High premiums in non-group insurance markets

Identifying Causes and Possible Remedies

By Linda J. Blumberg and John Holahan
Urban Institute, January 2017

The Issue

The authors note that premiums are not high in all areas, and that many areas with large recent premium increases reflect needed corrections to very low premiums in the early years of reform. However, in sources with high premium levels, the causes differ, as should appropriate solutions.

Key Findings

The first potential problem is that sicker people are more likely to enroll in marketplace plans;

The second, premiums tend to be higher if an insurer or provider group has a monopoly in their area; and

Third, inadequate risk adjustment, i.e., moving premium dollars from insurers with low-cost enrollees to insurers with high-cost enrollees, may also be a factor leading to high premiums in some areas.


Marketplaces in many areas throughout the United States have experienced high premiums and insurer attrition for the 2017 plan year. In this brief, we identify three potential sources of these difficulties, not all of which exist in all markets, and we emphasize that some markets continue to work competitively and without any significant difficulties.

First, some ACA-compliant nongroup insurance markets have enrolled a population with above-average health care risks relative to a broad cross section of the nonelderly population (e.g., the population enrolled in employer-based insurance). In these cases, above-average claims have led to high and increasing premiums and losses for some insurers that were large enough to force them out of the market. Second, some markets have high insurer and/or provider concentration. The exercise of monopoly or near-monopoly power in these markets can lead to high provider payment rates and high premiums. Third, in some markets, the risk-adjustment system may not have sufficiently shared health care risk across insurers within the nongroup insurance market, either moving too much money away from lower-cost insurers or moving too little money toward higher-cost insurers. In such cases, insurers may have increased premiums to compensate for their losses. An inadequate risk-adjustment system could also induce all insurers in a market to increase premiums because they fear that they may be selected against but not receive appropriate compensation through the risk-adjustment mechanism.

In this paper, we argue that each problem requires a different approach. Some policies may be implemented in some states and substate areas and not in others. However, a more uniform national strategy could incorporate a range of policies that collectively could address all these problems, with policies designed to be binding only in the specific areas where they are warranted. Because different markets have different problems, some policy solutions will be essential in some markets and have little or no impact in others. Implementing these policies nationally would obviate the need to design and legislate policies on an ongoing basis as market conditions change over time.

To address the problem of low enrollment and adverse selection against the entire nongroup market in an area, first, we suggest a significant increase in outreach and enrollment assistance and more generous premium tax credits and cost-sharing assistance. Second, penalties for not obtaining health insurance may also be increased, once coverage is made more affordable with increased financial assistance. Third, there should be a permanent reinsurance program funded from a broad-based source (e.g., general revenues or a tax on all insurance plans and stop-loss plans sold to self-insuring employers) that would subsidize nongroup insurance market plans with extremely high claims (e.g., $1 million or more).

Fourth, there should be increased regulation of sales of non-ACA-compliant nongroup insurance products (e.g., short-term policies, disease-specific policies, indemnity plans). These products are most attractive to those in relatively good health. Phasing them out, eliminating them entirely, or assessing them in order that they share in the risk of the ACA compliant market, will make individuals more likely to purchase ACA compliant plans and improve the risk pool. Finally, fixing the family glitch and lowering the Medicaid eligibility threshold to 100 percent of the FPL would also increase marketplace enrollment while improving the private nongroup insurance pools’ risk profiles.

In principle, the problems of insurer and provider concentration could be solved with a public insurance option, but this would be extremely difficult to implement. For starters, the government would have to develop a large new organization, and the approach is certain to face opposition from the insurance industry. An alternative that could be more politically palatable (at least to insurers) is a cap on provider payment rates paid by any ACA-compliant nongroup insurer within a particular market (e.g., Medicare plus some percentage). Insurers could develop networks at lower provider payment rates, but providers would be prohibited from charging nongroup insurers above the statutory limit. This would be more attractive than a public option to insurers, though many providers would probably be opposed to any legislation limiting their payment rates (even though these limits would apply to the fairly small population enrolled in nongroup coverage). In markets where provider payment rates are already low (e.g., those with high levels of competition), this policy would have little or no effect. But it could have a significant effect in markets where few insurers compete or where providers have significant market power and insurers have little or no negotiating leverage.

Problems with the risk-adjustment mechanism may persist even after CMS implements improvements in 2017. The inclusion of prescription drug use in the risk-adjustment formula is a step in the right direction. Other improvements are beyond the scope of this paper and require additional research. However, we believe the approach proposed by the Obama administration for 2018, which would add a reinsurance-type component for insurers incurring extremely high claims in a given year, could be very helpful.

Ideally, the federal government would adopt an array of these policies and apply them nationwide. As noted above, not all would be binding in all markets in any particular year. But implementing each policy on an as-needed basis in each particular time and place would leave the overall program vulnerable to political challenges, market instability, and logistical delays. However, in the face of federal inaction, individual states could choose to implement the specific policies that apply to them. These mechanisms should lead to increased enrollment, a fairer distribution of risk among insurers, greater insurer participation in ACA-compliant nongroup markets, more stable risk pools, and more predictable increases in premiums year to year.

Full report (9 pages):



By Don McCanne, M.D.

Linda Blumberg and John Holahan explain why premiums in non-group insurance markets can be high, and they suggest policies that can improve the stability of premiums. The proposals would increase the administrative complexity in our system when we are already tremendously overburdened with administrative excesses, and, further, they are designed to protect the private insurance industry - an approach in which providers and patients are considered a nuisance which must be accommodated to benefit the insurers.

The key findings - adverse selection, insurer or provider monopolies, and inadequate risk adjustment - would not be issues under a well designed single payer national health program. Our policy community is well meaning when they try to fix the the serious faults with our fragmented, dysfunctional financing system, but they do a disservice to all of us when they continue to exclude the single payer model from their policy briefs.