It happens often that calamitous election outcomes are attributed to democratic dysfunction, and reform proposals follow from there. Those Republicans who are not deeply cynical about voter ID–and there are a fair number of those–sincerely and mistakenly believe that they lose elections because of illegal voting. Years ago, they convinced themselves that structural bias would keep Democrats in power in the Congress until the nation adopted mandatory term limits. Democrats in 1971 famously believed Richard Nixon reinvented himself politically through dark manipulations of television advertising, and to weaken his reelection prospects, their ranks in the Congress enacted a short-lived law that limited candidate media expenditures.

Of course, not all reform proposals born of frustration are misconceived or doomed to fail. But there are risks, some of which are apparent in a new reform initiative from Professor Lawrence Lessig’s: litigation to establish that states may not allocate Electoral Votes on a winner-take-all basis (WTA). His argument is not entirely about a change in the rules to achieve preferred outcomes or prevent bad ones. He does contend that, by driving active campaigning to a handful of states, WTA limits meaningful participation in the election to a handful of states whose voters see the most paid advertising and enjoy–if that is right word–the most visits from candidates and their surrogates.

Professor Lessig’s next argument, however, is more political. He sees WTA as putting more power in the hands of older white voters in industrial heartland states. It is fairly clear, then, that this proposed litigation is a response to the election of 2016, grounded in the belief that a change in the allocation would work a shift in the balance of national political power. And we have seen a reaction like this before. In 2012, the Democratic nominee won Pennsylvania, and those who were bitter about WTA were Republicans state legislators who threatened to shift a Congressional-district based allocation. Indeed the same move among Republicans is now underway in the State of Virginia and Minnesota.

So this is one risk: guessing wrong, as the Republicans did in 2012. Had they succeeded then, the congressional district-based allocation of electoral votes would have benefited Hillary Clinton in 2016. Politics is not static, and the judgment about the political effects of specific allocations is hazardous.

Another risk is misreading, or reading narrowly, the requirements of democratic participation in the concrete setting of the politics of the era.

Progressives–who mostly make up the audience for the Lessig reform–are closely considering and beginning to practice the virtues of what Yale Law Dean Heather Gerken has termed a “progressive federalism” that mobilizes states and localities to give voice and effect to policies out of reach at the national level. In arguing the “progressive case for local power,” Gerken has written that:

Progressives have long leveraged local population concentrations into political power. Indeed, much of the most important work on progressive issues started at the local level. Take climate change: from green building codes to cap-and-trade, the bulk of the work on progressive issues is being accomplished outside of Washington. And even national standards on fuel efficiency and the like have emerged largely as a result of states like California using their policymaking power to prod national regulators to move forward.

Gerken’s analysis has found favor with analysts like Ilya Somin who would certainly not classify themselves as progressives but who agree that “federalism can help mitigate the dangers of a situation where there is severe ideological and partisan polarization.” Somin has written that by “decentralizing more power to the state and local level, or to the private sector, we can reduce the number of issues on which a one-size-fits-all decision has to be made.”

A state-centered progressive federalism is necessarily in tension with the Lessig proposal to change the allocation of electoral votes from WTA to a proportional system. Returning to the example of California cited by Gerken: It has adopted its environmental policies by majority decision-making rules. It is looking to break with the direction, or overcome the inertia, of policymaking at the national level. Those voters in California supporting the decision see it as their state’s choice, and most would likely resist the notion that, say, in 2016, they should have accepted a proportional vote allocation that would have awarded 18 electoral votes to Donald Trump.

This proportional share would not have elected Trump in 2016: he would have fallen short on a proportional allocation, and so would have Hillary Clinton. (The election would have gone to the House and Trump would have won there.) But we could imagine a case where this allocation would have made a difference, helping to elect a president who would then aggressively pursue policies opposed to those that most Californians actively support.

This national outcome might satisfy one criterion for democratic participation, “one person, one vote.” It would have failed if judged by another: the also democratic processes by which states and localities pursue vital policies within a polarized polity afflicted by increasing dysfunction and discord at the federal level.

There is more to be said about the complexities of a reform initiative like the one Professor Lessig has advanced, but the evaluation requires attention in the first instance to the question of the various ways in the current structure of our politics that democratic values are best advanced and respected.