Contandriopoulos D, Bilodeau H. Health Policy 90 (2009) 104-112.

Background: Where does the public turn for information about private healthcare delivery and funding? Surveys and public opinion polls are a major source of information. These polls and surveys are often derived from commercial firms, private interest groups or statistical collection agencies. The apparent simplicity of polling and survey results stand in sharp contrast to their methodologic and interpretive complexities. Using privatization of healthcare funding as a case study, the authors analyze the nature and role of opinion polls as a source that is instrumental in public policy-making. They take a three-part approach: elucidate the relationship between polls, public opinion and policy-making; portray the opinions of Canadians and Quebeckers about the healthcare system, using results from commercial polls and statistical surveys; and, use this portrait to highlight logical, theoretical and methodological flaws in using polls to inform political decisions or determine policy.

Objective: To examine the nature and role of public opinion polls in shaping the debate on privatization of healthcare funding.

Methodology and Limitations: This research collected all relevant and available poll data between 2001 and 2006 from Statistics Canada and four major polling firms (Ipsos Reid, Leger & Leger, Pollara, and Harris-Decima). The researchers examined 52 commercial polls to examine the meaning and use of polls, and they cited data from 14 polls. This is a policy analysis coming from well-versed academics of the University of Montreal and University of Quebec. As with any policy or political analysis, the perspective of the analysts influences the findings and conclusions. Moreover, the analysis is based on commercial polls, academic research and statistical surveys, the selective use and choice of which might bias or influence the findings. However, the analysis is largely a narrative discussion of methodological problems, discrepancies in results and underlying logical fallacies, such that problems with the original data would likely have little influence on the analysis.

Results: Polls are used in at least two different ways: to inform policy and research or to frame news reporting along political lines. The former uses polling data to understand how public opinion shapes policy, while the latter uses polling data as a tool for gaining political advantage, by influencing the interpretation of the polls to legitimize a point of view – the spin! Clearly, these two functions of polls are inseparably linked, with either or both being the underlying intents of any poll or survey. Because polling for political purposes is mostly carried out by private firms, the results may be withheld from public scrutiny unless they are consistent with the political purpose.

The authors admit that extracting a clear picture of perception on healthcare is a “daunting task”, given the difficulty of explaining the widely variable results. They conclude that people who have used the healthcare system have a more favorable view of the quality of the system than those who have not. The same holds true for perceptions on access. One speculation as to why this may be the case relates to the impact of media alerts on waiting lists and negative lingering perceptions among the public that there is a general problem with access, rather than any actual personal experience. These media-induced perceptions create worry and negatively impact on the public’s confidence in the system. For example, surveys conducted since 2000 consistently show that 82-84% of Canadians and Quebeckers who have used the health care system report being satisfied or very satisfied with the services they received. Yet in another poll, 46% of Canadians and Quebecers believed that the system’s quality and access will deteriorate in the next five years. Paradoxically, the population appears pessimistic about the future, even though individual personal experience is, in fact, overwhelmingly positive. This paradox, fueled by media attention, of course, underpins arguments for a private parallel system, guided by the myth and unsubstantiated assumptions that a two-tier system somehow improves access and quality in the public system. We have shown elsewhere in our CDM e-rounds that this is not the case (e-rounds 1-4, 7, 8, 14&16). Nevertheless, 49% of Canadians and 51% of Quebecers say that they are personally willing to pay directly to gain faster access to services. In contrast, when the public is presented with several options for improving access to care, private-pay options rate poorly. The authors conclude: “It … appears that the population at large prefers a publicly-funded system and … the attraction to a privately-funded system comes from the myth of improved access and better quality on the private side.”

Conclusions: Results from polls on healthcare reform show wide fluctuations and poor internal consistency. Some of this wide variability is caused by inadequate respondent information or insufficient response time. Less easy to explain are the opinions on specific subjects, for instance, the Chaoulli decision. Despite the fact that respondents have not ever seen or heard the actual judgment, they were left with strong opinions, raising legitimate questions about the validity of their judgments. Validity may also be sacrificed by efforts to maximize response rates, such as those that come from forcing responses to closed polling questions. As the authors point out, it is not uncommon for poll results to vary widely and for respondents to provide different answers if given more time and opportunity to reflect on the issues.

Relevance: Polls may be informative or political. They may or may not faithfully represent population values and perceptions. This appears particularly true for polls on privately-funded healthcare which are undermined by weak underpinnings in logic and methodologic validity. Moreover, polls used for political purposes are meant to create and influence public opinion, as opposed to attempting to understand it.

Note from E-rounds Authors: In Spring of 2009, the Canadian Medical Association conducted a survey of members to determine their views on proposed “transformations” to the healthcare system. We and others have raised serious concerns about the validity of the survey results because of the biases embedded in a number of questions, revealing a political agenda supporting investor-owned, profit-driven delivery of health care services. As noted in this review, “the poll is a powerful tool in purely political battles aimed at gaining acceptance for a specific opinion”. We would caution that conclusions based on such agenda-driven polls are suspect. They may be purely an artifact of ideology and political process serving either not to inform us at all, or to simply misinform.