Center for Studying Health System Change

Providing Insights that Contribute to Better Health Policy


Insurance Coverage & Costs Costs The Uninsured Private Coverage Employer Sponsored Individual Public Coverage Medicare Medicaid and SCHIP Access to Care Quality & Care Delivery Health Care Markets Issue Briefs Data Bulletins Research Briefs Policy Analyses Community Reports Journal Articles Other Publications Surveys Site Visits Design and Methods Data Files

Association Health Plans and Alternative Ways to Increase Health Insurance Coverage Among Workers in Small Firms

Congressional Testimony
Feb. 5, 2003
Len M. Nichols

Statement of Len M. Nichols, Ph.D.
Vice President, Center for Studying Health System Change

Before the Committee on Small Business and Entrepreneurship
U.S. Senate

My name is Len M. Nichols and I am the Vice President of the Center for Studying Health System Change (HSC). HSC is an independent nonpartisan policy research organization funded solely by the Robert Wood Johnson Foundation and affiliated with Mathematica Policy Research. Our recurrent nationally representative surveys of households and physicians, our site visits to monitor ongoing changes in the local health systems of 12 U.S. communities, as well as our monitoring of secondary data and general health system trends all enable us to provide policy makers with unique insights on developments in health care markets and their impacts on people. Our various research and communication activities may be found at

As an economist, I have studied the decisions of employers, and specifically small employers, to offer health insurance or not, as well as the general workings of small group insurance markets for the past 10 years. My research ranges from statistical analyses with nationally representative survey data gathered from employers to interviews with small employers, large employers, small business coalitions, insurers, insurance brokers, actuaries, state regulators, purchasing cooperatives, state legislators, and most recently site visit research conducted by myself and colleagues at the HSC.

I am sure this committee is well aware of the basic fact that small employers are much less likely to offer health insurance than are large firms. For example, in the most recent data, 47% of establishments with fewer than 50 workers offer health insurance, compared to 97% of establishments with more than 100 employees.

There are many specific contributors to this disparity in offer rates, but one conclusion stands out in all my research and in the professional literature of economics as a whole: employers offer health insurance if they think they need to in order to successfully compete for workers. If they do not offer health insurance, by and large it is because they can attract and retain the workers they need without offering it.

Most of us, happily, have enough education and training to work in labor markets where health insurance is a normal and expected part of any and all compensation packages. Seventy nine percent of workers are eligible for the health insurance their employers offer. Indeed, most of us have never had a job offer without some kind of health insurance attached.

But while a distinct minority, some workers, typically those without much education or marketable job skills, often can only get jobs with no health insurance attached. And some firms, indeed a disproportionate share of small firms, mostly need the skills of these kinds of workers. These workers, along with the sometimes prodigious efforts by the small business owner and his or her family members, are sufficient to generate marketable products or services for many small businesses’ customers. In competitive markets, if more skilled and highly compensated labor is not required, pretty soon profit margins are driven down to where more expensive labor actually could not be paid for, and that is when and why most small businesses who do not offer health insurance answer the question, "Why not?" with responses like, "Health insurance costs too much to provide to my workers. My business and I can’t afford it." This kind of response is the source of the shorthand but reasonable conclusion that cost is the single most important thing when it comes to health insurance and small business.

Now, in economic theory and even in real life, workers who do have health insurance through their employer implicitly trade at least some of what they could have had in wages in exchange for their employers’ premium contributions. (Whether wage offsets completely finance all employer payments is not settled in the economics literature, but this dispute is relatively academic for our purposes today, for the larger point is beyond dispute: all workers at offering firms trade some wages for employer premium payments, whether they know it or believe it or not). Workers with relatively low productivity or value added in the competitive marketplace, unfair though their compensation may seem from some perspectives, simply are not willing or able (given low discretionary income and competing resource demands) to trade some of their already low wages for health insurance, and this is how it comes to pass that small firms with low profit margins which employ these kinds of workers simply cannot and need not provide health insurance to remain in business.

So, the "root cause" reason some firms choose not to offer health insurance is that their workers’ wages are just too low (relative to the current cost of health insurance). But wages and compensation are tightly linked to the market value of a worker’s productivity, so it must be that their market productivity and purchasing power is simply too low to enable them to afford health insurance. It follows, then, that an efficient way to attack the root cause of lack of health insurance among workers in small firms, and among low income workers generally, is to increase their purchasing power with direct subsidies to them. Only with this increased purchasing power will they be willing to purchase health insurance for themselves and their families. This is the conclusion reached by two recent specific analyses of this question. These subsidies, by the way, could take the form of tax credits or other sorts of assistance, including eligibility for existing or modified public insurance programs like Medicaid or SCHIP, but those details and debates are also for another day.

For the agenda item at hand is association health plans. Where might they fit into this discussion, why are their advocates so strongly in favor of them as a solution for small businesses’ health insurance woes, why are their opponents just as convinced that this type of reform would do more harm than good, and what other non-subsidy alternatives to AHPs might this committee want to consider?

Before answering these questions, let me provide just one more background or contextual statement. My research over the years strongly suggests that small businesses have three primary goals when it comes to health insurance: affordability, simplicity, and stability. It is therefore efficient to examine AHPs and feasible alternatives in light of how well they might help small businesses reach these specific goals.

Affordability. It is well known that small firms have to pay more for the same health insurance policy than do large firms because the administrative costs of selling insurance can be spread over so many fewer workers, and thus premium loading factors - the difference between premiums and expected claims costs -- are higher for small firms, perhaps as much as 20-30% higher. Thus, any kind of larger group purchasing arrangement -- AHPs, small employer coalitions, allowing small firms to buy into state or federal employee plans or into Medicaid - could in theory lower the administrative costs per worker which now drive small business premiums higher than they have to be. Analysts might quibble about which of these types of group arrangements would lower administrative costs more, but it really depends on the details of each and the ultimate size reached, so as a first order approximation it may be reasonable to agree that AHPs can look attractive on this front and thus it should be no surprise they have received so much attention.

The second dimension to affordability that is often mentioned in AHP discussions is the cost of benefit mandates and the potential gains if they could only be avoided. The effect of benefit mandates is perhaps one of the most contentious issues of factual debate in the small group health insurance market. Careful academic research tends to find little to no net effect of benefit mandates on premiums or employer offer rates, yet some small employers, and particularly some small employer advocates, appear to be convinced that benefit mandates are the major culprit in health care costs today.

A few facts can help make sense of these competing and heartfelt claims. Most insured workers work for firms that are not small. Many medium sized and most large firms self-insure, outside the commercial market, and are not subject to benefit mandates. Nevertheless, most of the benefits that are mandated by states for their small group market are actually provided by large self-insured firms even though they do not "have to." Indeed, causation runs from large firm coverage decisions, which are driven totally by labor market competition, to state mandates on small firms, whose proponents argue for them as a matter of equity. In addition, many small firms that do offer are high wage firms, and they offer as rich a package as do the large firms, ordinarily. So, the bottom line here is that mandated benefits add relatively little to average insurance costs, and that is why most good academic research based on nationally representative data finds so little impact of them

At the same time, clearly benefit mandates must add to costs for those firms that were not offering particular benefits prior to the mandate’s passage. The Department of Insurance of the State of Texas did a very comprehensive study in 1998 which concluded that the 9 mandates in Texas, which are not atypical and include inpatient treatment of alcohol and substance abuse, represented about 3% of claims paid in the two most recent years available. I and colleagues, as well as the CBO in a later study both concluded that a reasonable interpretation read of the entire literature calls for an assumption that exemption from benefit mandates would save those firms that joined AHPs about 5% off their premium on average. Now 5% of a family premium these days is not a trivial amount of money, but please note that administrative load savings are likely to be 4 to 6 times larger as a percent of premium. It should also be noted that benefit mandates do add some value that should be weighed against cost. Research has shown that employee takeup of employer offers is higher in states with more mandates.

By far the most important determinant of the cost of health insurance in any alternative arrangement, of course, is the risk pool one has access to, or more precisely, the relative health risk of those which whom you are pooled. I will return to this important point below, for it is complex enough to be best addressed when we consider the stability of premiums.

Simplicity in health insurance is practically a contradiction in terms, and most small business owners must depend on agents or brokers to guide them through the inherently complex maze of details involved in health insurance choices today. Information and education functions are essential in any small group market solution, and agents may provide the cheapest and most widely trusted source of this information that anyone can devise. If any entity like an AHP or purchasing cooperative tried to avoid agents altogether, they certainly could, but they would have to then build in the costs of providing information about health insurance choices to all participating employer and employees. This will necessarily "take back" some of the administrative savings from self-insuring and forming a larger agglomeration of small groups. On the simplicity front, allowing firms to buy into existing employee pools, like state employee pools, might be the clear winner.

Stability. Small business owners are typically not in the health insurance business. They do not have enough time to master constantly changing details that are not directly related to their own production, sales and delivery problems. They want a basic package to be competitive in the labor market, and they do not want to have to worry about premium fluctuations that will force them to start over at ground zero with a new set of insurance options. For lots of reasons, premiums in the small group market are more variable than in the large group market. Just last week brokers reported to HSC researchers, including myself, of premium increases for small businesses in a given market that ranged from 10-90% in the last year. No business nor worker can sustain 90% increases of an item already as large as health insurance premiums are now. Now 90% might be an outlier, but there is considerable evidence that the variance of premiums over time for small groups is indeed larger than for larger groups, and so the basic interpretation of relative instability remains. Just hearing about these kinds of experiences, and enough small employers have experienced them that almost all have heard about the possibility over time, makes a small employer think twice about offering health insurance since the prospect of having to take an important compensation piece away in the future is more painful than continuing to live with not offering it as they have in the past. So stability is extremely important to any effective reform of small group options.

And it is precisely along the stability dimension that I fear AHPs earn the lowest marks. We learned on an HSC site visit just last week that in 2001 the Arkansas legislature passed a law which allowed small employers to join together and purchase insurance together as a large group and to avoid benefit mandates. But no insurer there would agree to offer coverage to these groups because of their fear of eventual adverse selection. That is, insurers were afraid that any premium rate that would cover the average costs of those who would be attracted to join together in an association-like arrangement in the first place - an association whose sole purpose was to purchase health insurance - would be higher than the average costs of specific groups. The insurers expected specific low-cost groups then to opt out over time and the insurers feared being left holding the deteriorating risk pool bag of the shrinking association plan. If the Arkansas insurers were correct, and it is telling that all reached the same conclusion, this kind of association could lead to rapidly increasing but actuarially fair premiums over time for well-intentioned members of the association plan. In the limit, this process is known as a death spiral.

Second, precisely because exemption from benefit mandates is such a strong motivation of proponents, many opponents of AHPs, including large insurers like Blue Cross Blue Shield plans which sell in the small group market, fear the opposite kind of selection into AHPs. They fear that more parsimonious benefit packages, and the ability to actively exclude small groups from the association and insurance product if not carefully proscribed, would siphon off all the good risks they need to keep in their blocks of business in order to keep their small group premium rates down. The CBO analysis of AHPs cited earlier estimated that 20 million workers would see their premiums rise if the AHP legislation of that day became law.

Aside from the obvious self-interested conflict here, there are fundamental issues at stake here. Essentially, self-insured AHP options, regardless of who runs them, could compete with commercial insurers and Blue Cross Blue Shield plans and anyone else - over who can package the best set of risk pools for the most small businesses. The analogy which proponents of AHPs sometimes invoke is that of large self-insured firms like General Motors or Xerox or Honeywell. Why should small firms not have the same power to self-insure? (They have the right now, but self-insuring a group with fewer than 100 members is rarely cost effective and highly risky).

Well, perhaps as a matter of fundamental liberty they should (and do) have the right to self-insure alone, but the analogy between AHPs and General Motors fails for a simple but powerful reason: General Motors self-insures all of its workers, it does not allow workers in California to select one self-insured pool and workers in Flint to select another, nor can workers in Flint select among competing risk pools (as some AHP proponents would allow). Whereas an AHP by construction would be a set of very small employers, any of whom could bolt or join at open enrollment time or at will if another insurer was willing to take or free them. Thus the inherent stability of any AHP risk pool cannot be as great as the largest self-insured employer pools can maintain. Either the self-interested search for more homogenous risk pools and insurance products among low risk firms will exert constant pressure for the pool to deteriorate, as the Arkansas insurers feared (perhaps new AHPs will repeatedly be re-constituted as a best case scenario), or the AHP will likely attract the lowest risk businesses who are willing to forego mandated benefits (in the mistaken belief that down that path lies huge savings) and the commercial risk pools would deteriorate, as the Blues fear and CBO predicted. The analysis I and my colleagues did at Urban a few years ago (cited earlier) suggests that this fear of commercial risk pool meltdown is probably exaggerated by some, but some deterioration is inevitable and the risk of large effects is certainly not zero.

There is one other concern about AHPs that can be solved but needs attention and that is the matter of reserves and regulatory oversight of those reserves, since they would not be typically regulated by state insurance departments. There have been some relatively rare but painful episodes of multiple employer insurance arrangements in the past being operated by people who took the premiums and left employers and workers with an unenforceable guarantee of the coverage they had paid for in good faith. I am no lawyer but have talked to enough to caution you to pay particular attention to these details in any enabling legislation to ensure the creation of adequate reserves and the operation of some kind of guaranty funds oversight mechanism in the event of miscalculations or financial meltdown or outright fraud, which does indeed happen in the real world.

So, are there better ways to provide small businesses with more affordable, simple, and particularly more stable health insurance options?

You may find it useful to have your staff review the set of proposals in an RWJF funded effort, spearheaded by Jack Mayer of the Economic and Social Research Institute, which were compiled into a book entitled Covering America. There you will find quite a few specific ideas about how better pooling and purchasing arrangements - specifically for small employer groups and for those relatively few uninsured with no employment connection at all - may be constructed from existing institutions. I was one of the co-authors of one of the proposals, the risk sharing theory of which has been accepted for publication in an upcoming issue of the American Economic Review, and I am also on the Advisory Panel to the Covering America project and so am familiar with the arguments advanced by the other nine proposal teams. I will summarize some options that come out of that work for you, omitting much subsidy mechanism, cost-control, and quality enhancement detail that your staff may want to peruse at some length later.

Allow small firms to join existing state employee pools. In many states, the single largest employer is the state itself. As such, states are often able to offer their employees a choice of plans and competitive low-load rates that are rarely possible for small firms. Many states also allow counties and even smaller administrative units to opt into their state employee plan. It would be fairly easy to allow small businesses to bring their employees into the state pool. Enrollment forms could be sent to small businesses with their tax forms each year, made continually available on web sites which most if not all states maintain now, and contributions - shared as employer and employees agree to on their own -- could be sent into the state monthly along with income tax withholdings. State employees might fear that small businesses who would join would be sicker than those who would remain outside the state pool, and that would be a risk, but two facts should calm. First, state employees tend to be older and sicker than workers in general. Second, since HIPAA imposed guaranteed issue for all products in small group markets, there is relatively little underwriting at the small group level anywhere any more, at least considerably less than there was prior to 1997 when typically only two products were guaranteed issue. This means any new firms likely to add coverage post-reform are more likely to be relatively low risk, and not high risk since universal guaranteed issue has already pulled the higher risk into the small group market. Finally, one could imagine requiring that all small employers purchase health insurance through the state pool, if they chose to offer health insurance (I am not advocating an employer mandate), which would purge any remaining fear or risk of adverse selection against the combined state employee-small business pool. This option would provide the maximum stability, simplicity, and affordability of all the options I can think of, with the added bonus of adding considerable choice of private health plan options for small firms’ employees, something very few of them have today.

Allow small firms to form purchasing cooperatives for the purpose of buying health insurance. They have this right in most states now, and while some work well, this movement has surprised analysts with how it did not exactly take the small business sector by firestorm, despite some obvious advantages over going it alone in small group purchasing. There are many reasons for the disappointment, but the relevant advantages of the best of these vis a vis AHPs is that they have similar insurance rules as the outside market - AHPs by construction would be exempt from mandates and perhaps able to underwrite more aggressively as well - and they could achieve critical administrative economies of scale and sufficient size to reach risk pool stability over time. A fair read of our experience to date, however, must admit that pools formed wholly by small employers are not likely to be as large or as stable as pools that would marry state employees and small employers together. Federal policy makers could insure that any pools that are formed are governed by the same market rules as prevail outside the pools in each state, but short of subsidies there is not much policy can do to ensure critical mass.

Allow small firms to buy their employees into Medicaid or SCHIP or some new hybrid state-employer program. This option is more attractive for extremely low wage workers who might actually be eligible - for themselves or for their children - for public insurance today. The concept is similar to allowing employers to buy into the state employee plan. This is a bit more complex because the benefit package for Medicaid is typically more comprehensive than the private employment-based plans that are offered to state employees, and policy makers may prefer to allow small employers of low wage workers to somehow buy less generous plans, or provide subsidies to help them afford them afford the more generous Medicaid package. Plus this option requires employers to deal with Medicaid stigma issues, which some feel more strongly about than others. Opportunities to reduce stigma are manifest in many SCHIP plans and the Basic Health Plan in Washington state, for example. Also, Medicaid enrollees are probably more expensive than small employers’ workers and their families, and Medicaid typically pays providers lower than other payers, so these this too could create some start-up costs. Still, allowing small employers to opt-in to the Medicaid purchasing apparatus would clearly offer administrative savings and a stable risk pool compared to buying insurance alone in the small group market.

I would conclude by iterating the most effective way to bring about more coverage of workers in small firms is to subsidize workers directly. I would be glad to elaborate on those options at the convenience of the Committee. My bottom line judgment is that AHPs do not score as well as allowing small firms to buy into state employee plans on the three criteria I think small businesses care the most about, affordability, simplicity, and stability. However, as in all policy choices, there are complex tradeoffs involved in any change along these lines, only some of which are amenable to technical analysis or which can be articulated in brief testimony on one day.

I would glad to answer any questions this testimony may have raised, now or at a future time.


Back to Top
Site Last Updated: 9/15/2014             Privacy Policy
The Center for Studying Health System Change Ceased operation on Dec. 31, 2013.