Scholars have long recognized the importance of comparisons for envisioning, formulating, and implementing new public policies. But cross-national comparisons of health systems and policies can prove challenging—sometimes potentially misleading—unless one carefully attends to the nature of comparative reasoning and the inferences that it allows. This article explores the distinctive logics of two forms of comparison: analogies and metaphors. Each offers potentially important insights regarding the performance of the US health care system and appropriate aspirations for American health policy. But they do so in very different ways, each form of comparison singling out particular other countries, policies, and health outcomes as the most appropriate or meaningful comparators. These patterns are illustrated with examples drawn from the articles and commentaries in this special issue. These explorations of comparative reasoning also highlight some gaps in the topics addressed in these other contributions—gaps that merit attention in future research.
Comparison comes naturally to us all. We rely on comparisons to assess worth, to make choices, and sometimes to discern how things work. Comparisons can also help us envision how things might work differently in the future. Comparing health system performance is thus a natural and intrinsic part of health policy discourse—always has been, and always will be.
Yet such comparisons can easily go awry. Metrics of performance that make comparison possible may prove deceptive if the meaning of those metrics varies over time or across national borders. Comparisons based on past performance may highlight causes that mattered greatly in years gone by but are less relevant to the future. Comparisons of complex systems can easily fixate on aspects of performance that are most readily measured or simplest to comprehend, obscuring or neglecting other equally vital outcomes.
Perhaps most challenging when it comes to assessing national health systems is the fact there are so many metrics for performance, so many ways in which they might be compared, so many criteria for making those comparisons more rigorous, so many comparators from which we might learn—so many that analysts can easily become ensnared in the demands and opportunities of careful measurement and analysis, losing track of the broader purposes of the comparative inquiry.
This special issue is filled with comparisons—comparisons across countries, across time, across industries, across regions within countries, across localities within those regions, across health services, across different venues for delivering services, across forms of collective assistance (health vs. social care), across different determinants of health outcomes, across different definitions for those outcomes—a veritable cacophony of comparisons. Each comparison comes with its own important, insightful cautionary tales related to the need for more precise measurement, more careful analysis, and more consistent interpretation.
And therein lies the rub. The rigorous, methodologically careful comparisons that populate this special issue were intended to be an antidote to simplistic judgments about American health care—whether the facile proclamations of politicians asserting its preeminence in the world or the harsh judgments of critics who ridicule that same system as bloated, ineffective, and disconnected from the true threats to health and well-being. Each article strives to be antidotal in its own distinct way; each is effective on its own terms, as are the thoughtful commentaries that accompany the articles.
All this is very useful. Yet precisely because these essays are so attentive to the nuances of careful measurement and analysis, they may strike some readers as being a bit reductive, losing touch with some of the larger purposes to which sound health policy and governance ought to attend. In short, they may seem to be skirting some key questions: Do Americans currently benefit from a “high-performance” health care system (Davis 2005)? Might they be able to do so in the future (Schneider and Squires 2017)? In fact, I believe that the articles do help answer these questions, but to see how, we must be as rigorous in thinking through the logic of comparative analysis and the ways in which comparisons can be used to assess health systems as we are in measuring and analyzing particular outcomes.
This final commentary introduces such rigor. It is my hope that these comments will provide an additional lens for interpreting the findings presented in this special issue and that readers will find it useful to return to the articles using these additional perspectives as an interpretive guide. I examine below some illustrative examples, drawn from these articles and commentaries. But these examples only touch the surface; each of the contributions to this special issue embodies a rich variety of comparative reasoning, all of which deserve to be more carefully explored and more fully explicated than is feasible here.
This more rigorous conceptualization of comparative reasoning also helps identify some gaps in these earlier contributions. These include aspects of health system performance that were overlooked or treated in too cursory a manner. I also highlight questions about system effectiveness that went unasked because the articles focused on particular aspects of performance rather than their interaction and overall impact on the public. These gaps also merit greater attention; here I explore some reasons why the gaps emerged and how they might be addressed.
The Logics of Comparative Reasoning
Scholars have long recognized the importance of comparisons for envisioning, formulating, and implementing new public policies (Bardach 2004; Mossberger and Wollman 2003). Comparisons take two basic forms: analogies and metaphors. There is a rich literature on the role that each plays in the thinking of political elites and the general public (Bougher 2012; Schauer 2008). The impact of comparative reasoning is evident across a variety of policy domains (including health care) and appears most pronounced when problems are highly complex, rendering expected costs and benefits of collective action most uncertain (Bougher 2012; Chan, Paletz, and Schunn 2012; Barry et al. 2009; Houghton 1998; Gavetti, Levinthal, and Rivkin 2005).
Although sometimes elided in the literature, analogy and metaphor shape our understanding in quite distinct ways (Schlesinger and Lau 2000). Learning from analogies depends on the similarity of the two comparators. When the objects of comparison are sufficiently similar, analogies can be helpful for both diagnosing problems and identifying potential remedies. The article on infant mortality illustrates diagnosis. Because the medical technologies for identifying fetal distress and keeping newborns alive are similar in the United States and other high-income countries, it was plausible to infer that the relatively high residual rates of infant mortality that persist in American households (particularly in lower-income settings) most likely reflect social determinants of ill health.
The articles on health care prices and waste/efficiency exemplify the prescriptive potential of analogies. This can be perhaps best illustrated by contrasting two forms of health care. On the one hand, because the pharmaceuticals being purchased in high-income countries are quite similar (most often identical), it is plausible to infer that price control strategies deployed in other countries ought to have comparable impact reducing medical costs in the United States. On the other hand, the nature and structure of primary care differ considerably across national boundaries. Consequently, it is more difficult to claim that strategies to reduce prices or curtail inefficiencies based on changes in primary care in other countries will necessarily translate well to the United States—or even those that have worked well in some parts of the United States will translate effectively to other regions. In each case, the reliability of the prescription—the extent to which policy transfers can be expected to yield comparable results in their new context—depends crucially on the similarity of the two contexts (Cook 2008; Mossberger and Wollman 2003).
Metaphorical reasoning works quite differently, because metaphors generate insights by comparing dissimilars. Consider a familiar literary example: when Romeo compares Juliet to the sun, it does not detract from the illuminating power of the metaphor that Juliet was neither rotund nor likely to have her visage obscured by passing clouds. The metaphor reminds us of particular attributes—Juliet's warmth and radiance—that embody the very nature of romantic love. So too, metaphors deployed for the purposes of health policy analysis call our attention to certain facets of complex delivery systems, patient experiences, or policy outcomes that might otherwise be overlooked or undervalued.
That gives metaphors a very different sort of utility than what analogies bring to comparative policy analyses—more limited in some ways, less limited in others. Unlike analogies, metaphors are not particularly useful for prediction or prescription, any more than likening Juliet to a celestial object will encourage her to have a sunny disposition. Consequently, when the articles in this special issue compare the United States to countries that have single-payer (or, more relevant for this example, single purchaser) health financing, the comparisons are more metaphors than analogies.
As such, they are not as much claims about how comparable single-payer systems might work in the United States as they are reminders about the benefits of leveraging purchasing power to promote lower prices, greater efficiencies, or improved quality. Those same outcomes can be at least partially achieved through comparable pathways that do not require a single payer. For example, price reductions have been achieved through either public policies (e.g., state-level all-payers systems) or private initiatives (e.g., consolidated purchasing among managed care plans with market power). The metaphorical power of the single-payer comparisons helps to remind US policy makers why they might wish to revitalize these past efforts, though both have fallen into disfavor among health policy elites.
Metaphorical comparisons can have greater breadth and wider impact than analogies in several different ways. First off, one can gain metaphorical insights from a wider range of comparators, since overall congruence is not a prerequisite for generating those insights. Quite the contrary, metaphorical comparisons may prove most insightful when they compare systems that are radically different, making emergent forms of commonality more surprising and their implications richer in innovative potential. Following this logic, one can draw useful metaphorical insights for health policy in the United States from policy experiences in low-income countries (Knaul et al. 2015), countries that have no health insurance at all (Parston et al. 2015), or countries with very different governance arrangements for health policy making (Campbell and Morgan 2005). These sorts of metaphorical lessons are most evident within this special issue in the articles on the clinician workforce and on public health investments.
Unlike analogies, metaphors also have the potential to reshape how we think about particular issues, as well as what ought to be done about them. This more conceptual aspect of metaphorical reasoning was also evident in the article on the clinical workforce. The key terms primary care and generalist deployed in that article are, in addition to being antonyms for specialization, each a sort of metaphor. Put differently, although both terms are used to categorize clinicians (based on training, practice setting, or practice style), they are also more than categorical labels. They represent configurations of expectations regarding clinical practice—what we hope a certain proportion of clinicians in every health care system will do in order to foster coordinated, patient-centered care. Invoking both in the context of this article highlights the need to think more carefully about which set of expectations ought to be given greatest priority in shaping workforce policy.
But it is the article (and associated commentary) on academic medical centers (AMCs) that perhaps most powerfully embodies the importance of metaphorical thinking in shaping health policy. Both article and commentary make a compelling case that critics of AMCs often blame these institutions for the broader failings of American health care and health policy making, characterizing this as a “fallacy of misplaced concreteness.” (Given the construction techniques used at many AMCs, this critique may be literally as well as logically accurate). But viewed through the lens of metaphors, the propensity to view certain problems as manifestations of AMC practices is not as much a logical fallacy as a psychological necessity.
In this context, AMCs are functioning as what linguists and cognitive psychologists refer to as “condensing metaphors” (Cameron 2007). In these metaphors, the comparator (in this case, the AMCs) stands in for something more complicated and harder to interpret (in this case, the health care system). Take income-related disparities in health care access and outcomes as a case in point. These disparities are more pronounced in the United States than other countries with comparable economic development (Schneider and Squires 2017; Hero, Zaslavsky, and Blendon 2017). But for all their import and extensive documentation, these disparities can seem a little murky for policy elites and the general public alike, because their causes and consequences so permeate health care and society at large. AMCs render disparities easier to envision: by attracting wealthy patients to seek care and well-off clinicians to care for them in the lower-income neighborhoods in which many AMCs are located, they make evident ethical tensions when services feel less accessible to those who live nearby (Gray et al. 2009).
In short, taking AMCs to task for broader system and societal failings may seem illogical yet still be a psychological necessity. Indeed, one might anticipate that the use of condensing metaphors will be more common in American health policy discourse than elsewhere. Precisely because the financing and delivery systems in the United States are more fragmented and complex than those in other high-income countries, they are also more confusing and seemingly opaque, requiring heuristics to clarify the diagnostic process.
Once these comparisons are understood to be metaphors, they are less likely to generate unreasonable expectations that all health system shortcomings and disparities should be rectified by interventions set within AMCs. That said, as metaphors these critiques can usefully remind those in charge of AMCs that they ought to bear some responsibility for addressing the shortfalls, whether or not their practices have been a primary cause. How much and in what ways they should address them can then become the focus of future policy deliberations.
Distinct Purposes of Comparative Analysis and Contributions from This Issue
The foregoing exploration of comparative reasoning identifies several ways in which comparisons (cross-national or otherwise) can enrich health policy discourse. These can be grouped into four categories:
diagnosis: identifying when a health system could perform better;
assessment: identifying the causal pathways that account for the diagnosed shortfalls;
prescription: identifying new interventions or revised policy configurations likely to enhance future performance; and
prioritization: identifying which aspects of health system performance most merit policy makers' attention and which should be most emphasized when striking trade-offs among competing outcomes.
The articles and commentaries in this special issue focus primarily on diagnosis and assessment. In so doing, they contribute an array of useful insights into the comparative performance of the American health care system. Of course, no six articles, however capably compiled, could ever comprehensively represent that performance. And three missing pieces seem particularly crucial for assessing the distinctive ways in which the US health care system is assembled, governed, and experienced by patients treated under its auspices.
The first involves equity: to be sure, a number of the contributions here—including the articles on infant mortality, clinical workforce, and AMCs—make reference to income-related differences in outcomes or care experiences. However, the broader cross-national differences in how health equity is defined and experienced are striking and far more pervasive than the manifestations identified in these articles (Hero, Zaslavsky, and Blendon 2017). These merit more careful and complete consideration. What is particularly striking, for example, is the extent to which many commonwealth countries that have political cultures and public values very similar to those in the United States have nonetheless created health systems yielding more egalitarian patient experiences with access and quality than those reported by Americans, or conversely, the extent to which some countries with quite egalitarian health care access (e.g., the United Kingdom) nonetheless experience income-related disparities in health outcomes as large as those in the United States.
A second essential missing piece involves financial security. Even after one takes into account cross-national differences in insurance enrollment, Americans report far higher levels of financial exposure and uncertainty related to medical expenses (Schoen et al. 2013). This dramatic contrast in perceived risk is arguably most important as a metaphor. Most Americans have lived for so long with incomplete and unreliable insurance that, in the absence of effective cross-national comparisons, it may have become virtually impossible for either the public or policy makers to envision insurance for medical care that effectively buffers financial insecurity.
The third essential element for comparing health system performance involves choice. Although American health policy has historically placed a distinctive emphasis on consumer choice, over the past fifteen years there has been a marked increase in choice-promoting health policies across most market democracies (Schlesinger 2010). This growing enthusiasm for consumer choice has become something of a shibboleth and an element of faith around which an elaborate belief system has developed among policy elites (Paulus et al. 2003). It is therefore remarkable how little is known about the extent to which the policies pursued in any of these countries actually empower consumers to make informed choices among either providers or health plans (in systems that incorporate multiple insurers or equivalent pooled financial arrangements).
Augmenting the articles in this special issue with comparative analyses of equity, financial security, and choice would yield a reasonably comprehensive diagnosis and assessment of comparative health system performance, that is, would provide a reasonably complete portrait of how the health system that Americans experience compares to those in other market democracies. However, for these comparisons to meaningfully guide prescription and prioritization, one other essential perspective is missing: the ways in which governance structures, political culture, and policy legacies constrain policy makers' efforts to reform policies that shape health and health care.
To be sure, some of the articles and commentaries (most notably those related to public health and AMCs) attend to some of these factors. What is missing, however, is the sort of systematic structural analysis that has long accompanied social scientists' study of comparative welfare states (Esping-Andersen 1990). How much do the basic elements of health system design (e.g., health insurance vs. national health service; single insurer vs. multiple insurers; primary care gatekeepers vs. free consumer choice of any clinician) effectively determine health care use and outcomes, whatever other health policies are pursued in that polity? How much do these structural elements or deeper invariant features of political culture constrain both which policies will be seen as politically feasible and how effective policies would be if adopted?
Structural analyses can be empirically challenging. Because these attributes tend to be stable over time, there is often too little variation to reliably assess the statistical relationship between attributes and outcomes. But there are just enough tantalizingly anomalous patterns that can be discerned from even simple cross-national comparisons to suggest that these structural analyses may yield some intriguing findings. To illustrate, national health services do not always lead to more equitable distributions of health outcomes—the United Kingdom is the striking exception in this regard (Hero, Zaslavsky, and Blendon 2017). And single-payer systems do not always yield reduced confusion and complexity for people dealing with insurance administrators—France proves to be the exception in this case (Schoen et al. 2013). In both cases, the crucial (and as yet unanswered) question is this: Why are these particular countries exceptions to the more general patterns? What might the answers tell us about how structural features of the health care system influence valued health outcomes?
Ultimately, our capacity to learn from cross-national comparisons will depend on both the rigor with which those analyses are conducted and the care with which the findings are interpreted. Scholars may have more leverage over the analyses than the interpretations, but they need to be attentive to both. Only by insisting on rigor for each of these stages will comparative studies yield the sort of meaningful lessons that can promote more effective and equitable health and social policies.