Consumers and Choice: the Forgotten Element in Health Care Cost Containment Policy

ABSTRACT - The responses of 365 potential consumers to three cost containment strategies in health care are examined. The aggregate sample proves negative toward two strategies and positive toward the third. Also, demographic characteristics which predict approval of one strategy predict disapproval of other strategies. These findings are discussed in the context of recent events in California.


Harris M. Allen Jr. (1985) ,"Consumers and Choice: the Forgotten Element in Health Care Cost Containment Policy", in NA - Advances in Consumer Research Volume 12, eds. Elizabeth C. Hirschman and Moris B. Holbrook, Provo, UT : Association for Consumer Research, Pages: 258-262.

Advances in Consumer Research Volume 12, 1985      Pages 258-262


Harris M. Allen Jr., The Rand Corporation

[Views expressed in this paper are the author's own and are not necessarily shared by Rand or its research sponsors. The author wishes to thank John Ware, without whose help this manuscript would not have been possible. The author is a consultant to the Health Sciences Program at Rand.]


The responses of 365 potential consumers to three cost containment strategies in health care are examined. The aggregate sample proves negative toward two strategies and positive toward the third. Also, demographic characteristics which predict approval of one strategy predict disapproval of other strategies. These findings are discussed in the context of recent events in California.


Numerous strategies to contain the spiraling costs of health care have been introduced in both the public and private sectors over the last two decades. These strategies have been extraordinarily varied in scope and have spanned a wide array of prescriptive approaches (competitive, regulatory, innovative, voluntaristic; McNerney 1980). By and large, however, they have been unable to blunt the persistent inflation of health care costs.

This paper explores one of the factors presumed by many observers to be responsible for this disappointing track record: consumer responses to containment strategies. To be effective, a cost-saving strategy must force a reduction in the overall utilization of the health care system by consumers. To do so, the strategy must impose new costs on consumers to develop cost-saving responses among the latter. Since those responses often require dramatic changes from established consumption patterns, the long term prospects of any strategy depend in part on how accurately it has anticipated the responses of consumers to its new costs.

Yet, over the years little systematic effort has been made to incorporate an understanding of consumer responses into the development of containment policy. Consumers have rarely been asked how they react to care delivered under a new and different containment strategy. As a consequence, consumer responses have often not been anticipated very well by providers and reimbursers seeking new cost-saving ways to deliver and finance health care.

In the discussion below I illustrate the use of one framework for addressing these concerns, using a study that probed potential consumer responses to three cost containment strategies: (1) use of insurance incentives to persuade consumers to choose providers who have contracted with reimbursors to provide services at a discount (preferred provider); (2) use of administrative guidelines to ration the availability of health personnel and resources (health planning); and (3) use of provisional arrangements to encourage consumers to take responsibility for personal health maintenance (self care).

Three key questions are addressed in the analysis. How do consumers respond in the aggregate to specific scenarios they are likely to encounter under each of the three broad containment strategies? What segments of consumers are likely to respond positively or negatively to the strategy scenarios? And, what do consumers think will happen to their care under each strategy?

This study finds that aggregate consumer responses are negative to two of the three strategies--preferred provider and health planning--and positive to the third-self care. Second, demographic characteristics which predict approval of one strategy predict disapproval of other strategies. No one strategy, therefore, is likely to be accepted by everyone.


A successful cost containment strategy for health care restructures patterns of care for consumers. In the course of this restructuring, consumers agree either to absorb novel and unfamiliar costs or to endure an increase in familiar costs. This section of the paper introduces the three containment strategies with an emphasis on the costs each is likely to generate for consumers.

Strategy I: Preferred Provider. The preferred provider strategy is a relatively new, yet swiftly evolving concept in health care. In theory, it is designed to take advantage of competitive principles. It seeks to increase "marketplace pressures" in health care by making both providers and consumers more accountable for the financial impact of their treatment decisions.

In practice, this strategy has produced a wide variety of organizational forms all of which are known as preferred provider organizations (PPOs). As a rule PPOs share the following description: "a group of fee-for-service providers who have a contractual agreement to provide health care services at a discount to a defined pool of patients who have free choice of provider but have an economic incentive to utilize PPO member providers ("Preferred Provider Organizations" 1982). The economic incentives to choose "member" providers are typically manipulated through insurance reimbursement arrangements (Enthoven 1978). For example, consumer coinsurance rates, copayments, and deductibles may be lower if "member" providers are chosen.

Preferred provider tactics, however, are not necessarily restricted to PPO fee-for-service arrangements. For example, reimbursors may require consumers to obtain estimates for a given service from multiple providers and base reimbursement on the lowest estimate. Such a tactic does not distinguish "member" providers from nonmember providers though it uses a similar coinsurance incentive to steer consumers to the preferred (less expensive) providers.

For consumers accustomed to traditional fee-for-service insurance arrangements, preferred provider tactics introduce a tradeoff between freedom of choice and out-of-pocket payments which in fact becomes a new cost. If the provider of their choice is not a preferred provider, they have to pay more. Accordingly, this study presented scenarios which described health care encounters in which restriction of choice for provider was emphasized.

Strategy II: Health Planning. In contrast to the preferred provider strategy, health planning is regulation-based. It attempts to coordinate the resources in organized health care so as to alleviate duplication and maldistribution of these resources. It assumes that coordination problems stem primarily from two factors: the independence of providers, particularly physicians and hospitals, and the multiplicity of reimbursors (Newman, Elliot, et al. 1978).

To address these factors, health planning seeks to impose resource limits on providers through a secondary layer of health provision which administers resources. This administrative structure supplements traditional cost sharing methods with either implicit or explicit rationing. Implicit rationing acts through constraint. It places broad limitations on the treatment decisions of providers via prospective budgets, need assessments, etc., but otherwise allows them to remain relatively flexible. Explicit rationing, in contrast, acts through restraint. It curtails provider autonomy through such mechanisms as pre-review, concurrent review, etc., all of which reflect an enhanced influence for the administrator (Mechanic 1978).

In practice, tactics for both types of rationing attack coordination problems by placing the initial burden for cost containment on providers, not consumers. For example, health administrators can minimize duplication of inpatient hospital bed supply and high cost technological equipment in a given area and thus limit the availability of these items to providers (implicit rationing). They can also mandate the curtailment of ancillary or "extra" services by providers (explicit rationing).

Both types of rationing do, however, generate secondary consequences for consumers, including increased inconveniences and reduced access to service. Since a major objective of health planning is to shift the burden of rationing from an economic to a non-economic basis, these consequences represent non-economic costs. Consumers, for example, may wait longer for appointments, experience longer delays in obtaining these appointments, etc., when seeking care under arrangements guided by health planning. Hence, the health planning scenarios in this study described situations which increase the access inconveniences that routinely occur when receiving care.

Strategy III: Self Care. This third strategy is based on recent innovations in health care. It seeks to tap a larger social movement in which the individual consumer participates more fully and accepts greater responsibility for the services received from professionals (Gartner and Reisman 1974). In health care this movement has produced a shift from the traditional medical model of "expert" physicians and "passive dependent" patients to the partnership model which emphasizes the many ways consumers can assume greater individual responsibility for health matters (Levin, Katz, et al. 1976).

Thus, in contrast to the other two strategies, self care seeks to place the initial burden of containment on consumers. Specifically, Fry (1978) has outlined four spheres of self care activity: (l) health maintenance, (2) disease prevention, (3) self diagnosis, treatment, and recovery from illness, and (4) patient participation in professional care. Expanded consumer activity in each sphere, if properly undertaken, is thought to reduce overall utilization of the health care system.

Tactics designed to promote such activity have recently begun to appear in a wide variety of governmental and corporate settings. For example, with respect to self recovery providers can opt for ambulatory (outpatient) surgery whenever medically appropriate instead of inpatient surgery. Without an inpatient stay, consumers (and their families) are forced to take more responsibility for patient recovery in the immediate postoperative period (Marks. Greenlick, et al. 1980).

For consumers, the acceptability of these tactics depends largely on their capacity to assume the increased personal responsibilities. That is, some segments of the population (e.g. the old and infirm) may be considerably less capable than others (e.g. the young and healthy) of responding to the demands implicit in the transfer of responsibilities encouraged by self care. Those segments may therefore perceive that the transfer places their health unnecessarily at risk. Accordingly, the self care scenarios in the study stressed the personal responsibility component.


Sample. The data for this study were gathered from a random sample of 365 adults. The sampling frame was the west side of Los Angeles. There was substantial variation on the education and income continuua though, on average, respondents (Rs) were relatively well educated and had substantial income. Likewise, the variation in age was considerable with Rs ranging from 18 to 84. Rs were more likely to be white than nonwhite, female than male, and to have been married at least once.

Procedures. Rs were contacted in their homes via a computer assisted telephone interviewing system on the UCLA campus for a 20 minute telephone interview during March-April, 1981. The interview protocol benefitted from three pretests involving some 130 respondents. Also, among the analytic methods used in the study were recently developed structural equation or causal modeling techniques (Bentler 1980). (A detailed discussion of the study's use of CATI for data collection and of structural equation techniques for data analysis can be found in Allen (1983).)

Variables. For each strategy a set of five scenarios was generated. The tactics emphasized in each set of scenarios were: for the preferred provider strategy, (l) using coinsurance incentives to pressure consumers to choose "member" providers, (2) basing reimbursement rates on an average of the charges of local "member" providers, and (3) basing reimbursement on the lowest of three estimates from (either "member" or non-member) providers; for health planning, (l) regulating the availability of physicians, inpatient bed supply, and technological inputs by increasing access inconveniences, and (2) curtailing ancillary services by providers; and for self care, (l) using coinsurance incentives to promote preventive behaviors,-(2) opting for ambulatory surgery whenever medically appropriate, and (3) promoting increased consumer record keeping.

For example, one scenario (for the preferred provider strategy) read as follows: "You would receive the full amount usually paid by your health insurance policy for any surgery received from the surgeon of your choice. However, coverage would be based not on the rate of this surgeon, but on the average rate of 30 local surgeons from a list approved by your insurance company." All scenarios were responded to on a five point approval-disapproval continuum.

For each strategy, nine expectation judgments were also obtained referencing three dimensions of health care: quality of care, access, and finances. Each judgment was framed so as to compare Rs' anticipation of care under the proposed (or hypothetical) strategy with their perception of the care they were receiving from their current health care arrangements. Quality of care judgments encompassed the projected doctor's thoroughness, courtesy, and efficiency. Access judgments included projected waiting times, both for getting an appointment and for office visits, and the level of inconvenience associated with getting care. Financial judgments incorporated the projected consequences of the strategy in question for amount of health insurance premiums, breadth of coverage, and insurance deductibles.

Finally, six demographic variables--ages education, income. sex. marital status. and race--were employed.


Statistics for the scenario responses (grouped by strategy) revealed that responses to the self care scenarios were more favorable than responses to the health planning scenarios which, in turn, were more favorable than responses to the preferred provider scenarios. Specifically, the self care strategy drew a largely positive response; nearly two-thirds of the sample were either strongly or somewhat approving. The health planning strategy, in contrast, drew a comparatively negative response; nearly two-thirds of the sample were either strongly or somewhat disapproving. The preferred provider strategy drew an even more negative response with nearly three-quarters of the sample classifiable as either strongly or somewhat disapproving. Statistical tests confirmed the significance of these differences.

Subsequent analyses turned from treatment of the sample in the aggregate to an examination of segments within the sample via structural equation techniques. Specifically, data were analyzed to predict scenario responses from two sets of variables: (l) the six demographic variables and (2) the three dimensions of expectation judgments made with reference to the strategies. Each strategy was treated separately in its own hierarchical model. In the model, each demographic variable was hypothesized to affect each expectation variable directly and the scenario response variable indirectly through each expectation variable. Furthermore, each of the three dimensions of expectation judgments and the demographic variables was hypothesized to affect the scenario response variable directly.

Results indicated that quality of care expectations exerted the greatest direct effect on preferred provider scenario responses. Quality of care expectations, in turn, mediated the effects of race and education. In addition, income exerted a negative direct effect on the response variable.

Thus, in the context of a sharply negative response by the aggregate sample to the preferred provider strategy, low SES non-whites were evidently less opposed to the restriction of choice imposed by the strategy. Rather, the strategy's screening function appears to have been seen by these Rs as an assurance for provider competence and courtesy that was unavailable in their current health care arrangements. High SES whites, in contrast, clearly based their negative responses on the perception that the strategy, in restricting their freedom choice, would funnel them toward providers of poorer quality. Providers approved by an insurance company would be "cheaper," less competent, less courteous, etc. than providers of their own choice.

For health planning, age showed the greatest direct (positive) effect. Also, financial expectations exerted a moderate direct effect by mediating a negative effect for age. In addition, quality of care expectations yielded a moderate effect on health planning scenarios by mediating a negative effect for education. Also, being white led to a direct positive effect on health planning.

The sample's aggregate negative response to health planning, therefore, was not prompted by expectations regarding the increased access difficulties hypothesized above to constitute new costs for consumers. The predictive findings here would, instead, seem to suggest several anomalies. For example, though whites were more likely to approve the strategy than non-whites, highly educated Rs also foresaw decrements in competence and courtesy among providers under the strategy relative to their current providers. Perhaps, highly educated whites tend to be sympathetic to the overall objectives of health planning, but distrust specific consequences for the strategy in implementation. Such anomalies will require future research to resolve.

Finally, for self care access expectations the greatest direct effect. They in turn mediated the effects of income and age. Financial expectations exerted the next greatest direct effect, mediating a negative effect for age. In addition, being white and having high education was associated directly with approval for self care.

Thus, the sample's positive aggregate response to self care derived principally from those segments in the sample which would seem most able to assume the transfer of personal responsibilities encouraged by the strategy. Those most likely to be healthy and independent of others for their well being (younger, high SES whites) linked their approval to expectations that the transfer would result in shorter waiting times, fewer inconveniences, etc. as well as lower insurance premiums, lower deductibles, etc. Those less likely to be healthy and independent (older, low SES non-whites), in contrast, foresaw worsened access and financial outcomes for themselves and responded accordingly.

Overall, it can be noted that demographic characteristics which predicted disapproval of the preferred provider strategy predicted approval for self care (being high SES and white) and to a lesser extent health planning (being white). Similarly, a key characteristic which predicted disapproval of health planning predicted approval of self care (youth). The sample's responses therefore proved highly complementary across the three strategies.


As has been the case historically, in passing recent legislation to encourage statewide PPO development, California policy-makers have not used empirical data as the basis for anticipation of consumer responses. Instead, they appear to be using untested assumptions provided by the competitive approach that underlies the strategy. These assumptions hold that consumers will respond positively in the aggregate to the marketplace pressures brought to bear by the strategy. Or, consumers across the demographic spectrum will be equally likely to respond positively. Or at the least, those consumers most affected by the changes brought about by the legislation will be the most likely to respond positively.

The data here, however, support none of these assumptions. The data of course are clearly preliminary and need further replication, yet they sound a note of warning. Specifically, they indicate that the overall response of consumers to care scenarios they might experience under California's implementation of the preferred provider strategy is negative. Moreover, young low SES non-whites are more likely to respond positively to the kind of care described by these scenarios than old high SES whites. By implication, consumer responses to the preferred provider-guided interventions will be far from homogenous across the demographic spectrum. Indeed, insofar as there is likely to be a positive response, it will probably occur among those least affected by changes in Medi-Cal--or even private insurance companies.

The data further indicate that consumer responses to the preferred provider strategy are likely to be more negative than responses to two alternative containment strategies available to California--health planning and self care. Thus, from the viewpoint of consumer responses the prospects for cost containment in California might well be improved by an upgraded emphasis on these and/or other strategies as well.

Calls for the use of multiple strategies can be found elsewhere in the containment literature. For example, McNerney (1980) has suggested that each cost-saving method--whether regulatory, competitive, innovative or voluntary can make unique contributions to a long-range, effective containment solution. The viable strategies which derive from various methods, therefore, should be developed together by testing them in combination with one another. Indeed, consumer responses can provide the blueprint for this development. Discriminations in strategy responses across segments of consumers can guide policymakers toward an optimal combination of the strategies.

The data suggesting a complementary response structure across the three strategies are consistent with this theme. If no one strategy is likely to appeal to all consumers, then it would appear advantageous for policymakers to explore which strategy(ies) various segments of the consumer population in question is (are) most likely to find appealing. Furthermore, if the segments respond most positively to several distinct strategies, then policymakers should endeavor to include each in the combination of strategies to be tested and developed.

Policymakers can exploit data on consumer responses for more than just the determination of strategy inclusion, however. Specifically, consumer response data can be used to devise appropriately formulated marketing initiatives when implementing strategies. These marketing initiatives should seek to address why consumers respond as they do to each strategy. By identifying and effectively addressing consumer concerns, policymakers can maximize the chances that consumers will accept and ultimately behave in ways which are conducive to the success of containment interventions.

The data above provide preliminary information for formulating marketing initiatives for two of the three strategies. For the preferred provider strategy, the crucial issue is likely to be quality of care. Specifically, (in contrast to low educated non-whites) high educated whites will tend to foresee poor outcomes along the quality of care dimension resulting from the restriction of choice by the strategy. Therefore, if preferred provider tactics are marketed in terms which strive to address the quality of care concerns of high educated whites (and the hopes for better quality of care outcomes of low educated non-whites), their containment prospects should improve.

For the self care strategy, the crucial issues are likely to be access and finances. That is, (in contrast to younger high income consumers) older low income consumers will tend to foresee poor access and financial outcomes resulting from the increased personal responsibility objective of self care. Thus, if self care tactics are marketed in terms which tackle the access and financial fears of older low income consumers (ant exploit the advantages along the two dimensions anticipated by younger high income consumers), the containment prospects of the tactics should similarly improve.

For the health planning strategy in contrast, the implications for marketing are less straightforward. Both quality of care and finances are likely to be important issues. Expectancies along neither dimension, however, are likely to mediate the positive (negative) responses that older (younger) respondents will show toward health planning. Other psychological variables could plausibly mediate this relationship, for the hypothesized mediators tested in this study certainly were not exhaustive (e.g. perceptions of the efficiency levels of health care bureaucracies).

Other limitations respecting the study should be noted as well. For example, sampling is a concern, both for respondents and for dependent variables. Clearly, this study's inferences to statewide developments would be more persuasive if based on data which systematically sampled both urban d rural consumer population; throughout the state. In addition, longitudinal data would considerably strengthen the cross-sectional prospective data gathered here, particularly if the attitudinal responses were supplemented by behavioral response data from those actually receiving care under the aegis of each strategy.

Second, expectation judgments may not be sufficient in and of themselves as a basis for determining the "why" underlying consumer responses. Rational choice theories in the social sciences, for instance, would stipulate that the expectation judgments which consumers make along the three dimensions of care need to be considered in conjunction with the values placed on the dimensions (Allen 1983). The effectiveness of marketing initiatives stand to be considerably enhanced if devised on the basis of the two components together, rather than expectation judgments alone.

Third, to simplify the introduction of consumer-oriented analyses 9 this study has assumed that the three strategies are more or less interchangeable as vehicles for cost containment. Obviously, this assumption bears little weight in the real world. For instance, whereas the full cost containment potential of the self care strategy has clearly yet to be tapped, it likely will be only a fraction of the potential achievable under purely competition-based or regulation-based strategies. Such considerations introduce layers of policy-oriented complexity featuring relative cost d price issues which effective marketing initiatives will have to incorporate. The present analysis, however, only acknowledges them.

These limitations notwithstanding, this study introduces a methodology for incorporating consumer responses into containment policy that would appear particularly useful either prior to or in the preliminary stages of strategy implementation. In so doing, this study suggests serious reconsideration of the well-established precedent for pursuing cost containment in health care in the absence of data on consumer responses. Containment policy utilizing empirical data on consumer responses has substantially greater chances of successfully anticipating these responses and, ultimately, of achieving its cost saving objectives.

Indeed, this study outlines a framework for an iterative process that could serve as s integral component to containment policy. Briefly, this iterative process would seek to optimize the match between alternatives for cost containment and consumer acceptance. For example, policymakers might begin by surveying various strategy-inspired delivery systems and marketing initiatives that are feasible alternatives for a given population of consumers. Feasibility in this case would be determined by such variables as current modes of health care practice, the quantity and quality of available resources, the experience of health care personnel, etc Policymakers would then seek to establish a feedback loop based on multiple contacts with samples from this population. This feedback loop would monitor consumer responses for the purpose of formulating and revising the systems and initiatives. Its objective would be to develop promising long term matches between the systems and initiatives on one hand d segments of consumers on the other. It would continue through the implementation period to monitor consumer responses for the benefit of future implementation efforts. Such an iterative process could be executed relatively cheaply, yet it could greatly facilitate a long range solution to the cost containment crisis that now confronts health care.


Allen, H. M., Jr. (1983). Alternative Strategies for Cost Containment in Medical Care: Lay Public Responses under Conditions of Uncertainty (Doctoral dissertation, University of California, Los Angeles, 1982). Dissertation Abstracts International, 43(10).

Bentler, P. M. (1980). Multivariate Analysis with Latent Variables: Causal Modeling. Annual Review of Psychology, 31, 419-456.

Enthoven, A. C. (1978). Consumer-choice Health Plan: A National Health Insurance Proposal based on Regulated Competition in the Private Sector. New England Journal of Medicine, 298(13), 709-720.

Fry. J. (1978). A New Approach to Medicine. Baltimore: University Park.

Gartner, A., and F. Reissman (1974). The Service Society and the Consumer Vanguard. New York: Harper and Row.

Levin, L. S., Katz, A. H., and E. Holst (1976). Self Care: Lay Initiatives in Health. New York: Prodist.

Marks, S. D., Greenlick, M. R., Hurtado, A. H., Johnson, J., and Henderson, J. (1980). Ambulatory Surgery in an HMO: A Study of Cost, Quality of Care, and Satisfaction. Medical Care, 18(2), 127-147.

McNerney, W. J. (1980). Control of Health-care Costs in the 1980's. New England Journal of Medicine, 303(19), 1088-1095.

Mechanic, D. (1978). Approaches to Controlling the Cost of Medical Care: Short Range and Long Range Alternatives. New England Journal of Medicine, 298(5), 269-256.

Newman, J. F., Elliott, W. B., Gibbs, J. O., and Gift, H. C. (1978). Attempts to Control Health Care Costs: The U.S. Experience. Social Science Medicine, 13A, 529-540.

Preferred Provider Organizations: A Developing Concept in Health Delivery (1982). Socioeconomic Report. Bureau of Research and Planning, California Medical Association, November/December, pp. 1-7.



Harris M. Allen Jr., The Rand Corporation


NA - Advances in Consumer Research Volume 12 | 1985

Share Proceeding

Featured papers

See More


What Converts Webpage Visits into Crowdfunding Contributions: Assessing the Role of Circumstantial Information

Lucia Salmonson Guimarães Barros, Universidade Federal de Sao Paulo
César Zucco Jr, Brazilian School of Public and Business Administration, Brazil
Eduardo B. Andrade, FGV / EBAPE
Marcelo Salhab Brogliato, Brazilian School of Public and Business Administration, Brazil

Read More


Time-of-Day Effects on Consumers’ Social Media Engagement

Ozum Zor, Rutgers University, USA
Kihyun Hannah Kim, Rutgers University, USA
Ashwani Monga, Rutgers University, USA

Read More


Hindsight Value: Failed Transactions Inform Willingness to Pay

Masha Ksendzova, Boston University, USA
Carey K. Morewedge, Boston University, USA
Dan Ariely, Duke University, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.