Modern liberals embrace an ideal of democracy that requires relative equality of political voice. To advance that ideal, they demand government regulation of political speech. And they are troubled that our Constitution has failed to evolve in line with their theory.

Buckley v. Valeo (1976), the foundation of contemporary campaign-finance regulation, rejects government efforts to equalize relative political voice as “wholly foreign to the First Amendment.” According to Buckley, the only basis for regulating campaign finance is the prevention of “corruption and the appearance corruption.” Austin v. Michigan Chamber of Commerce (1990) strained to characterize large inequalities as a form of corruption, illustrating the awkward position of liberals forced to milk justice from the bull of anti-corruption. To make matters worse, Citizens United v. FEC overturned Austin, reaffirmed Buckley’s quid pro quo approach to corruption, and opened the floodgates for unlimited “independent expenditures” by for-profit and nonprofit corporations.

Lawrence Lessig aims to get the Constitution back on the right track. He criticizes the quid pro quo paradigm of corruption as unjustifiably narrow and defends a more expansive conception of institutional corruption. This more expansive conception would justify bans on the political speech of organizations taking the corporate form, and also mandate government-financed elections. The Court’s big mistake in Citizens United, Lessig argues, was to engage in footloose speculation about the conditions that lead citizens to lose faith in democracy. Lessig’s big mistake is to engage in his own footloose speculation about those same conditions.

Institutions exist within a nexus of social forces or, as Lessig likes to say, an “economy of influence.” That economy of influence is corrupting in his expansive sense just in case it undermines an institution’s effectiveness, “especially by weakening public trust in that institution.” Lessig is concerned with Congress in particular. The framers intended, he says, that Congress be “dependent upon the People alone.” But when political campaigns raise private funds from willing members of the public (instead of, say, receiving grants from the government), candidates become dependent on these funds, thus not dependent on the people alone. Dependency on a bunch of people’s money is not dependency on the people, since some people give a lot more than others. We know which side our bread is buttered on, and lavish attention on that side. Thus are “good soul” legislators bent toward the service of special interests and away from the public interest.

The share of the population describing government as corrupt went down even as soft-money contributions skyrocketed.

So far, so logical. But what exactly is the evidence about the effect of campaign contributions on legislators? Studies that control for party, ideology, and constituency show little if any relation between campaign contributions and roll-call votes by members of Congress. There certainly are studies that show small effects, though Lessig mentions none. Perhaps the paucity of evidence about corrupting influence explains why Lessig at times seems to abandon his account of how public policy should depend on public opinion alone and to rely instead on the idea that insufficiently regulated spending diminishes popular trust in Congress.

Lessig expresses dismay at Justice Kennedy’s claim in Citizens United that “the appearance of influence or access . . . will not cause the electorate to lose faith in democracy.” Kennedy hasn’t proven this, Lessig insists. “He hasn’t even offered evidence to suggest it is true.” Lessig calls this practice of, as he sees it, dogmatically protecting liberties on the basis of evidence-free speculation, “Lochnerism,” after the early twentieth–century case of Lochner v. New York. But his response to Lochnerian armchair speculation is to offer his own evidence-free speculation about what “a voter could well reason” that could well lead her to lose faith in democracy. Lessig’s vehemently empiricist rhetoric encourages the reader to expect rather more.

Of course, that Justice Kennedy offers no evidence to the effect that “the appearance of influence or access . . . will not cause the electorate to lose faith in democracy,” does not mean there is none. A glance at long-term trends in American trust in government shows variation that appears to have nothing to do with campaign finance or speech regulation. Over the past 40 years, trust in government briefly peaked after 9/11 and then plummeted rapidly to dark post-Watergate depths. The 2002 passage of McCain-Feingold, meant to restore Americans’ faith in Congress, did nothing to forestall the profound deterioration of public trust.

In a recent paper, Nathaniel Persily and Kelli Lammie, scholars concerned with political communication, report, “the share of the population describing government as corrupt went down even as soft-money contributions skyrocketed.” Their data—which, on the whole, show that a majority of Americans distrust Congress—indicate that beliefs are unmoored from the facts. Similarly, political scientist David Primo and economist Jeff Milyo find “little impact of state campaign finance laws on perceptions of government.” Elsewhere Primo and Milyo have reported further “evidence that campaign finance laws are unlikely to be an effective means for improving the ‘quality of democracy.’” If Lessig knows something Justice Kennedy doesn’t, he should share.

Moreover, Lessig’s idea that public perceptions can justify the abridgment of a fundamental constitutional right is troubling. Basic constitutional rights are supposed to be protected from the fleeting disposition of the volkgeist—sheltered from that particular economy of influence. The public perception of the relationship between legislation and the people’s will is just one aspect of public opinion among many, and no less likely to be in error. Why should one potentially mistaken aspect of public opinion stand as valid justification for the restriction of the fundamental right to attempt to alter public opinion through speech? This is an especially pertinent question when we consider the real possibility that the sources of speech Lessig wants to ban might improve the quality of political information available to the public, including the quality of information about such things as the dependency of policy on public opinion.

In short, Lessig offers no evidence that the reforms he seeks would have any effect on the distrust he laments. And even if he has some evidence, we should find ways to improve popular trust in democracy other than restricting the liberties that make democracy possible.