Aaron Katz likes to buy his food from member-owned nonprofit co-ops. Health care is another matter.

After President Obama's plan for a government-run health plan fizzled, the idea of a national network of health care co-ops took the place of a public option. While it's "a warm and fuzzy idea," says the University of Washington professor who studied health policy for 30 years, co-ops don't lower health care costs or cover the uninsured, two central goals of the president's reform. More than 18 percent of the state of Washington's adult population under 65 (the entry age for Medicare) is uninsured.

As a University of Washington employee, Katz has a choice of at least seven health plans, including private insurers and a co-op that is being touted by some as a model for national reform: Group Health Cooperative of Seattle. Katz chooses another option: a state-run insurance plan offered to university employees. Or as Katz puts it, he chose the public option.

Katz isn't the only skeptic. Timothy Jost, a health policy expert and law professor at Washington and Lee University in Lexington, Virginia, says health care cooperatives may be an ineffective method of fixing an ailing health care system on the national level. A proposal by Senate Finance Committee chairman Max Baucus calls for nonprofit co-ops to be set up around the country to compete with commercial insurers. But there's no conclusive data showing a co-op does what President Obama initially set out to do: create a plan that provides real competition for insurance companies and covers tens of million of uninsured Americans.

Even if the co-ops provided effective competition, lowering costs or decreasing the number of unsinured, they face a number of barriers to entry into most markets. It will be difficult for co-ops to vie for insurance business on price or doctor networks because they lack the scale and market power of large private insurers. There are hurdles to getting started: It takes a lot of money to capitalize and market an insurance company, likely much more than the $6 billion Baucus estimates.

"The big problem is that it's really hard to enter a market that is already heavily concentrated," Jost says. To compete effectively, he says, insurers need to affiliate with doctor and hospital networks. "You're not going to get the kind of discounts that big insurers get. Forming your own network is a huge burden and this can be a barrier to entry into a market."

Group Health was founded in 1947 by farmers and union workers. The group bought a medical clinic and a small hospital to provide coverage for middle-class people who didn't have insurance. The co-op continued to grow over the decades, opening new medical centers, adding doctors, and creating a research arm. It now has 600,000 customers and ranks among the top-three-largest health insurers in Washington with Regence Blue Shield and Primera Blue Cross. The three companies make up almost 60 percent of the market, according to state insurance data.

To be sure, Group Health is as an innovative company that plays the part of insurer and care provider. It reinvests all profit back in the system to pay for new programs and technology. For all the talk at the national level about modernizing medical records, Group Health already made the conversion to digital records. Doctors are full-time employees of Group Health and are paid salaries rather than the traditional fee for services performed. Group Health says its doctors are paid competitively and that it has no problem hiring staff. Patients give the insurer and its doctors high marks, according to a survey by Consumer Reports.

But for all its years innovating and providing good patient care, Group Health doesn't radically change the insurance market in Washington, which means a key rationale for creating a nationwide system of co-ops makes little sense.