Over the course of this year, Boston Review has published four essays drawn from a lecture series at Harvard University. That series launched a five-year research project to understand and help to remedy the problem of “institutional corruption.”
Institutional corruption does not refer to the knowing violation of any law or ethical rule. This is not the problem of Rod Blagojevich, or, more generally, of bad souls acting badly. It instead describes an influence, financial or otherwise, within an economy of influence, that weakens the effectiveness of an institution, especially by weakening public trust in that institution. (An “economy of influence” rather than the simpler “system of influence” to emphasize the reciprocal character of such influence, often requiring little or no direct coordination.)
Congress is a paradigm case. Members of Congress run privately financed campaigns. The contributions that fund those campaigns are not illegal, or even unethical. To the contrary, they are protected speech under the First Amendment.
Yet arguably—or maybe obviously—those contributions are (1) an influence (2) within an economy of influence that has (3) (quite likely) weakened the ability of Congress to do its work, by (4) (certainly) weakening public trust in Congress. The vast majority of Americans believe money buys results in Congress; less than a quarter of Americans believe the institution worthy of their trust. When “free-market” Republicans vote to support milk subsidies or sugar tariffs, or when “pro-consumer” Democrats vote to exempt used-car dealers from consumer financial-protection legislation, it is easy to understand the mistrust and hard to believe that the influence of money hasn’t weakened the ability of members to serve the principles, or even the interests, they were elected to represent.
This is “corruption” not because it describes the acts of evil or corrupted individuals. Members of Congress are insanely hard-working, decent souls, the overwhelming majority of whom entered public service to do good, as they see it. And indeed, the traditional corruption of politics—bribery—is likely at its lowest point in American history. From the perspective of criminal law, this is among the cleanest Congresses ever.
Instead, this is “corruption” because it weakens the integrity of the institution, of Congress itself. The framers intended Congress to be “dependent upon the People alone.” But the private funding of public campaigns has bred within Congress a second, and conflicting, dependency. As with an alcoholic mother trying to care for her children, that conflicting dependency does not change the good intentions of members of Congress—they still want to serve the public interest they thought themselves elected to serve. But as with an alcoholic mother trying to care for her children, that conflicting dependency distracts members from their good intentions, directing their focus more and more toward the challenge of raising money.
Such a dependency would distract any of us. Institutional corruption is not a particular weakness that affects a particular class within a particular institution. It is instead, as cognitive psychology and behavioral economics increasingly reveal, among the most human of reactions that evolution has produced. We all reward those whom we depend upon, whether or not such reward is consistent with our ideals or objectives. And we are especially prone to reward those whom we depend upon when our conduct is viewed as justified within the institution with which we are affiliated.
The solution to this form of corruption is not to extirpate human nature. Nor need we accept the corruption as simply part of the human condition. The solution, as the framers of the Constitution self consciously described, is to design institutions that avoid these conflicting dependencies. We can’t presume that “angels [will] govern.” But neither need we believe that every member of Congress is evil in order to recognize that a proper constitution prevents systems of dependence that draw representatives away from the interests they were elected to represent and the values they were chosen to serve. Indeed, when one studies the tiny dependencies that our framers sought obsessively to avoid (my favorite are the constitutional prohibitions intended to prevent dependence upon “foreign princes”—if only the framers had said “princes, foreign or domestic”), it is impossible to believe they wouldn’t be astonished by the enormous dependencies that have evolved within their design. The one dependence they had in mind —“upon the People alone”—has almost become an afterthought.
• • •
On January 21, 2010, a year and a day after Barack “change the way Washington works” Obama was sworn into office, the Supreme Court rendered the environment for this institutional corruption much worse. By a five-to-four vote, the Court in Citizens United v. Federal Election Commission (FEC) secured to corporations a right that individuals have had since the last equally fundamental Supreme Court decision on campaign finance (Buckley v. Valeo, 1976): the right to spend an unlimited amount on independent campaign expenditures. “Independent expenditures”—not contributions to candidates. And “independent expenditures”—not coordinated with the campaign. No doubt, as corporations exercise this new right, candidates will become skilled in the dance necessary to get enormous corporate wealth spent for their political benefit.
At issue in Citizens United was a limitation in the Bipartisan Campaign Reform Act of 2002 (BCRA, commonly known as McCain-Feingold). BCRA forbade any company from using corporate funds to influence an election within 60 days of that election, even if the funds were being used simply to discuss an issue, although in a way that plainly demonstrated support for or opposition to a candidate. Citizens United, a nonprofit corporation, wanted to use funds from its own treasury to promote a film critical of Hillary Clinton. BCRA blocked that use because the film in effect advocated voting against Hillary Clinton. Citizens United challenged the law as an unacceptable abridgment of its freedom of speech.
Citizens United might have spoken through other channels—for example, by setting up a political action committee (PAC) with its own treasury to which individuals could contribute. And the individuals who had created Citizens United were still free (because of Buckley) to spend independently as much as they wanted to advance their interests. Nonetheless, the Court held that Congress had no power to block the corporation itself from spending funds on political speech.
Under our current system of campaign finance, there is a fundamental gap between the interests of voters and of contributors.
There is a core aspiration in the Citizens United opinion with which everyone ought to agree, but there also is much to quibble with, and there is much more that is very confused.
The Court was right to raise a red flag about a law that purported to block political speech.
Such a law should face a significant burden of justification, regardless of whether the target is the Chamber of Commerce or the Chinese government, and even though the relevant “speech” is not the direct words of a citizen, but words and video purchased with corporate money. I don’t believe vigilance against improper government regulation of speech should turn upon whether the regulated entity is a “person,” or whether the speech is pure speech or money-as-speech. Nor did the Court. Instead, the Court focused its inquiry on Congress’s power. Congress, the Court held, has no power to remove this liberty to speak, because, regardless of the speaker and irrespective of the means, Congress has no legitimate interest in silencing political speech.
Unless that speech amounts to corruption. But here the Court was extremely narrow, and mistaken in my view, in its understanding of corruption. The only corruption the Court found that justifies the suppression of political speech is “quid pro quo” corruption. And as independent expenditures are, by definition at least, made independently of a political campaign, they cannot constitute quid pro quo corruption. There may be a quid. There may be a quo. But because the two are independent, there is no pro.
This clarity was not cut from whole cloth. Since Buckley there has been a recurring suggestion that the only corruption that is constitutionally cognizable—the only kind that suffices for restricting political speech—is quid pro quo corruption. As the Court itself said in FEC v. NCPAC (1985): “Corruption is a subversion of the political process,” and the “hallmark of corruption is the financial quid pro quo: dollars for political favors.”
But in 1990 the Court suggested a more expansive definition of corruption. In Austin v. Michigan Chamber of Commerce the Court upheld a Michigan law that prevented corporations using money from their general treasuries to fund independent expenditures in candidate elections. The Court thus recognized another kind of corruption, beyond quid pro quo, that would come from the
corrosive and distorting effects of immense aggregations of wealth that are accumulated with the help of the corporate form and that have little or no correlation with the public’s support for the corporation’s political ideas. (Emphasis added.)
So a divided Court held that campaign spending could be regulated in order to prevent these effects as well. And after Austin, no opinion of the Court again said that quid pro quo was the only way to corrupt the political process capable of justifying a restriction on speech. Until, that is, Citizens United, which overruled Austin.
Yet why this is the only sense of corruption that might justify a restriction on speech is not clear. The argument on this essential point in Justice Kennedy’s majority opinion in Citizens United seems almost an afterthought. It is perhaps best captured in two closely related passages.
First, to the suggestion that there may be a corruption beyond quid pro quo tied to the special influence that money has within our political system, Justice Kennedy quotes an earlier opinion of his:
Favoritism and influence are not . . . avoidable in representative politics. It is in the nature of an elected representative to favor certain policies, and, by necessary corollary, to favor the voters and contributors who support those policies.
Notice the words “and contributors.” Without those two words, the statement is certainly true. The claim could be made even more strongly: favoring the policies that one’s constituents favor is the essence of representative democracy (on at least one dominant conception of it). It was for the purpose of establishing precisely this sort of dependency of representatives on constituents that the framers created frequent elections in the House. As Federalist No. 57 put it, “The House of Representatives is so constituted as to support in the members an habitual recollection of their dependence on the people.”
But adding the words “and contributors” not only makes the statement not obvious (as Justice Kennedy seems to believe), but in my view, plainly wrong. The framers did not intend to make representatives dependent upon contributors. Representatives were to be dependent upon voters, or, more generally, “the People alone.” It is conceivable—assuming many contingencies—that a dependence upon contributors could in effect be the same as a dependence upon voters. For example, to borrow an idea from Yale’s Bruce Ackerman and Ian Ayres, if elections were funded by what we could call “democracy vouchers,” where citizens could allocate up to $50 to any candidate and candidates could receive money only from citizens in their districts, then a dependency upon voters might well be the same as, or close enough to, a dependency upon contributors. Candidates would have to spend time talking to voters, if only to raise money from them. And because of the small size of the vouchers, the dependence would be relatively equal. In this system, the sort of thing you need to do to make voters happy is the same sort of thing you need to do to make contributors happy.
Under our current system of campaign finance, however, there is no such overlap between the interests of voters and of contributors. There is instead a fundamental gap. That gap has many dimensions, but the simplest to remark is the geographical source of campaign funds. In the most critical cases, the vast majority of contributions are not even from the voters. Opensecrets.org reports that 67 percent of contributions to John Kerry’s 2008 senate-reelection campaign came from outside Massachusetts. His Republican opponent received 73 percent of his funding from out-of-state donors. Maplight.org reports that since January of 2007 79 percent of contributions to California state legislators have come from out-of-district contributors. This gap between contributors and voters means that responsiveness to one is not necessarily responsiveness to the other. Once more, the sort of thing you need to do to make contributors happy is not the sort of thing you need to do to make voters happy.
A vibrant free-speech tradition need not prevent the regulation of institutional corruption in Congress.
This leads to the second critical passage from Citizens United, casually dismissive of the problem of institutional corruption:
The appearance of influence or access . . . will not cause the electorate to lose faith in our democracy.
Among constitutional lawyers, this sort of claim is called “Lochnerism,” and the moniker is not high praise. It derives from an infamous 1905 case, Lochner v. New York, in which the Supreme Court struck down New York’s Bakeshop Act of 1895. That progressive statute included a maximum-hours requirement for workers in the baking industry and established minimum conditions for plumbing, flooring, product storage, and sanitation. The Republican-dominated state legislature passed the regulation twice, each time unanimously. In defending this statute before the Supreme Court, lawyers for the State of New York said the regulation was a health measure, intended to improve the well-being of overworked bakers.
The Supreme Court, however, rejected New York’s argument. On the Court’s view, the regulation of hours worked had no bearing on a worker’s health. The statute thus must have had a different motivation behind it. And as the Court sussed that motivation out, it claimed that New York’s true aim was to benefit labor at the expense of capital. The law was simple redistribution, and the Fourteenth Amendment, the Court held, banned such a taking from A to give to B.
But the question of whether the regulation does or does not affect the health of bakers is a factual one, not a matter of constitutional interpretation. So one might rightly ask: on what factual basis did the Court rest its judgment?
The answer is none. The Court had no evidence for its claim, while the State of New York had plenty. An 1894 report by the New York State Labor Commissioner found that baking was an especially dangerous occupation, in part because of the extended hours of the baker. After a 30-year war between courts and legislators, the Supreme Court finally decided firmly that it was no longer in the business of second-guessing state or federal legislatures about the social conditions they purported to regulate. As the Court said in West Coast Hotel v. Parrish (1937): “Even if the wisdom of the policy be regarded as debatable and its effects uncertain, still the Legislature is entitled to its judgment.”
It is the view of every current justice on the Supreme Court that Lochner was mistaken, and that the Court has no business second-guessing legislatures, at least when they pass regulations that restrict the liberty of contract. But when it comes to the First Amendment, Lochnerism is alive and well, and Justice Kennedy’s opinion in Citizens United may well mark an extreme. For whether or not “the appearance of influence or access” will “cause the electorate to lose faith in our democracy” is a factual issue. It is not resolved by the observation that the expenditures merely are intended to persuade voters, and thus that the system, as Kennedy put it, “presupposes that the [voters] have the ultimate influence over elected officials.” “Ultimate influence” isn’t necessarily the issue; it may be that relative influence is all that matters to whether “the electorate [will] lose faith in our democracy.”
And given the radical inequality between interests and expenditures, a voter could well reason: “even if it is ultimately only votes that count, whether you get votes depends upon whether you can outspend your opponent; whether you can outspend your opponent in turn depends upon whether you can raise enough money or inspire others to spend enough money on your behalf; thus, it is the race for campaign money that is the real race in American democracy; and thus,” this voter could well conclude, “there is every reason in the world to lose faith in this democracy—conceived of as a system in which representatives are to be dependent upon the people.”
Kennedy’s opinion puts the factual question off the table. He hasn’t proven his conclusion. He hasn’t even offered evidence to suggest it is true. He has, like the Court in Lochner, simply asserted what those without life tenure on a federal bench would have to prove. And as empirical evidence certainly contradicts Justice Kennedy’s assertion—many Americans have “lost faith in our democracy,” and the role of money in elections is high on the list of reasons for their skepticism—this keystone to the Court’s decision is not likely to survive very long.
To be clear: I am not saying that the Court should treat First Amendment questions the way it treats questions about the regulation of contracts. I, along with every justice on the Court, believe that free speech deserves a much more vigorous defense against improper legislation than the liberty of contract. But one could believe that more stringent standards for regulating speech are justified, yet also reject what the Court did in Citizens United. The Court should not be able to assert the facts that purportedly justify its negation of legislation with nothing more than analogical reasoning to back its claims. The rejection of Lochner could mean that, in most contexts, the legislature gets an almost-free pass from review. But the Court should never be entitled to a similar free pass. When invalidating legislation, courts should have to work harder than this.
• • •
Maybe Kennedy could have done better. Maybe his point is that legislatures only can regulate the First Amendment right to speak in the name of preventing quid pro quo corruption, that no other corruption of democracy justifies a regulation of speech. Bribery, Kennedy would say, is a special case. It is of course legitimate that a legislature forbid “speech” that amounts to bribery: regulations against quid pro quo corruption are thus acceptable, even though they restrict speech (and especially, the power of “money as speech”). But beyond that narrow class of corrupting speech there is no other that could weigh against the right to speak.
The point is clear enough. But is it true? Do the reasons for regulating speech that offers a bribe only justify regulating bribery? Or do the same reasons cover a wider range of conduct that is arguably at odds with the democratic process—with the exclusive dependence on the people intended by the framers?
We recognize the right of the state to ban bribery, even though bribery is speech. The interest of the state in banning this form of speech overwhelms the right of the briber or the bribed to engage in it. That is because bribery short circuits a deliberative or democratic process. When a congressman inserts an earmark into an appropriations bill in exchange for $10,000, we presume the reason he had for inserting that earmark had nothing to do with the merits of the earmark. We presume, in other words, that the legislative decision was guided by a venal interest rather than a public interest.
The Court was not oblivious to the ways in which expenditures might affect the political process. But the justices certainly were ignorant about them.
But why do we so presume? What if the congressman could prove that he would have inserted the earmark anyway? What if he said, “I only take bribes to do what I otherwise would have done”? And what if we could be certain that the venal interest was irrelevant to the inclusion of the earmark, that the congressman had only the public interest in mind?
I take it as uncontroversial that the law should reject such a defense. Maybe the congressman’s claim is true. But as well as protecting against decisions by legislators that are in fact guided by venal rather than public reasons, it is perfectly appropriate for a legislature to protect against decisions that appear to be guided by venal rather than public reasons, even when they are not. As Buckley says, the concern is with “preventing corruption and the appearance of corruption.” It would be too much to expect the public to ignore the potential venality of a bribe—no one believes the First Amendment requires that prosecutors prove that, but for the bribe, the decision would not have been made.
Bribes are thus instances of personal corruption. They are regulable because we presume they corrupt legislators by distracting them from their proper focus—on legislation that serves the public interest.
Institutional corruption brings this same concern to the level of the legislature. If a legislature can regulate to keep individual legislators from making decisions that are dependent upon venal rather than public interests, why can’t it regulate to keep the legislature as a whole from making decisions based on improper dependencies? If it can act to assure that individual legislators don’t act, or seem to act, on an obviously improper dependency, why can’t it act to assure that the legislature itself not act, or seem to act, on a different, but equally improper, dependency?
A dependency, for example, upon funders rather than voters. If the framers were clear that legislators are to be dependent only on the people, is it permissible for a legislature to protect itself from other dependencies? Kennedy has not shown why not.
In short, bribery is regulable because it manifests personal corruption, the deviation from a public to a venal interest. Yet Kennedy has offered no argument as to why acts that constitute institutional corruption shouldn’t likewise be regulable, as manifesting, improperly, a dependency upon funders rather than upon voters.
My suggestion is that we therefore recognize two kinds of corruption. One is personal, or venal. This is bribery. The legislator takes money for a decision, giving rise to the presumption that he was acting not for the public interest, but for his own private, venal interest.
The other is institutional, or what we could call, building on the work of Fordham University professor Zephyr Teachout, “dependency corruption.” Dependency corruption happens when an institution develops a dependency that conflicts with its intended dependency. This corruption doesn’t look to evil acts by evil individuals; it looks instead to the economies of influence that have developed within a designed system. Regulation of dependency corruption aims to prevent the deviation of an institution from its proper or intended focus, just as regulation of quid pro quo corruption aims to prevent individual deviation from proper focus.
The charge against the current system of funding congressional campaigns thus comes to this: not only has it eroded trust in Congress (the Pew Research Center’s latest numbers indicate trust in government and faith in Congress both at historic lows), it has also engendered a focus on interests distinct from the interests of the voters. A people should have the power to avoid just this sort of distraction, a.k.a., corruption. A vibrant free-speech tradition need not disable that power.
• • •
Even if my analysis is correct, a defender of the decision in Citizens United could insist that the expenditures at issue in that case were still independent expenditures, a step removed from any economy of dependence. And if they are indeed independent, how could they generate an improper dependency?
They need not, of course. My argument is not an argument of grammar. It is perfectly possible that massive expenditures made independently of a decision-maker have no plausibly corrupting effect on the decision-maker. For example, millions are spent each year writing amicus briefs for the Supreme Court—sometimes on a single case. Yet no one thinks that the expenditures themselves, as distinct from the arguments in the briefs, affect the decisions of a justice. As the money spent there is irrelevant to the justice, the money does not in this sense corrupt.
But independent expenditures are not benign in every case. There was ample evidence presented in Citizens United that campaigns are aware of the independent expenditures made on their behalf. Indeed, there was plenty of evidence that the expenditures were reported directly to the campaigns. And while only the venally corrupt or the stupid would expressly link those expenditures to some kind of action, there are plenty of ways to coordinate without coordination. To paraphrase Boston political boss Martin Lomasney, Washington is the kind of city where one never writes if one can call, never calls if one can speak, never speaks if one can nod, and never nods if one can wink. Little doubt that in such a rich environment of reciprocity, open deals are rarely essential.
The Court, of course, was not completely oblivious to the ways in which such expenditures might affect the political process. But, with respect, the justices certainly were ignorant about them. The Court conceded that large expenditures by a corporation could raise questions. But the Court also insisted that those concerns could be addressed adequately by a regime of transparency in which all expenditures must be disclosed. The First Amendment, that is, need not be seen to deny the danger of corruption through independent expenditures; it instead simply limits the remedy to an acknowledged danger. And here, the Court thought, the constitutionally permissible remedy to a danger from large, tacitly coordinated expenditures is to require the speakers to disclose.
So long as First Amendment ‘Lochnerism’ prevails on this Court, Congress will be unable to address institutional corruption.
But in this simple solution, the Court is overlooking an important way in which money—the right to spend—actually influences campaigns.
The assumption of the Court was that $1 million of influence would be disclosed as $1 million of influence and that voters could then reckon whether that level of influence should matter to their decision. Yet as economists Marcos Chamon and Ethan Kaplan have argued in their “iceberg theory” of campaign finance, the influence from campaign expenditures may be independent of the amount actually spent. The influence on any particular candidate, they maintain, also would depend on the credible threat of expenditures to benefit the candidate’s opponent.
For example, the incentive effect of a $10,000 contribution to a candidate could be the equivalent to the incentive effect of a $2,000 contribution if the $2,000 contribution is bundled with a credible threat to contribute $8,000 to the candidate’s opponent. The threat creates its own incentive. The more credible the threat, the greater the incentive.
These threats, however, are not reported on any campaign disclosure form. In the example above, the $2,000 contribution would have to be reported, but the $8,000 threat would not. The $2,000 is thus the visible tip of the iceberg, while the $8,000 is the bulk, hidden from the public’s view.
Chamon and Kaplan wrote in the pre-Citizens United world, where the maximum “corporate contribution” through a corporate PAC was $5,000 per cycle. But the significance of their theory in a post-Citizens United world is much greater, for the power of a potential threat is limited by the maximum contribution allowed. After Citizens United those limits are removed. And while the threats must be of independent expenditures, there are obvious new ways in which corporate wealth can now translate into significant political influence.
Imagine, for example, that Exxon let it be known that it was willing to spend up to $1 million in any congressional district to promote candidates who are skeptical of global-warming science. Or imagine that Google let it be known that it would run up to $1 million in online ads to defeat global-warming skeptics. Neither position would necessarily be “coordination” sufficient to render the expenditures non-independent: both announcements could be made well before candidates are even chosen by parties. Yet, if the iceberg theory is correct, in neither case would all the money have to be spent in order to have its intended effect. Google might actually spend only a thousand dollars, and it would report that amount. But its influence would be far beyond what it reported, so long as its threat was credible.
Does the influence or threat of independent expenditures within the economy of influence that is privately funded campaigns further weaken the effectiveness of Congress to do its job? Does it further weaken public trust of Congress by confirming that “money buys results”?
Again, the answer depends on the facts. It is not simply a matter of logic, but instead hangs upon the actual expectations of an actual public. I have my intuitions. Maybe they are wrong. Maybe American cynicism is already so great that even a radical increase in a conflicting dependency won’t further weaken public trust.
Lochnerism, however, will not permit us to know. So long as First Amendment Lochnerism prevails on this Court, so long as Justices are prepared to let their own factual speculations trump legislative fact-finding, and so long as judicial intuitions about the impact of disclosure requirements are permitted to decide the issue, Congress will be unable to address institutional corruption directly by limiting expenditures that create a reality or appearance of unacceptable dependency.
Yet Congress could still act in less-direct ways to prevent dependency corruption. It could, for example, adopt a voluntary, small-dollar contribution system under which candidates agree to take no more than, say, a hundred dollars from any citizen, and the government matches that amount with some multiple. Or Congress could enact the Ackerman-Ayres system of democracy vouchers. Both alternatives would provide important reform to the current system. The former is the idea behind the Fair Elections Now Act, whose sponsors estimate that it could raise more money for campaigns than the current system does. The latter would be even more dramatic. At $50 a voter, the voucher system would pump almost four times the amount spent on congressional campaigns in 2008 into a non-corrupting form of campaign finance.
But the blindness to this more fundamental corruption—a corruption that has likely done more damage to the United States in the last decade than all of the quid pro quo corruption in American history combined—will continue to limit the imagination and conduct of both courts and legislatures. We need a clearer way to mark this type of good-soul, institutional corruption. We need a broader recognition of its harms. Because the idea that the protections of the First Amendment would bend to allow punishment of small-time crooks like Rod Blagojevich and Duke Cunningham, yet block efforts to reform the economy of influence that has produced the Great Recession, or to confront global warming, is just crazy. We need a better lens to show why.