Jake M. Grumbach has received well-deserved attention for his book Laboratories against Democracy: How National Parties Transformed State Politics, which helps to reignite the debate over federalism in the United States. It stands alongside Jamila Michener's Fragmented Democracy: Medicaid, Federalism, and Unequal Politics (2018) by calling out the shortcomings of a system that delegates more power to subnational governments than nearly any other political system. Both works elucidate the links between devolution of authority and inequality in policies ranging from access to the health care safety net to the ability to cast a ballot, showing how minority rights are put at risk by federalism in a multiracial society. Michener's book is rooted in the qualitative testimony of those who navigate state Medicaid programs, while Grumbach's book brings together impressive quantitative tests conducted on original datasets, paying close attention to causal inference.
The central argument that Laboratories against Democracy advances is that federalism, which has always left minority populations vulnerable to authoritarian and antidemocratic majorities, is particularly ill-suited for today's politics. Decentralizing authority to the states empowers polarized national parties and the coordinated coalitions aligned with them, advantaging well-resourced groups and opening the door to democratic backsliding. Grumbach makes this argument with great force, writes convincingly, and supports it with new data and sophisticated tests.
While not every piece of the argument is perfectly convincing on its own, when viewed as a cohesive whole, this set of theoretical contentions and empirical investigations will help to reframe the way that our discipline thinks of the role of states in the American system. Those who wish to challenge portions of it or defend federalism overall will be motivated to assemble evidence for joining a vital debate that has been dormant for too long.
The empirical portion of Laboratories against Democracy begins by drawing on a massive dataset of state policy choices and outcomes—with Grumbach gathering data on 35 policies to add to 100 policies measured by earlier works in the literature—to establish two central descriptive facts. First, the states have increasingly diverged in their policy directions over the past generation. Second, state divergence is increasingly tied to party control, buttressing Grumbach's argument that national, polarized politics have taken advantage of the flexibility of federalism to drive the United States in increasingly divided directions. He arranges these distinct policies into 15 different areas, looking at how states diverged on them in the 1970–1999 period compared to how they diverged in 2000–2014. Before reading this analysis, many state politics scholars might have expected there to be little difference between the eras. After all, states diverged in many key areas a generation ago in the massively different ways in which they crafted their Medicaid and welfare policies in the wake of the Great Society, in laws reflecting the culture wars that emerged in states during the 1970s, and in their fiscal policies during a series of recessions.
Yet by conducting the first systematic and comprehensive analysis of this key empirical question, Grumbach shows that there is far more variation across states today than there was in the 1970s through the 1990s and that there is now a stronger correlation between this variation and party control. Electorally, we know that red states have become redder and blue states have become bluer since the 2010 election especially, making the divided government structure that was once so common in statehouses much less frequent. What Grumbach's analysis demonstrates is that each party was able to move the states that they captured in the policy direction that they preferred, causing the states to diverge and to do so along party lines. Grumbach clearly lays out the normative implications. These trends exacerbate the “policy patchwork” problem of federalism—the fact that where you live in America determines which laws you live under and even what your life and health prospects may be. Breaking down these patterns by issue area shows that the partisan divergence has been particularly prominent with regard to civil rights, environmental policy, guns, health and welfare, immigration, labor, taxation, LGBTQ rights, and abortion rights. And note that Grumbach's analysis was all conducted before the Supreme Court issued its Dobbs decision, supercharging state policy variation in a most salient realm. In a null finding that Grumbach does not shy away from, there has been no divergence in the realms of education and criminal justice, which together account for the lion's share of state and local expenditures. Still, the overall message here is that states are pulling apart.
What is driving them toward the ideological poles? Grumbach argues that the polarizing force is not public opinion. This is the portion of the book most likely to generate academic debate that may lead to a revision of the initial finding. There is a strong tradition of empirical work showing a link between public opinion and public policy, demonstrating that states are indeed representative and responsive. The most analogous study here is Caughey and Warshaw's (2022) Dynamic Democracy, a book based on a series of papers published over the past several years in the discipline's top journals that, like Grumbach's book, draws on many surveys, multilevel regression, and poststratification to estimate public opinion over time in states to see whether it predicts shifts in policies. Caughey and Warshaw find strong evidence of responsiveness to citizen demands, while Grumbach's analysis reveals no linkages (except in the areas of LGBTQ rights and marijuana policy). Perhaps the difference stems from the longer time period that Caughey and Warshaw study, the way that they combine issue areas, and differences in the estimation strategy they employ. It would have been helpful if Grumbach had directly addressed the differences between his work and this other influential piece of scholarship, because it has important implications for his overall argument. If states are shifting and polarizing in response to changes in public opinion, then federalism can increase the representativeness of American government. But because Grumbach does not find evidence of dynamic responsiveness, he turns next to other explanations of who governs in the states.
For that, he assembles important new evidence that the inequality plaguing American politics at the national level is even more acute in the states. Drawing on original data about the background of campaign donors, he shows that donors to state legislative candidates are even more likely to be wealthy and to be white than are presidential or congressional donors. The unheavenly chorus sings with an even thicker upper-class accent at the state level, it appears. Grumbach also shows that donors affiliated with activist organizations such as environmental or antiabortion groups are more likely to contribute to the campaigns of the most ideologically extreme state legislators. As he admits, there is a chicken-and-egg problem here: do the contributions drive extremism, or do polarized positions attract money? But the starkness and scale of the apparent effect is striking, and Grumbach shows just how closely polarization in state legislative chambers tracks over time with the average percentage of funds candidates receive from donors linked to these activist organizations.
Laboratories against Democracy also adds to the literature on state policy diffusion when Grumbach takes on the quote by Justice Louis Brandeis that he plays off of in his book title. Brandeis saw it as a “happy incident” of our federal system that states could innovate and then learn from each other in their “laboratories of democracy.” Grumbach challenges this notion of “unbiased learning” by instead contending that states engage in “partisan learning,” emulating only the successes of states that match their party brand. He is not the first to critique the Brandeis model of policy diffusion. For instance, Graeme Boushey's Policy Diffusion Dynamics in America (2010) makes its central contention that “the ‘laboratories of democracy’ metaphor for incremental policy evaluation and emulation is insufficient to capture the dynamic process of policy diffusion in America.” And while work by Craig Volden has provided evidence of partisan emulation, the substantial contribution that Grumbach's analysis makes here is to uncover evidence of “partisan learning.” With measures of the political and the economic successes of new policies, he first tests whether success in one state predicts adoption in all others, regardless of party. It does not. Yet when he interacts success with a match in party control between the first state and other potential adopters, he finds that political success breeds policy adoption among copartisans. The states are thus laboratories for political experimentation, without any policy learning and with a partisan tilt to the political emulation. This clearly departs from Brandeis' vision, helping to put a final nail in the coffin of that justification for federalism.
Grumbach's closing analysis presents a new dataset, the State Democracy Index, which shows that states led by the Republican Party have adopted more of the backsliding policies that make up the index. It once again implicates the states, which were the main locus of the racially disfranchising laws in place from the late 19th through the mid-20th century, as the weakest point in American democracy. Empowering states to determine the “times, places, and manners of holding elections” is one of the Constitution's greatest vulnerabilities, it appears. Grumbach's original measure of “democracy indicators” in the states includes voter registration and turnout rates, voter wait times, access for voters with disabilities, and gerrymandering measures. Each of these is an objective measure of democratic performance. But also playing a major if not predominant role in the index are same-day registration, early voting, youth preregistration, no-fault absentee voting, and automatic voter registration policies—essentially the set of laws that today's Democratic Party has pushed for nationally and in the states. Even those strongly supportive of these laws, or opposed to the Republican-backed voter ID and felon disfranchisement laws that also drive Grumbach's measure, may admit that they do not necessarily belong in an objective measure of democratic performance without a much more extensive analysis or discussion. Also included are a set of criminal justice laws such as DNA exoneration, determinate sentencing, three-strikes laws, and the presence of the death penalty, all of which seem even further afield from democracy. Although all of these are vital state policies, they are measures of a concept distinct from democratic backsliding.
When Grumbach looks at what explains the position that states take on an index combining all of these items, he finds, unsurprisingly, that party control is the primary determinant. Given that the parties so clearly diverge on these crime policies and the mechanisms of voting, this is not shocking. The most convincing case that Grumbach could make, perhaps in a future analysis, would be to create a State Democratic Performance Index composed entirely of objective measures to see whether Democratic states outperform Republican-run states in the inclusiveness and fairness of their elections and district lines.
Overall, though, Laboratories against Democracy uncovers a series of original findings in areas foundational to the study of federalism and assembles them into a provocative and convincing argument that will be broadly influential. State politics scholars typically either celebrate federalism or take it for granted. Grumbach instead argues against devolving power to the states. He shows that states are being pulled further and further apart, that it is nationalized parties and activist groups rather than localized voters doing the pulling, that states do not truly act as laboratories, and that democracy may be dangerously backsliding in a subset of states. This is a powerful argument that will stimulate much debate to come.
