citizens of these states really need that many more dialysis stations, ambulatory surgical centers and open heart surgery programs? It hardly seems likely.

    But given what we know about how supply drives the use of medical facilities, the odds are high that many borderline patients soon found themselves either on dialysis or undergoing open-heart surgery.

    It's worth noting that belief in CON laws waxes and wanes as the political mood of the country changes. In 1974, the federal government passed a law requiring all states to beef up their CON regulatory structure at the risk of losing federal funding-a move which effectively led to the passage of CON laws in all 50 states. Thirteen years later, at the height of Reaganism, deregulation was the order of the day and Congress repealed its federal requirement. Within the next decade 14 states rolled back or eliminated their CON regulations. (For a list of which states still have them, and which facilities and services are regulated, click here).

    Today, as concern about health care inflation once again rises, some of these states are toying with the prospect of reinstating their CON requirements in order to take another crack at reining in health care spending.

    For health care policy observers, the empirical and political confusion surrounding CON regulations is pretty disorienting. Does the conflicted record of CON laws mean that regulating supply-side medicine isn't an effective cost control measure in health care? Should Maggie and I throw in the towel and join the ranks of folks like Palin, who insisted in a February op-ed that it's essential to "allow free-market competition and reduce onerous government regulation" in health care?

    Not so fast-there's one aspect of CON laws that may help explain why the results have been so muddled, and this has to do with the way CON measures whether the proposed services and facilities are "necessary" and how it assures "quality." How do state health planning agencies make sure that these two priorities are achieved? By setting some minimum volume of procedures that all approved bodies must meet. That's right: CON regulations may limit the number of hospitals in a community, but they actually mandate that existing hospitals do more.

    Health care entities seeking CON approval need to project, and then meet, a certain benchmark of procedures or treatments in order to justify their existence. The logic behind this requirement is that, if a surgical center doesn't perform at least X amount of surgeries, it's not worth the investment-because these surgeries are clearly not essential to the community. Concerns over quality also drive CON regulations to encourage a high-volume of care: studies show that facilities which do the same procedure more often tend to produce better outcomes. And so, because practice makes perfect, CON laws seek to concentrate a high-volume of care at fewer medical centers.

    But here's the problem: the fact that CON laws demand that hospitals and other facilities meet yearly quotas in terms of procedures and treatments may well impact the regulation's effectiveness when it comes to containing costs. When a hospital knows that bureaucrats are going to crack down on it unless it performs cataract surgery on at least 50 patients per year, it's going to scramble to find and operate on those patients. In fact, research from Baylor University and Rice University suggests that CON minimum volume requirements on angioplasty and open heart surgery make hospitals feel that "they need to perform the extra procedures to meet the regulation's minimum volume quotas." Indeed, facilities in CON states take on many more open heart patients than those in non-CON states: studying more than 900,000 cases of open heart surgery performed from 1994 to 1999, researchers at the University of Iowa found that the volume of procedures per medical program was 84 percent higher in CON states.

    Volume requirements are part and parcel of CON laws around the nation. In New Jersey, an aspiring hospital must commit to performing at least 15 pancreas transplants a year to garner CON approval; in West Virginia, a hospital must perform no fewer than 300 cardiac catheterizations a year; and in Washington state, "hospitals applying for a pediatric cardiac surgery and interventional center certificate of need must demonstrate that they can meet one hundred ten percent of the minimum volume standards" - one hundred cardiac surgical procedures and one hundred fifty cauterizations every year. In Michigan, overall volume is used as a proxy measure for quality of care in state calculations.

    Given the fact that CON laws seek to regulate the distribution of care, but not the overall volume of care, it's not really surprising that they've become so difficult to assess. One supply-side variable (number of facilities) is being restricted, while another (volume of procedures) is potentially magnified. It may well be that, when it comes to cutting down on over-utilization of health care, CON regulations are fighting only half the battle.

    Let's be clear: this doesn't mean that the concept or intention of certificate of need programs is somehow unsalvageable. It's just that, like most policies, CON laws are imperfect, and could benefit from refinement and rethinking.

    Instead of placing so much value on the number of procedures facilities are doing, lawmakers should focus on how many patients in the community say that they needed a procedure and had to leave the community to get it, or had to wait an unreasonably long time before a slot opened up. That's a better indication of whether or not a community needs more facilities and services, because existing facilities are always likely to operate at full capacity-again, build the beds and they will come. But a greater number of filled beds doesn't mean that those beds are medically necessary.

    So be careful when you hear conservatives like Sarah Palin condemn CON's admittedly less-than-spectacular results as confirmation that free markets are the answer in health care. Just because a policy is imperfect doesn't mean that the absence of that policy-or an abandonment of its intentions-is the answer. The conversation on certificate of need programs isn't over yet.