One often hears the phrase “laboratories of democracy” in connection with federalism — the Brandeisian idea that state and local governments can “test” new ideas without doing any harm to the rest of the country. It’s an attractive idea, but there are some complications:
(1) Nationalists, i.e., advocates of central authority, are inclined to think about races-to-the-bottom rather than races-to-the-top, and to focus on the possibility that state and local experiments might indeed harm the rest of the country, e.g., the commonly-held beliefs that lax regulation of firearms might impact neighboring jurisdictions or that inadequate social services might lead to outmigration (and a kind of adverse selection dynamic might take hold), etc. In the U.S., this concern is most commonly associated with the political left.
But in a somewhat different vein, at least some conservatives, myself included, are skeptical about cooperative or cartel or polyphonic federalism, in which federal and state governments concurrently regulate various domains. We prefer the idea of dual or competitive federalism, in which states and the federal government stick to their respective lanes. State and local experiments that intrude on federal authority, e.g., nontariff barriers or state regulations in domains that ought to be regulated exclusively by the federal government, ought to be discouraged under this framework.
(2) The mental model of some advocates of a “laboratories of democracy” framework is that states ought to try new things — e.g., created subsidized exchanges designed to restructure the individual and small-group market for medical insurance — and the federal government ought to mimic policies deemed successful, or rather politically attractive. Assuming the “success” we have in mind is that a given program alleviated the problem it was designed to solve at an acceptable cost — as opposed to success in helping elected official X win reelection or secure donations from a particular constituency — we’re faced with a new problem: it is possible that success in Mississippi doesn’t scale well to success nationally, for the obvious reason that every state is unique. Jim Manzi addressed exactly this issue in Uncontrolled. And he suggests that rather than think about test-and-learn experimentation in public policy as a matter of identifying “what works” and then scaling it up, it ought to be embedding a capacity to test-and-learn in all service delivery organizations to better determine what works where, as a literacy program that works in Tulsa might not work in Austin, etc.
Instead of scaling up state-level innovations by having them go national via Congress, we might instead work to facilitate the diffusion of successful state policy initiatives, i.e., we might encourage Texas to learn from Arkansas or New Jersey to learn from Massachusetts. (The convergence of local public school districts around the standard U.S. summer vacation is an intriguing example of how diffusion has worked in practice.) This is one of the goals of organizations like the American Legislative Exchange Council (ALEC), which has been widely condemned in left-of-center circles.
Gabriel Rossman kindly pointed me to the work of political scientist Graeme Boushey on state-to-state policy diffusion:
Although federalism makes policy coordination difficult, it also creates opportunities for considerable policy innovation, as municipal, county, and state governments develop new policies to address local concerns. Federalism encourages venue shopping, a process where activists and interest groups strategically exploit the multiple venues of government to secure support for their legislative programs (Baumgartner & Jones, 2009; Holyoke, 2003; Pralle, 2003). This process increases the number of new ideas entering the political systems and can create conditions where “new ideas or policy images may spread rapidly across linked venues, thus setting in motion a positive feedback process” (Baumgartner & Jones, 2009, p. 240). This is especially likely when a focusing event directs national political attention on a local policy reform that appeals broadly to voters. Federalism therefore creates both opportunities and constraints for rapid policy diffusion.
Baumgartner and Jones (2009) observed two different ways that issue expansion can lead to sudden patterns of policy diffusion. First, issue attention can lead to the nationalization of policy attention as federal institutions focus policymaking attention on a policy program that had previously been viewed as a concern of local and state governments. When this occurs, the federal government emerges as a strong catalyst for state-level policy adoption, employing grants, mandates, or other signals to simultaneously change policy across the states. In the 1960s, growing concern over highway conditions led to direct federal intervention in state transportation laws, most prominently linking federal highway funding to state adoption of mandatory motorcycle helmet legislation. Furthermore, Baumgartner and Jones (2009) suggested that the nationalization of issue attention on transportation safety legitimized sustained federal involvement in this issue area, resulting in a sustained period of rapid policy change.
Positive feedback cycles can also be triggered absent the direct involvement of the federal government. Boushey (2010) noted that issue attention can also be nationalized through sustained interest group campaigns as prominent advocacy groups organize to implement a policy from one state to the next. For example, state-level adoption of child abuse reporting statutes occurred not through top-down coercion but through the rapid communication of a model state law across a professional interest group network of advocacy organization (Baumgartner & Jones, 2009).
One issue Boushey highlights is that state-to-state diffusion doesn’t generally reflect a careful, deliberative process of incremental learning, along the lines Jim has in mind. Rather, they take the form of rapid-fire imitation. I’d like to buy Boushey’s book, but it is crazily expensive, thus limiting the diffusion of his ideas.