Diffusion of Innovation & Change in Health Care Policy: Why We Just Can’t Seem to Learn!
BY DAVID WILSFORD
When you think about public policy issues that are not working well, it does not take a rocket scientist to identify the health care system in the United States as one of them. For those who study public policy and particularly those who look at how other countries do it, a central problem found everywhere is the lack of learning about how others deliver the best health care they can to the most people in their country (and do so without breaking the national accounts).
Nowhere is this truer than in the United States, where it is not difficult to argue that, among all the advanced democracies of the world (the Organization of Economic Cooperation and Development countries, known as the OECD) we find the worst system, which we could more charitably refer to as the “most suboptimal.” The lamentable state of health policy in the US is the most flagrant example of the failure to learn how to do it better. The US spends more than any country in the world on health care in the aggregate (over 15 percent of GDP) for the least favorable health status results, all the while leaving 45 million US citizens without health care coverage. No wonder that filmmaker Michael Moore turned his considerable polemical skills to this mess in the 2007 film “Sicko.”
Contrast this state of affairs in policy with the very effective and rapid diffusion of medical advances in science and technology across all national boundaries. The latest HIV/AIDS research flashes around the globe between research scientists and their labs at a breathtaking pace. The top researchers in France always know where their American colleagues (and others) are in the search for bio-medical progress in this or that medical area.
The spread of new drug therapies, to take another example, occurs with amazing swiftness as long as you can pay for it. The introduction of a major new drug on the market happens at virtually the same moment across all the advanced countries. What’s the latest cancer treatment? Everyone knows . . . and moves to adopt it. What’s the latest drug for this or that? Everyone knows . . . and moves to prescribe it. What’s the latest problem uncovered with stent devices? Everyone knows . . . and watches closely for solutions to emerge from research and development, no matter what country that is located in.
If globalization is epitomized by anything at all, it is by the free and fast flow of information across the planet in the bio-medical sciences. Why are the natural sciences so effective in spreading innovation and progress across nations but scientists and politicians are not?
The primary reasons for this disjuncture in cross-border learning between the natural and social sciences, and particularly in the policy realm, lie in two main areas.
First, the natural and physical sciences have a strong common understanding of what their dependent variables are. This understanding is shared internationally across all the national communities of researchers and laboratories. By contrast, in the social sciences and across policy communities, there has never been a single commonly accepted articulation of how the most important dependent variables are defined.1
Additionally, the organization of the vast and overlapping networks and hierarchies of actors in the bio-medical universe is not structurally characterized by the same degree of stickiness and isolation as their policy science counterparts. In other words, the structure of networks and hierarchies of relevant actors in the natural sciences facilitates the flow of ideas. In contrast, the structure of comparable networks and hierarchies in the policy sciences and in the policy communities does not facilitate flow, rather it facilitates isolation and barriers.
This point is developed further through examination of the phenomenon of “diffusion effects.” In the world of ideas, the great sociologist Reinhard Bendix, in his masterwork Kings or People (1978) argued that the great political-social movements across the world have been triggered by the flow of ideas from other societies and countries. The American Revolution was heavily influenced by European ideas of the Enlightenment; the Bolshevik revolution in Russia was likewise influenced by the development of participatory democracies and Marxist thought from Western Europe. And so on.
Clearly, the world of modern bio-medical science that has been briefly described is characterized by extensive diffusion effects that flow and spread with lightning speed. Yet the flow of ideas and innovation in health care policy moves as slowly as Grandma’s robust molasses—if it even moves at all. The wealth of evidence from the diverse ways that other countries “do” their health care system (the United Kingdom, France, Canada, Japan, Germany, among others) is either minimized or altogether ignored by most actors in the policy space— public and private—of the US health care system.
The high fluidity of communications in, through and around the bio-medical world can be attributed to the fact that the structure of the networks and hierarchies of this world are less characterized by rigid, non-porous boundaries. That is, because the entire worldwide bio-medical community shares a common understanding of the important dependent variables (which one might simply label the “problems at hand”), the natural divisions of and boundaries between the hierarchies and networks that characterize the structure of this environment are very porous. Because all the important dependent variables are the same, the flow of ideas and communications about who is doing what and why is fluid and extensive.
Unfortunately the structure of networks and hierarchies in the policy world is very segmented and characterized by more impervious boundaries. In the policy realm, for an area such as health care policy, the chief units of analysis are the different countries, and the boundaries between them to the flow and acceptance of policy ideas and practices are rigid and impermeable. Even within countries, the actors in the health system, because there is little common understanding or agreement on the most important dependent variables, find themselves working in hosts of segmented isolated pockets, the boundaries within the system proving far less porous and much more impermeable than those of the physical sciences.
In an overview of networks and hierarchies across the sciences versus the policy world, identification of some common structural elements reveal how their characteristics enhance the flow of ideas and practices in the sciences while impeding the flow of ideas and practices in the policy domains.
Both ‘networks’ and ‘hierarchies’ are referred to because it is a simple way to depict how complex adaptive systems, either scientific communities or policy communities, are organized. Horizontally, networks of actors spread throughout a given scientific space—across universities, across laboratories, across institutions, across governments—all interacting back and forth as time unfolds in the environment. Vertically, hierarchies organize the accountability relationship within a given structure according to lower levels reporting up to higher ones and higher levels deciding down for the lower ones. (Both network and hierarchical structures are very clean in theory, but often very messy in the real world; nonetheless, the two dimensions usefully characterize most of the actor relationships of a given universe or community.)
Populating all these networks and hierarchies, which naturally overlap and cascade throughout any system, the individual actors (sometimes called “decision agents” or with even more precision, “quasi-autonomous decision agents”), holding their identifiable responsibilities that are more or less clearly defined—vary in their own respective attributes. We can think of this variance as the difference in factor endowments, which is a technical way to refer to the fact that every single actor (or decision agent) on a given scientific or policy landscape poses greater or less authority, greater or lesser resources, greater or lesser aptitude herself or himself to deploy whatever putative authority or resources she or he has.
Moreover, every environment is characterized by a high degree of contingency. That is, what one person thinks or does in a given context is being influenced and impinged upon by what others are thinking and doing. The policy communities however are characterized by much higher contingency than even the considerable contingency of the scientific communities. This even greater degree of contingency introduces a far greater degree of uncertainty in the environment for the policy actors, making it more difficult to coordinate and further impeding clear flows of communication.
For any policy arena, compared to any scientific community, these differences mean that the structures are stickier, less elastic and that the entanglements of uncoordinated, indecisive agents are more immense. The interaction of multiple networks and hierarchies that cross-cut and cascade through time in this policy world is more confused and less porous.
The stickiness and entanglements of the policy environment thereby reinforce, even magnify the effects of little agreement on what should be the commonly accepted dependent variables. It is logical that the flow of ideas and learning across policy communities and, especially, across countries would be considerably more impeded than in the scientific world.
These are some of the structural reasons that the United States just never seems to learn from others in regards to gains in national health care policy. While the effects of globalization have changed many areas of trade and commerce that affect the United States, and while its own bio-medical community can clearly be seen to be an integral part of a global one, its policy communities in health care remain segmented, isolated and impermeably shut off from much outside influence that might enter from outside the country’s borders.
And so the suboptimality of the currently constituted health care system is likely to persist—or at the very least, remain very hard to change. In health care policy, it results in an unhappy picture for those who believe that the current system is not up to speed and needs to be improved. Change (or reform) is possible, but it remains unlikely, and insofar as it does ever occur, it will be a long, hard slog.
David Wilsford (dwilsford@gmu.edu) is a professor of government in the department of Public & International Affairs (http://pia.gmu.edu/) and at the London School of Economics. This article was first published in print and citations have been removed due to space limitations, but are available from the author.
- “Dependent variable” is a scientific term to signify the specification of the problem at hand. “Independent variables” can be thought of as the causes of the effects, the “effects” being the problem at hand being researched. Technically, all science is the search for independent variables that explain the variance on the dependent variable. This characterization is as true for the social sciences as it is for the natural or physical sciences. [↩]
