STANFORD GRADUATE SCHOOL OF BUSINESS—It's a poorly kept secret in the healthcare industry that nobody—except maybe the uninsured—pays list price for hospital services. State deregulation of hospital pricing has pushed insurance companies into aggressive negotiations with hospitals to get the best deals, and the resulting discounts vary greatly from insurer to insurer. But why some insurance companies get much better deals than others has only been guessed at until recently, when Alan Sorensen, an assistant professor of strategic management at the Stanford Graduate School of Business, began researching the topic.

"The conventional wisdom was it was all about volume," says Sorensen. "The bigger you were, the bigger discount you could get." This notion is so pervasive that some industry insiders believe it may have led to the consolidation waves of the 1990s, when both insurers and hospitals raced to increase their bargaining clout by becoming bigger.

"That explanation never really did it for me," says Sorensen. After all, even some small insurers were able to negotiate substantial discounts.

The real key, Sorensen suspected, was not size per se (that is, not the total number of patients the insurer represents), but rather the ability of insurers to take their business elsewhere. "In any bargaining situation, the more credible your threat to withdraw from the bargain, the better you're going to come out in the negotiation," he explains. For insurers, he reasoned, the bargaining power comes from the promise to channel all or most of their patients to a small set of hospitals, those who'd agree to substantial price breaks. This would mean that traditional indemnity health coverage plans, which offer enrollees free choice on where they get their medical services, would pay more for hospital services than insurers with restrictions on access, such as preferred provider organizations (PPOs) that limit reimbursements for hospital visits outside their network or health maintenance organizations (HMOs) that typically have even more restrictive plans.

But while this insight is intuitive, finding the highly confidential pricing data to test the idea proved tricky. The one exception was Connecticut, whose deregulation laws required every hospital to list overall discounts for each insurer. Thus, while Sorensen couldn't compare the price of, say, an appendectomy negotiated by insurer A versus insurer B, he did know which insurers negotiated aggregate discount levels at a particular Connecticut hospital. As expected, the discounts varied greatly. At a single hospital one payer might be getting a discount of 3 percent while another had negotiated 25 percent. And, as predicted, managed care plans (PPOs and HMOs) were getting bigger discounts than traditional indemnity plans.

But that data wasn't enough to show that better discounts were going to insurers because they were pushing their customers to certain hospitals. Other reasons—such as size of the insured population—could explain differences in negotiated price. Besides, Sorensen wanted to show why even insurers of the same type, say, the CIGNA PPO in Hartford and another regional PPO, were getting different prices from the same hospital. So Sorensen needed data showing actual charges incurred by different payers at the various hospitals. Luckily, this charge data for almost all the hospitals was available from the Connecticut Hospital Association.

By merging the discount data with the charge data, Sorensen could at last perform the kind of analysis needed to test his hunch. "For plans that can channel their patients, you should see big discounts associated with big market shares," he explains.

And, indeed, he found that in cities with two competing hospitals—such as New Haven's St. Raphael and Yale hospitals—insurers that didn't receive a good discount from a given hospital incurred few patient charges there. In other words, the insurer channeled few patients to the hospital that offered the worst negotiated rate, just as Sorensen had predicted. And, on the whole, HMOs received substantially better discounts than did PPOs. Sorensen's estimates suggest that while an insurer's size does affect bargaining clout, the impact of size is small compared to the impact of the insurer's willingness and ability to control the flow of patients to one hospital or another.

Fair enough, but it's easier said than done. "Although it seems clear that [health care insurers] can increase bargaining power by tightening their network, the cost is that consumers don't like it," says Sorensen. If patients are loyal to a particular hospital because of its reputation in labor and delivery, for example, or because it's the only hospital that can perform a particular procedure, insurers can't credibly threaten to completely exclude that hospital from their network.

The same tradeoff to restrictiveness may well hold in other industries, Sorensen believes. Retailers can negotiate big discounts with manufacturers by promising to stock their brands exclusively, but merchants risk losing their customers to stores that offer wider selection.

Related Links

Insurer-Hospital Bargaining: Negotiated Discounts in Post-Deregulation Connecticut, Alan T. Sorensen, The Journal of Industrial Economics, 2003, 4 (December), 469-490

Additional Reading

"Change, Consolidation, and Competition in Health Care Markets," M. Gaynor and D. Haas-Wilson, Journal of Economic Perspectives, 1999, 13 (1), 141-164

"Hospital-Insurer Bargaining: An Empirical Investigation of Appendectomy Pricing," J. M. Brooks, A. Dor, and H. S. Wong, Journal of Health Economics, 1997, 16 (4), 417-434

"Countervailing Power in Wholesale Pharmaceuticals," S. F. Ellison and C. M. Snyder, MIT Working Paper, 2001 (July), 01-27