Introduction
We all know the story. Received wisdom says that our federal system of government encourages high levels of policy experimentation. “It is one of the happy incidents of the federal system,” Justice Louis D. Brandeis wrote, “that a single, courageous State may . . . serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.”
The basic idea is simple: The Constitution divides power and responsibility between the federal and state governments, giving states reign over their own affairs. That arrangement encourages state officials to compete for a mobile tax base by inventing better policies, and it allows them to tailor state law to local conditions. “Our Federalism” thus creates fifty state “laboratories,” whose officials toil to “devise[] solutions to difficult legal problems.”
And the nation benefits from knowing whether, and under what conditions, those solutions work.
In the decades since Justice Brandeis penned his memorable opinion, countless scholars and judges have spun the same yarn about federalism’s role in promoting policy experimentation. The laboratories account has been deployed in thousands (yes, thousands) of academic works.
And the Supreme Court has invoked it in scores of decisions on topics far and wide.
In those decisions, the laboratories account often fits into a larger theory about how best to promote federalism values, such as choice, participation, competition, the diffusion of power, and experimentation.
In particular, the Court has sought to achieve those ends by carving out a policy space where states are autonomous—where federal law and federal officials may not intrude. The laboratories account aligns with this vision of federalism, as it seeks to prevent a federal behemoth from displacing states from domains where they can experiment.
The laboratories account has had a remarkable run, but it is little more than a campfire story. Even a cursory glance at states and their officials suggests that they are poorly equipped innovators. For starters, there’s the problem of resources. State officials tend to be overwhelmed by the many demands on their time and hamstrung by tiny budgets, short legislative sessions, and low levels of expertise. There are also problems of incentives. As Professor Susan Rose-Ackerman and others have argued, a state often has little reason to pioneer new policies when it can simply copy successful ones from other jurisdictions at a fraction of the cost.
Further, the potential electoral costs of endorsing unsuccessful policies will often outweigh the potential gains from endorsing successful ones.
Taken together, these obstacles to innovation suggest that major parts of the laboratories account are mistaken.
These obstacles, however, can’t be the end of the story. States are, in fact, flourishing sites for policy innovation. Just in recent years, they’ve pioneered a huge range of policies—from fracking to climate-change laws; from voter ID laws to sanctuary cities; from LGBTQ civil rights to protections for the “right to work”; from enhanced firearm restrictions to stand-your-ground defenses.
On the right and the left, ideas for new policies are often first enacted in the states. While some legal scholars have criticized the laboratories account, no one has offered a satisfactory affirmative account of how states and their officials manage to enact major innovative policies despite the obstacles to innovation mentioned above. The primary task of this Article, then, is to figure out where the laboratories account goes wrong and to propose a better account in its place.
This Article maintains that the laboratories account focuses on the wrong actors. It focuses inward, viewing state policies as the output of officials working within state governments to promote local interests and concerns. But it should focus outward—on interest groups, activists, constituency-mobilization organizations, advocacy coalitions, donor consortia, and other third-party organizations that aggressively advocate for their preferred policies.
To illustrate the influence of these groups, consider the state of Iowa in the aftermath of the 2016 elections. Republicans had just gained total control of Iowa’s state government for the first time in nearly two decades. One of the first items on the agenda was changing the legal rules governing public sector labor unions.
The bill that was ultimately signed into law drew heavily from a proposal by the conservative-leaning American Legislative Exchange Council (ALEC) and was championed by several ALEC members.
Policy briefs published by an affiliate of the State Policy Network (SPN)—a national association of right-leaning think tanks—advocated for the bill.
And lawmakers were pressured to publicly pledge support for the bill by the Iowa chapter of Americans for Prosperity (AFP)—a national libertarian advocacy organization established by industrial magnates Charles and David Koch—which launched a “grassroots” campaign that included mailers, advertisements, and constituent outreach.
Together, these groups both made the lawmaking process easier for Republicans who antecedently supported the union-busting legislation and exerted considerable electoral and social pressure on any lawmaker who would have preferred to move more slowly.
As this example illustrates, policy innovations are often devised and then propagated by third-party organizations connected to state officials through political networks. For many of the most important state policies, these organized interests are the true “laboratories of democracy,” as they catalyze policy experimentation in several crucial ways. They inform public officials about important social issues; propose solutions supported by bespoke research; provide model legislative text and talking points tailored to local conditions; create electoral incentives and social connections that conduce to policy experimentation; and use the federal government’s power to spark state innovation.
State officials aren’t so much the scientists responsible for many of the most important policy innovations as they are the test subjects on which the real laboratories of democracy can experiment.
The prominent role of intense policy demanders in creating and diffusing state policy is familiar to close observers of American politics. But it has failed to elicit a change in federalism doctrine, and a fair amount of federalism scholarship remains oblivious to it. If one takes that role seriously, it becomes clear that the Court’s federalism doctrine needs to be revised. For starters, state autonomy shouldn’t be the desideratum. There is no reason to think that leaving states to their own devices will result in higher levels of experimentation since the third-party organizations behind many of the most important policy innovations aren’t typically motivated by the possibility of giving one state a competitive advantage over others or tailoring its policies to local conditions. Instead, these intense policy demanders seek to push their agendas in any jurisdiction where they have political leverage.
Moreover, many influential state policy experiments take place within federal regulatory regimes. Creating separation from these regimes would thus sometimes inhibit state experiments that wouldn’t otherwise occur. In short, there is little gained and much lost from attempts to maintain separation between the state and federal governments.
At the same time, it’s important to acknowledge that the federal government’s wide-ranging power can be used to interfere with state experiments. The federal government can impose draconian funding conditions designed to put an end to state policies; it can try to co-opt state officials into regulatory regimes with which they disagree; and it can broadly preempt state law. But the axis along which federal–state rivalry occurs is less intergovernmental than it is interpartisan.
The threat to experimentation with which we should be most concerned isn’t federal encroachment on state domains—rather, it is a federal government controlled by officials with one set of partisan interests using its power to thwart the experiments of states controlled by another set of partisans. The problem with the autonomy model is that it proposes a separation-of-powers solution to what is fundamentally a separation-of-parties problem.
This Article proposes an alternative to the autonomy model that doesn’t require the futile task of keeping the federal government out of the states’ domains. Instead, the proposal aims to obtain the right conditions for federal–state bargaining within areas of overlapping jurisdiction. Put differently, the goal isn’t to prevent federal–state fights by keeping the two sides separated—it is to establish ground rules to push the fights toward beneficial forms of contestation. While the Article does not aim to identify the complete set of doctrinal rules that courts following this model should adopt, it concludes by sketching some potential implications for several areas of doctrine, including federal preemption of state law, conditional spending, and the anticommandeering principle.
The argument proceeds in three parts. Part I describes the laboratories account in more detail and outlines several obstacles inhibiting state officials’ propensity to experiment with new policies. Part II explains the role of networks of organized interests in the creation and diffusion of state policy innovations and how they help state officials overcome obstacles to innovation. Part III identifies the implications for legal theory and doctrine.
Before proceeding, two preliminary points are in order. First, this Article follows the convention of political scientists in using the term “innovation” to refer to “a program or policy which is new to the state[] adopting it.”
That sense of the term differs from another common sense according to which something is an “innovation” only if it’s better than the thing preceding it.
This Article uses the more value-neutral definition because it’s more tractable—not everyone will agree about which policies are genuine improvements over their predecessors. Indeed, one of the most powerful justifications for our federal system of government is that Americans have different conceptions of the “good” and thus different ideas about which policies to adopt. Attempting to study “innovation” in its value-laden sense would lead to disagreement as to which phenomena count as part of the study. Further, there is value in determining whether particular doctrines, frameworks, and systems conduce to producing “democratic churn,”
quite apart from whether the changes they tend to produce align with any particular conception of the good. When jurisdictions implement new policies, they show us how those policies fare relative to their predecessors, and that information may allow us to improve our decisions in the future.
Second, this Article is primarily interested in the influence of the laboratories myth on our constitutional discourse and the development of constitutional law. Accordingly, its inquiry is limited to policy domains where clashes between the federal and state governments are likely to occur—where the threat to state experimentation is at its peak.
The Article thus sets to one side policies that are idiosyncratic to a particular jurisdiction or that otherwise lack national salience. The potential for our federal structure of government to encourage the latter sort of experimentation is a point in its favor and would be relevant to someone trying to design a political system from scratch. But our interest is in whether and how courts can encourage policy experimentation by policing disputes between the federal and state governments.