Yale Law School Professor Heather Gerken has long argued that liberals should take a more favorable view of federalism. In an important new article on Vox, she outlines a variety of ways in which they could potentially use state and local power to resist the Trump administration. There are a number of parallels between Gerken’s argument and my own analysis of the same subject. For example, we both outline similar strategies that sanctuary cities could use to resist Trump’s efforts to deport undocumented immigrants.

Perhaps the most notable difference between Gerken’s position and mine is that she puts little if any emphasis on judicially enforceable limits on federal power. This is part of a longstanding disagreement between Gerken and other modern progressive champions of federalism on the one hand, and conservative and libertarian federalists on the other. In my view, the new progressive federalism would be on a sounder footing if its advocates accept the need for strong, binding constitutional limits on federal power rather than resist it. In some of her recent writings, Gerken has shown greater openness to such limits on than in the past. But I think she and other liberals should move further in that direction.

Gerken’s Vox article actually underscores this point very well, even if perhaps unintentionally. Many of her suggested strategies for resisting Trump implicitly depend on constitutional limits on federal power for their effectiveness. For example, her (and my) recommendation that sanctuary cities should refuse to cooperate with federal deportation efforts relies on Supreme Court decisions forbidding federal “commandeering” of state and local governments. Otherwise, Trump and the GOP-controlled Congress could simply enact laws ordering the states to comply, and potentially imposing severe punishment on officials who refuse to do so.

Both the sanctuary city policy and some of Gerken’s other ideas might be undermined if the federal government could use conditional spending grants to pressure dissenting states into obedience. As Gerken briefly notes, such pressure tactics are rendered more difficult by Supreme Court decisions requiring that any such conditions be unambiguously clear, related to the purpose of the federal grant in question, and not so sweeping as to be “coercive.”

Tighter constraints on federal power could also curb other dangers posed by Trump to blue states, such as efforts to undermine state-level marijuana legalization. In addition, liberal efforts to use federalism to resist Trump are more likely to succeed in both courts of law and the court of public opinion if they attract at least some conservative and libertarian support. That support is more likely to be forthcoming if it is based on acceptance of generalized limits on federal power that can protect right and left alike, as opposed to ad hoc opposition to specific Trump policies.

Some liberals will likely continue to oppose nearly all hard-wired structural constraints on federal power for fear that they might be used to impede federal efforts to protect racial, ethnic, and other minorities. But as both Gerken and I have explained in the past, greater political decentralization can often benefit vulnerable minorities, particularly under modern conditions. Moreover, it is possible to impose tighter limits on federal power in many other areas, while still leaving the federal government broad power to combat invidious discrimination by state and local authorities.

Ultimately, the greatest threat to both unpopular minorities and many other groups is a largely unconstrained federal government dominated by their political enemies. In a diverse and increasingly polarized society with deep reservoirs of partisan hatred, both right and left have much to fear from such concentrated power. Recent political history shows that neither side can hope to stave off the threat by establishing a stranglehold over Washington that eliminates the possibility that the other will return to power. Rigorous enforcement of tight constitutional constraints on federal authority cannot completely eliminate the danger posed by the combination of polarization and the vast power of the modern state. But it can make it less menacing than it would be otherwise.

I don’t expect Gerken and myself to reach a complete meeting of the minds on constitutional federalism questions. Still less do I expect the emergence of a more general cross-ideological consensus on the subject anytime soon. But it should be possible to narrow some of the differences, and to recognize that many on both right and left from keeping a tighter constitutional leash on federal power.