Blog Home


Bigger May Not Be Better: Does Scale Matter For Payors?

November 15th, 2013

Editor’s note: In addition to Shubham Singhal (photo and linked bio above), this post is authored by Rohit Kumar and Jeris Stueland. Rohit Kumar is a consultant in McKinsey’s Chicago office. Jeris Stueland, an expert in McKinsey’s Healthcare Systems and Services Practice, is also in the Chicago office. The authors would like to thank Ellen Rosen, Jim Oatman, and Michael K. Park for their contributions to this article.

This is the second in a periodic series of posts by McKinsey analysts on the landscape facing payors in the post-reform world. You can read the first post in the series here.

Whether scale brings competitive advantage to payors is a topic of hot debate. Many believe that consolidation is likely as the industry goes through the disruptive changes set in motion by reform. Some contend that anticipated margin compression and medical-loss-ratio floors will make scale efficiencies critical for achieving sustainable economics in the future. Others, however, note that managing the total cost of care is becoming central to a payor’s success, and question what advantages scale provides in such a world. If most health care is locally delivered, they argue, how much of the value created by cost-of-care management can scale drive?

Our research and experience suggest that for payors, the minimum threshold for efficient and effective scale is low. The primary rationale for scale emerges from the large fixed investments payors must make to develop the new capabilities needed to compete effectively in a rapidly changing regulatory and market environment (and to comply with evolving regulations). This rationale holds particularly true for payors that choose to build these capabilities themselves rather than through partnerships with external vendors, noncompeting plans, or other stakeholders in the value chain.

Yet, once the minimum level of scale is achieved, performance variability on administrative costs continues to be quite high. This suggests that for many payors the bigger opportunity to achieve administrative efficiencies is through operating model and organizational redesign, productivity enhancements, and application of design-to-value principles to core processes.

Minimum Efficient Scale For Payors

We used National Association of Insurance Commissioners (NAIC) filings to assess sales, general, and administrative (SG&A) costs for payors of different sizes across states. Exhibit 1 shows the industry scale curve we derived. We found that costs began to converge above 100,000 lives, a relatively modest level of aggregate scale. (Exhibits appear at the end of this post.)

Interestingly, it appears that, as scale increases, the incremental costs driven by greater complexity begin to counteract the economies of scale. As Exhibit 2 shows, payors with greater than one million covered lives tend to have more lines of business and to operate in many more states than smaller payors do, and they seem to have higher administrative costs. In our experience, smaller payors often have much greater standardization of products and processes, and are more likely to outsource IT platforms and core functions. Because their business is less complex, they often appear to be better able to make the most of the efficiencies derived from economies of scale.

In a post-reform world, the disadvantages of complexity could further increase. For example, as the public exchanges and much talked-about private exchanges take hold, large payors will have to be able to offer products and pricing in numerous different geographic ratings areas if they want to reach the consumers buying insurance on those exchanges. (California alone has 19 rating areas.) In this scenario, local and regional payors may be better positioned to capture opportunities in the various rating areas, since they may have greater flexibility to customize locally.

If significant numbers of large employers opt to move to defined contribution models, the disadvantages of complexity could become even greater. Companies that today cover all of their employees through a single national group plan or administrative-services-only product could decide instead to offer their employees a range of health coverage and supplemental insurance products at different price points; private exchanges could help the companies tailor the offerings to different geographic rating areas.

In addition, the ability to customize locally could give smaller payors an advantage in the growing Medicaid and dual-eligible businesses. Care management and provider collaboration are important for controlling the cost of care in those businesses, and understanding specific local health needs and establishing relationships with local providers can be crucial for executing these efforts well.

As a result, the pursuit of aggregate scale is unlikely to be fruitful for payors unless accompanied by material changes in operating model design.

Conduct Has Greater Impact On Administrative Costs Than Scale Does

Our analyses reveal that the impact of scale on lowering SG&A costs is even less evident within a single line of business. In the small-group segment, for example, we have found that some payors have per-member, per-month costs that are more than twice those of payors of a similar size (Exhibit 3). The conclusion is inescapable: some payors simply operate much more efficiently than other payors do, regardless of scale. In other words, conduct matters more than scale.

We have also assessed if there are differences in margins or earnings growth between payors that spend more on administrative costs and those that spend less. That is, are parsimonious payors shortchanging themselves by investing less in the key capabilities required to ensure business performance? As Exhibit 4 shows, we could find no correlation between the level of SG&A spending and either profit margins or profit growth.

In our experience, payors that operate more efficiently and perform better have put considerable thought into their operating models to ensure that the companies are optimized to deliver against the value proposition they promise to their customers and sales channels. These payors make sure that their resource allocations are ruthlessly prioritized to the capabilities that help the companies succeed against competitors. They also make careful decisions about the level of standardization or managed customization (e.g., through modular product design) in their offerings, and they ensure that their IT architecture is capable of delivering changes and new capabilities at low cost (e.g., through easy integration with standard third-party solutions).

In addition, these payors make thoughtful choices about the use of in-house versus outsourced processes and capabilities — they keep in-house those areas in which they can be distinctive or create a competitive advantage, and they outsource as much else as possible.

Benefit Of Scale

Scale does have its benefits, however. With reform, the level of uncertainty and thus the level of risk in the health insurance business have increased. Among the unknowns: likely uptake rates among the previously uninsured, the behavior of newly insured individuals and those with a change in insurance type, the composition of risk pools, the effect of risk adjusters, and the conduct of existing competitors and new entrants. We assessed the volatility in likely performance of differently sized books of business to understand the benefits of scale and found that scale does help to mitigate volatility (Exhibit 5). Smaller companies can purchase reinsurance to mitigate volatility but incur the additional cost.

An even stronger argument in favor of scale arises from the fact that today’s payors have a significant need for new capabilities if they are to compete effectively. For example, they must learn to operate in the new individual and small-group exchanges (as well as in the completely redesigned off-exchange market in those segments), cope with the emergence of private exchanges and small-group self-insurance in the employee benefits market, adjust to rate pressures and changes to revenue programs (e.g., star ratings and risk adjusters) in the Medicare market, and accommodate expanded Medicaid eligibility rules. At the same time, payors must accustom themselves to new payment models and learn how to comply with the new and often incremental regulatory mandates that some of the market changes are creating.

Each of these new capabilities requires significant investments in IT systems, new business processes, and talent. Payors that can spread the costs of these investments over more lives can often have a significant cost advantage. Furthermore, larger payors can invest in many more capability areas, thereby diversifying their risks until the uncertainty resolves. Exhibit 6 illustrates the effects of scale on two examples of one-time, fixed investments.

In sum, the value of scale is relatively small in terms of operating efficiencies, and many times those efficiencies are offset by increased complexity. However, scale can make it easier for payors to build superior capabilities and mitigate volatility. All payors, regardless of their size, should assess and consider evolving their operating models and organizational designs to unlock efficiency opportunities. Those payors that do opt to pursue scale should be thoughtful in how they undertake mergers or partnerships — the best approach to use will depend on the deal’s rationale. (More details about deal execution can be found in the white paper by Celia Huber et al, “Riding the next wave of healthcare payor M&A.” July 2013. For a copy of this paper, contact

EXHIBITS (Click to enlarge):

Exhibit 1

Singhal Exhibit 1

Exhibit 2

Singhal Exhibit 2

Exhibit 3

Singhal Exhibit 3

Exhibit 4

Singhal Exhibit 4

Exhibit 5

Singhal Exhibit 5

Exhibit 6

Singhal Exhibit 6

Email This Post Email This Post Print This Post Print This Post

 to the #1 source of health policy research.

No Trackbacks for “Bigger May Not Be Better: Does Scale Matter For Payors?”

2 Responses to “Bigger May Not Be Better: Does Scale Matter For Payors?”

  1. Bob Says:

    Volatility can be dealt with (at a cost) through reinsurance. More and more lines are subject to risk adjustment which also helps.

  2. Thomas Cox PhD RN Says:

    That this question can even be asked shows how much re-education is needed if sound policy is going to be feasible.

    The question comes down to this: Are large insurers more efficient risk managers than small insurers when the health care system is providing services efficiently? This question was answered centuries ago unless your perspective is dominated by Austrian School ideology.

    Given a population of potential policyholders P and different sized insurers, drawing policyholders at random from this population receiving health care services from efficient health care providers, who will have the greatest variation in their medical loss ratios, the large insurers, or the small insurers?

    Let’s suppose we have just two: A large insurer issuing 1,000,000 policies dispersed throughout a state and A small mom and pop insurer operating in a city with a similar epidemiological profile as the entire state.

    The key to insurance is the standard error for the loss ratio for an insurer. We can assume that the large insurer’s standard error is 1/10th as large as the standard error for the small insurer – A simple consequence of the Central Limit Theorem (See “Standard Errors: Our Failing Health Care (Finance) Systems And How To Fix Them” at my website.

    Both insurers have expected loss ratios equal to the population loss ratio – Let’s say that is 0.7500. Let’s assume they both have non-loss related operating costs of 15% of their premium revenues, seek profits of 5% and charge identical, market based risk premiums of 5%. We can assume that the larger insurer’s standard error is 0.0500 which works very well in real markets and in this example.

    The smaller insurer’s standard error is therefore 0.5000 (0.0500 * 10).

    Both insurers earn profits greater than 10%, at loss ratios less than 0.7500, in about half the years they operate because they both have normally distributed loss ratio probability distributions with mean values 0.7500. The Central Limit Theorem applies to insurers as well as small sample statistics. But that is the only loss ratio at which both insurers have identical probabilities.

    Both insurers will earn profits greater than 5%, at loss ratios less than 0.8000. The large insurer’s probability is 0.8413 because a loss ratio of 0.8000 is one standard error above the mean. But the insurer with 10,000 patients has probability 0.5398 of earning profits greater than 5% because a loss ratio of 0.8000 is only 1/10th of a standard error above the mean for its loss ratio probability distribution.

    The larger insurer’s probability of breaking even, at a loss ratio lower than 0.8500, is 0.9772 because a loss ratio of 0.8500 is two standard errors above the mean. But the smaller insurer’s probability of breaking even is only 0.5793 because a loss ratio of 0.8500 is only 2/10ths of a standard error above the mean for the small insurer.

    How about really high operating losses, 25% or more of the insurer’s premiums at loss ratios above 1.1000? The large insurer’s probability of a loss ratio that high is 0.0000 because a loss ratio of 1.1000 is 7 standard errors above the mean and there is virtually no probability that far out on a normal curve. But the small insurer has a significant amount of probability of loss ratios that high, or higher, 0.2419.

    The small insurer should anticipate catastrophic operating losses about one year in four because a loss ratio of 1.1000 is only 7/10ths of one standard error above the mean for its probability distribution.

    Then too, we have to consider conflagration risk. With policyholders distributed throughout an entire state a local health event is unlikely to effect the large insurer’s overall operating results. Not so for the smaller insurer. A local health event like a factory accident or virulent flu season could result in much higher costs in a small geographic region.

    This, of course, is all applicable to insurance risk assuming health care providers receiving capitation payments, or fixed payments through the Medicare/Medicaid Prospective Payment Systems, episode based care schemes and bundled payment schemes. Health providers make bad health insurers for the same reason small insurers do, they are very inefficient risk managers.

    Small insurers are simply inefficient insurers. They do not benefit as much from the homogenizing effect of the Central Limit Theorem. Unfortunately the differences in risk management efficiency between large and small insurers is so profound that no amount of operational efficiencies in the processing of claims is likely to influence these results. The key to insurance is having a loss ratio close to the population loss ratio , not having a lower loss ratio than the population loss ratio.

    The target for improved efficiency is the expense ratio – but that is only 15% of the insurer’s premium revenues and even if the small insurer operates with no non-loss operating expenses its exposure to risk is so great that its probability of a loss ratio greater than 10% remains at 0.2419.

    So the answer to the question – Does size matter for payors is YES! The core myth of capitation-like health care finance mechanisms is that there is no end to the potential for cutting health care delivery costs. But there are limits beneath which one cannot cut costs without cutting medically necessary and appropriate care. When efficient health care providers are paid through capitation, the capitation payments are either excessive (More than the average cost to insurers to provide such care in an equally efficient fee for service system), or the payments are inadequate (Less than the amount necessary for providers to meet the target goals assumed in the the capitation payments), making them inefficient almost 100% of the time.

    100% inefficiency doesn’t measure up to the lofty promises of capitation financed health care.

Leave a Reply

Comment moderation is in use. Please do not submit your comment twice -- it will appear shortly.

Authors: Click here to submit a post.