Search tips
Search criteria 


Logo of hsresearchLink to Publisher's site
Health Serv Res. 2004 August; 39(4 Pt 2): 1159–1166.
PMCID: PMC1361060

Commentary—Defined Contribution Health Plans: Attracting the Healthy and Well-Off

Driven by a philosophy that favors unbridled faith in the free marketplace, the year 2003 may well go down in health care history as the year that the health care system officially abandoned the premise that the community has a responsibility to care for each member, replacing it with the philosophy that individuals should each look after themselves.

The most visible change that nudges the system toward self-insurance is the provision in the Medicare bill that expands and makes permanent “health savings accounts” (HSAs) (formerly known as “medical savings accounts” or MSAs). This provision allows most Americans to set up tax-advantaged savings accounts (no tax is paid when money is paid in or when paid out, an unprecedented new tax loophole), when they also have a high-deductible health insurance policy. These new accounts are likely to favor the healthy (who stand to benefit financially from a new tax shelter since their accounts need not be depleted on health care expenses) and the wealthy (the higher tax brackets mean higher tax benefits).1 In his State of the Union address, President George W. Bush's proposal for a new tax deduction for premiums for high-deductible policies introduced the possibility that health savings accounts' penetration of the marketplace—and the demise of the employer-based health care system—will be accelerated.2

The second development is the encroachment of so-called consumer-driven health care plans (CDHC) into the employer-based health insurance marketplace. This new approach is dressed up with a consumer-friendly name, but in reality, as noted in Christianson, Parente, and Feldman (2004, this issue), this new approach is characterized by higher deductibles for employees. A more apt label, and one that seems to have been overtaken by CDHC, is “defined contribution health care.” As a gentle reminder to health researchers and policymakers that a consumer-friendly name should not be used to mask a marketplace change that may be harmful to consumers, I will use the “defined contribution health plan” (DCHP) label to refer to these new plans. “Defined contribution” accurately connotes limited employer liability for health care costs. “Consumer-driven” implies that the consumer exerts considerable control—hardly an accurate portrayal of high-risk consumers' likely experience with a high-deductible plan.

The two studies raise red flags about the potential for these new plans to appeal disproportionately to the healthy and those with high income. They contribute to the dangerous distraction of policymakers from the goal of working toward a health care system that provides affordable, quality health care to all by spreading costs broadly and fairly across the community.

Comments on Study 1 (University of Minnesota)

Study 1 (Christianson, Parente, and Feldman 2004, this issue) considers the experience at the University of Minnesota, when 16,000 employees were offered several health insurance choices, including policies that combine relatively high-deductible health insurance coverage, a personal care/health care savings account check, and a gap between the amount contributed to the account and the deductible, assuring that employees would face some out-of-pocket costs before their health insurance policy provided coverage. This study does nothing to make DCHP appear to be consumer-friendly and confirms concerns about what a shift toward DCHP will mean for the health care system. This section summarizes and considers some of the key findings.

DCHP Appeals Disproportionately to People with Relatively High Income

The average income for employees who enrolled in DCHP (and responded to the survey) was 48 percent higher than the income for employees who did not enroll in DCHP ($71,406 versus $48,148) (Christianson, Parente, and Feldman 2004, Table 1, this issue). This wide disparity lends strong support to the notion that higher-income individuals are more likely to enroll in a high-deductible health insurance plan in which they could be at risk of large out-of-pocket costs before meeting a deductible.

DCHP Appeals Disproportionately to a Relatively Sophisticated Population of Faculty Members and Does Not Appeal to Union Members

Thirty-six percent of DCHP enrollees were faculty members; only 14 percent of non-DCHP enrollees were faculty members. Participants in the civil service/bargaining unit were more likely to favor non-DCHPs: 50 percent of enrollees in non-DCHPs were civil service/bargaining unit members, while only 23 percent of DCHP participants were. The DCHPs appeal disproportionately to relatively sophisticated participants (Table 1).

An Overwhelming Majority (96 percent) of Employees Favor Low-Deductible Coverage to DCHP, Based on Their Choices in the Marketplace

The low participation rate in DCHPs indicates that there is no groundswell of consumer demand favoring a health care system centered on high-deductible health insurance: 4.3 percent of the eligible population participated in the DCHP program. (This assumes that families do not have more than one employee eligible for this coverage. A total of 695 employees—349 individuals and 346 families—enrolled, out of a total population of 16,000 employees.)

The Study Design Is Inadequate to Allow Conclusions about Risk Segmentation by DCHPs

The study uses a self-reported measure of chronic illness to study the potential for risk fragmentation, and finds no significant difference among DCHP and non-DCHP enrollees. This measure is insufficient to draw a conclusion on risk fragmentation. A more in-depth measure of health care costs, possibly a time-series, for all covered individuals in each family is needed. The measure used does not take into account whether employees might anticipate certain health care costs in the future (e.g., a planned pregnancy, elective surgery), which would discourage enrollment in a DCHP for fear of high out-of-pocket costs. Some health conditions might have regular costs associated with them, but respondents might not consider them to be a chronic illness (e.g., back pain) but more of a chronic condition. This is an area where further expansion of the underlying health status of respondents is critical.

The Satisfaction Level with DCHPs Is Not Impressive

While respondents in DCHPs were somewhat less satisfied than respondents in other plans (7.46 versus 7.55, on a scale of 0 to 10, 10 is best), the difference can be considered trivial even if technically statistically significant.

Internet Support Tools, a Key Selling Point of DCHPs, Were Used Only Moderately

While 30 percent of respondents in DCHPs used provider directories, only 8 percent used disease management information, and only 12 percent used pharmacy-pricing tools. These numbers do not support the premise that DCHPs mobilize employees to comparison shop and access Internet resources to manage their care and control costs.

Overall, the first study paints a picture of highly educated and high-income faculty members gaming the health care system by selecting into the high-deductible plan if they believe that they will come out ahead financially. The limited measure of health status precludes drawing conclusions about the segmentation of the health risk pool, but overall there is nothing in this study to dispel the concern about risk fragmentation. Perhaps the strongest conclusion from this study is that DCHPs appeal disproportionately to highly educated, high-income members of an employee group. They appeal to a tiny portion of employees. The small fraction of employees who enroll do not make full use of the tools that they offer, and are not particularly satisfied with the plans' performance.

Comments on Study 2: Humana Employees

Study 2 (Fowles et al. 2004, this issue) reports the results of a survey of 4,680 employees of Humana Inc., 7 percent of whom selected a new “consumer-defined health plan option” (referred to as DCHC below). This is the epitome of a “defined contribution health plan”: the employer would pay a fixed amount, 79 percent of the reference plan, for each employee. This study provides troubling confirmation of the potential of DCHPs to fragment the health risk pool to the detriment of the less healthy.

Those Selecting DCHP Are More Likely to Be Healthy

The study found that enrollees in DCHP were “significantly healthier on every dimension measured.” This study used a more comprehensive measure of health status, including measures such as reported health status, likelihood of a covered member receiving regular medical treatment, likelihood of having a personal physician, and existence of a chronic health problem. Those who selected the DCHP were less likely to have a chronic health problem (54 percent) and more likely to have had no recent doctor visits (3.07). Enrollees in DCHPs were more likely to be in excellent health (31 percent versus 18 percent) (Table 1). The study found that employees reporting that a family member had a chronic health problem were half as likely as others to select the DCHP.

Enrollment in the New Plans Was Modest

Like the University of Minnesota employees, the Humana employees did not flock to the high-deductible coverage (despite the annual premium savings of $400 per year for an individual and $1,200 per year for a family): only 7 percent enrolled in the new plan. Individuals were more likely to enroll in a DCHP than families.

Sociodemographic Findings

Those enrolling in DCHPs were more likely to be college-educated, white, male, and in positions exempt (from a union) than those who enrolled in other plans. The finding that blacks are about half as likely to enroll in DCHPs is troubling, and suggests that just as policymakers are waking up to the magnitude of disparities in our health care system, yet another policy that separates blacks (and presumably other minorities) from whites is created. Income is not listed as an independent variable, ruling out the ability to estimate the relative importance of race and income.

This study clearly demonstrates that widespread expansion of DCHPs within the employer marketplace will fragment the risk pools in the employer-based health insurance marketplace, one by one. Employer-based health insurance coverage has been held up as the one place in which risk pools tended to be unified, with costs spread among employees (albeit paid directly in large part by employers). DCHP's have the potential to unravel this important risk-spreading role. This study clearly demonstrates that risk segmentation, to the advantage of the healthy and the disadvantage of the less healthy, will be a reality should the role of DCHPs expand in the health insurance marketplace.

Implications of the Studies for Public Policy

Members of the public and policymakers should view these two studies as the proverbial canary in a coal mine. They raise red flags about the potential that DCHPs (like their cousins Medical Savings Accounts) appeal disproportionately to the wealthy and healthy. The first study shows that the income level of employees selecting DCHPs is 48 percent higher than those not selecting them. The second study finds that those selecting DCHPs are healthier “on every dimension” than those not selecting them. The concern that this new model of health care will appeal more to the sophisticated who can “game the system” and shift costs to the sick becomes greater after reviewing these studies. They should set off alarm bells about the potential long-term threat to our health care system.

The scope and design of these studies did not allow consideration of some of the most important issues that will affect the long-term impact of this new type of plan. Some important areas for future research include:

  • To what extent will DCHPs merely shift cost to sicker employees, instead of truly lowering health care spending?
  • Over time, will sophisticated employees “game the system,” opting out of DCHPs when they anticipate high health care expenses related, for example, to pregnancy or elective surgery?
  • To what extent will employer's health care premium dollars be diverted from paying for health care expenses to paying to build health reimbursement accounts?
  • To what extent do these new health plans create new financial barriers to health care for low-wage workers?
  • Do consumers have the necessary information about quality of providers on which to make informed decisions?
  • What are true consumer/employee preferences regarding deductible levels?
  • To what extent will the gap between the health reimbursement account and the deductible pose a financial barrier to getting needed health care?
  • Will anticipated cost savings occur, or will they fail to materialize since so much health spending is concentrated among those with catastrophic expenditures?
  • Will the new high deductibles and sense of spending one's own money deter preventive care and early treatment for illness, ultimately leading to worse health outcomes and higher costs?

The findings from these two studies are troubling for another reason: because of the nature of adverse selection, over time, DCHPs may drive lower-deductible health insurance options out of the marketplace (Zabinski et al. 1999). Bolstered in the health care market with the enactment of the health savings account provision in the Medicare bill, in a few short years, it is very possible that unpopular high-deductible health insurance coverage will be the only choice that many employees may face for their coverage in the employer-based market. Those with high health care expenses will face higher out-of-pocket costs than they would in the absence of DCHPs. It is troubling that this type of change in the health care marketplace will take place in the absence of a public debate. Advocates of medical savings acccounts, for example, maintain that there should be a choice of plans. The reality is that over time, as adverse selection pushes the next “relatively healthy” group toward high-deductible plans, an insurance marketplace death spiral will result and ultimately will remove the very choice (a low-deductible plan) that employees want.

Both studies contribute to the body of knowledge about DCHPs, “as a first, limited attempt to shed light on the important issues” (Christianson, Parente, and Feldman 2004, this issue). In considering the health policy expertise and money devoted to these studies, it is important for health researchers and policymakers to ask fundamental questions about priorities for future health research. The buzz about DCHPs in health policy circles creates a sense that valuable dollars are being spent in an effort to rearrange the deck chairs on the Titanic. More resources should be devoted to charting the course to guarantee all U.S. consumers have guaranteed, quality, affordable health care. We should be moving full-steam toward this vision, not spending countless hours and resources analyzing new models that promise to split the healthy from the sick, shift costs to the sick, favor the highly educated and high-incomed, and grow the inequities on our system. The two studies confirm that DCHPs are a dangerous distraction from this mission; they undermine the important value of a community-wide approach to looking after one's neighbor in a health care system that would spread costs broadly in an effort to achieve affordable, quality health care for all.


1In addition to benefiting from a higher tax bracket (and higher tax benefit from HSAs), the wealthy are more likely than the nonwealthy to be able to risk the out-of-pocket costs of a high-deductible policy.

2Because healthy individuals may be able to get a lower premium for a catastrophic policy in the individual market, the new tax deduction available to individuals, when combined with the possibility that employers will increasingly “cash-out” health benefits when the healthy opt-out of coverage, could lead to rapid erosion of the employer-based health insurance market.


  • Christianson JB, Parente ST, Feldman R. “Consumer Experiences in a Consumer-Driven Health Plan” Health Services Research. 2004;39(4, part 2):1123–40. [PMC free article] [PubMed]
  • Fowles JB, Kind EA, Braun BL, Bertko J. “Early Experience with Employee Choice of Consumer-Directed Health Plans and Satisfaction with Enrollment” Health Services Research. 2004;39(4, part 2):1141–58. [PMC free article] [PubMed]
  • Zabinski D, Selden TM, Moeller JF, Banthin JS. “Medical Savings Accounts: Microsimulation Results from a Model with Adverse Selection” Journal of Health Economics. 1999;18:195–218. [PubMed]

Articles from Health Services Research are provided here courtesy of Health Research & Educational Trust