When Fewer Options are Better for Consumers: The Benefits of Narrow Health Insurance Networks


Should health insurers be required to allow their enrollees to visit any hospital or doctor?  Many insurers limit enrollees to “in-network” medical providers, forcing them to pay significant “out-of-network” costs if they seek care elsewhere. Our research examines the role that limited medical provider networks play in the U.S. commercial healthcare market and measures both their impact on spending and their potential for consumer harm. We show that by excluding certain medical providers, particularly those that are low-quality or high-cost, insurers can obtain significant rate reductions when negotiating with hospitals. These reductions may then be partly passed along to consumers in the form of lower premiums. While regulation that prohibits any exclusion of providers can lead to increased overall spending, we also show that targeted interventions protecting healthcare access for certain vulnerable populations may still be beneficial.

As policymakers, medical providers, insurers and consumers debate myriad proposed solutions to rising healthcare costs in the United States, the idea that insurers should be forced to expand their networks so enrollees can avoid expensive “out-of-network” charges is receiving renewed attention. This research examines the reasons insurers might choose to exclude certain providers from their plans and how this behavior can actually benefit consumers on average. Insurers may choose to offer a selective network in order to steer enrollees to lower-cost or higher-quality providers, deter the least healthy enrollees, or negotiate better prices with their “in-network” hospitals and clinics.

We apply a model of consumer choice and insurer competition to California public employee benefits agency (CalPERS) data to explore how adjustments in an insurer’s network can affect insurance premiums, negotiated provider prices, and consumer spending patterns. We find:

· By limiting the choice of medical providers, a CalPERS HMO plan would have been able to negotiate approximately 12% lower hospital prices on average, and up to 30% lower prices in specific markets.

· Consumers on average benefit from this more selective network, as rate reductions and the resulting lower premiums offset any harm from having fewer providers in-network.

· When there are many plans with large networks of providers to choose from, the average consumer actually prefers a plan with an even narrower network and lower premiums than the one chosen by their insurer.

· Regulations mandating insurers include all medical providers on their plan would result in higher hospital payments, higher premiums and reduced consumer surplus on average.

· Despite the benefits to the average consumer, significant costs may be borne by certain consumers who value the excluded medical providers most.

These findings suggest policymakers interested in reducing healthcare costs should focus on ensuring there is a competitive market in health insurance, rather than controlling which providers are included in insurers’ networks.

Main article

The use of selective or “narrow” networks by commercial health insurers in the United States is not a new phenomenon. Instead, it has grown and faded in popularity over the past several decades as a means of controlling costs—as seen during the rise of managed care in the 1980s and the subsequent “backlash” in the 1990s (Glied 2003). However, recent high-profile exclusions of providers from some insurance plans offered on state exchanges have reinvigorated debate over whether insurance networks should be more closely regulated.[1] Existing and proposed interventions include requiring certain services and geographic coverage and ensuring particular providers cannot be excluded.

The desirability and effectiveness of health network regulation depends on the reasons insurers might engage in exclusion in the first place, and whether the gains they realize are shared with consumers. In our research, we identify three main reasons why an insurer might wish to exclude a medical provider from its network, and we highlight when this exclusion actually benefits consumers.

  1. By excluding certain providers, an insurer can steer its enrollees towards alternatives. Steering can be helpful if the alternative providers are lower-cost or higher-quality than the excluded provider. This would not be harmful to consumers, as long as the insurer maintains adequate service and geographic coverage.
  2. An insurer can use its provider network to manage the selection of consumers onto its plan and potentially deter more costly enrollees by excluding providers valued by less healthy consumers.[2] This “cream-skimming” can have adverse effects if deterred consumers are left choosing worse plans, or having to go uninsured altogether.
  3. An insurer may be able to negotiate better prices with its existing medical providers as a result of offering a selective network. By committing to exclude providers, an insurer can persuade hospitals to bid down their rates in exchange for a place in its network. Whether this is beneficial depends on how large the savings are, whether the savings are passed along to consumers in the form of lower premiums, and the impact that any rate reductions have on providers’ quality, investment, or entry and exit decisions.


Research Strategy

As in prior related research,[3] we focus on the health insurance options provided by CalPERS, a major benefits agency covering California state public employees and their dependents (representing about 10 percent of the California commercial market). In 2004, CalPERS provided access to three large insurance plans for more than one million individuals across multiple geographic markets. These plans included a non-integrated HMO offered by Blue Shield of California; a vertically integrated HMO offered by Kaiser Permanente; and a broad-network PPO plan offered by Blue Cross. Unlike the other plans, the Blue Shield HMO plan faced regulatory constraints and could not exclude major hospital systems without regulatory review.

Our main exercise quantifies what would have happened if the Blue Shield HMO plan was unregulated and could have configured its hospital network in order to maximize profits. We compare these outcomes to a situation in which the insurer was prohibited from excluding providers, or instead had to select the network that maximized consumer or social welfare. Absent regulation, this shows whether an insurer would choose a broader or narrower network than that preferred by others, and the circumstances under which regulation might be warranted.

We simulate adjustments to the hospital network of the Blue Shield HMO plan across 12 distinct geographic markets in California. To predict outcomes under these counterfactual scenarios, we use a model of how consumers choose insurance plans and hospitals (estimated in our prior work), and pair it with a new theoretical model of how an insurer can leverage exclusion in order to play hospitals against one another when bargaining over reimbursement rates. This framework allows us to explore how adjustments in an insurer’s network can affect insurance premiums, negotiated provider prices, and consumer spending patterns.


Research Findings

In most of the 12 geographic markets, we find that both consumers and a profit-maximizing insurer prefer the exclusion of at least one major hospital system—and in many instances multiple systems—from the HMO’s network. We also find that the insurer’s incentives to exclude are not primarily driven by steering or cream-skimming incentives, but rather by increasing its bargaining leverage when negotiating reimbursement rates with hospitals. Compared to having to contract with all major hospital systems, we predict that the Blue Shield HMO plan would be able to negotiate approximately 12 percent lower hospital prices on average across markets (with reductions up to 30 percent in certain markets) with a narrower network.

On average, consumers would benefit from this more selective network, as rate reductions would result in lower premiums and offset any harm from having fewer providers in-network. We also find that, on average, consumers would generally prefer an even narrower network than the one chosen by the insurer. This is partly driven by the fact that consumers had access to generous insurance plans that included many providers as alternatives to the HMO product, and would value having an even lower-premium, narrower-network option.

We find that requiring that the insurer engage in negotiations with all major hospital systems would result in higher hospital payments and premiums and reduced consumer surplus on average. While consumers may benefit from the cost-savings generated by narrower networks, we show that the gains are not distributed evenly, and some consumers might be harmed. Certain consumers, particularly those who lived closer to excluded hospitals, would benefit significantly from prohibiting exclusion—by as much as nine percent of their annual out-of-pocket premiums. Those who face higher premiums without benefiting as much from the broader network would be worse-off by similar magnitudes.



Though we caution that many of our quantitative findings are likely to depend on the specifics of our particular setting, there are several conclusions from our analysis that we believe to be more generalizable.

First, there will likely be significant distributional consequences associated with any type of provider exclusion or network regulation: consumers, on average, may benefit from the cost reductions associated with narrower networks, yet significant harm can still be done to those who value the excluded providers most.

Second, the presence of multiple insurers competing for enrollees may prevent insurers from excluding too many providers from their networks. This suggests regulation may not be necessary in an adequately competitive market, and policymakers should remain focused on fostering insurer competition.

Last, given the presence of adequate competition, an insurer’s decision to exclude a medical provider may often be in the interest of consumers and their employers, particularly when it funnels enrollees to higher-quality, lower-cost providers. This, in turn, reduces rates and premiums. Regulators considering interventions to protect disadvantaged consumers should be wary of preventing these parties working together—potentially on customized networks or other types of sophisticated plan design—to control health spending and costs.


This article summarizes ‘Equilibrium Provider Networks: Bargaining and Exclusion in Health Care Markets’ by Kate Ho and Robin Lee, published in the American Economic Review in February 2019.

Kate Ho is at Princeton University. Robin Lee is at Harvard University.

[1] An Associated Press survey in March 2014 found, for example, that Seattle Cancer Care Alliance was excluded by five out of eight insurers on Washington’s insurance exchange; and MD Anderson Cancer Center was included by less than half of the plans in the Houston, TX area. See “Concerns about Cancer Centers Under Health Law”, US News and World Report, March 19 2014, available at http://www.usnews.com/news/articles/2014/03/19/concerns-about-cancer-centers-under-health-law

[2] For a previous working paper that assesses this incentive in detail, holding provider prices fixed, see Mark Shepard, 2019. “Hospital Network Competition and Adverse Selection: Evidence from the Massachusetts Health Insurance Exchange.” NBER Working Paper #22600.

[3] Kate Ho and Robin S Lee. 2017. “Insurer Competition in Health Care Markets.” Econometrica, 85: 379–417. See also, “Health insurance competition: effects on premiums, hospital rates, and welfare,”