By contrast, Congress could achieve the cost-control objective of the of the public option by considering the practical operation of the proposed mandate for all individuals to obtain health insurance . Under most proposals before Congress, the key to implementing an individual mandate is a Connector – sometimes called an exchange – a government-run board that would act as a group purchasing agent for all the uninsured in the relevant area.

The Connector would negotiate with multiple insurance companies and health care providers to offer a menu of different policies to the uninsured. The Connector also would distribute premium subsidies, financed by the federal government, to help moderate-income families (too well-off for Medicaid) buy health insurance. Specifically, the Connector would generate a menu of health care policies for all those falling through the cracks of the existing system – those not covered by Medicaid, Medicare, their union, or their employer. An uninsured individual does not have the bargaining power or economies of scale to negotiate a low-cost policy with a large insurance company. By representing hundreds of thousands of uninsured, the Connector can obtain better policies than any of those uninsured could have obtained individually.

To make Connectors more effective as group purchasing agents, Congress could require them to offer a public option as one of the health care policies on their menu. Its availability should serve as a competitive constraint on the prices charged by the insurance companies negotiating with the Connector to provide health care policies. So what type of public option would best play this role for the Connector?

An effective Connector must focus on the insurance companies and health care providers in its state, or adjacent states in the region. It would not make sense for the Connector in Massachusetts to negotiate for insurance policies with parties who operate mainly in Iowa or Oregon. Health insurance in the United States has long been organized state by state, each with its unique structure of hospitals, doctor organizations, and patient demography.

Since an effective Connector needs to focus on one state or region, this should guide the search for the most workable definition of the public option. Every state already has a variety of governmental plans serving public employees – e.g., employees of state agencies or local teachers. Therefore, each Connector should offer, as one of the choices on its menu, the opportunity for the previously uninsured to join such a state-based health care plan.

Compared with other approaches to the public option, health care plans for state employees or local teachers have several key advantages. First, these plans are public, but without the ideological ramifications of a national system. Second, these plans should assuage concerns about the possible erosion of state rights by national health care legislation. Third, these plans are already in operation in all states, as opposed to vague notions of health cooperatives that do not exist in most states. Most important, these plans would offer the uninsured the chance to obtain a high-quality health care policy at a relatively low cost.

In short, within the context of state-based Connectors implementing the individual mandate, the most viable definition of the public option is a state-based health care plan currently serving governmental employees. That approach would sidestep the ideological debate, while helping to achieve the consensus goal of constraining health care costs.

Robert Pozen is chairman of MFS Investment Management and a senior lecturer at Harvard Business School. His new book, “Too Big To Save?’’, will be published in November.