Chapter 34 — Money in Politics: Citizens United, Dark Money, and Whether Democracy Is for Sale

"Money is to politics as oxygen is to fire. You cannot run a competitive campaign without it. The harder question is what kind of fire money builds — and whether the public, knowing what it knows, has any way to control it." — paraphrase, Larry Lessig, Republic, Lost (2011)

"If you can suppress core political speech in the name of preventing corruption, you can suppress it in the name of preventing anything. The First Amendment was not written to be subordinated to good intentions." — paraphrase, Bradley Smith, Unfree Speech (2001)

The two epigraphs above are not opposed because one is right and the other wrong. They are opposed because they are answering different questions about the same phenomenon. Lessig is asking what an ungoverned money flow does to representative democracy; Smith is asking what a regulated money flow does to political liberty. Both questions are real. Both deserve answers. The American campaign-finance system is the unsteady compromise we have produced after fifty years of fighting about which question matters more.

This chapter opens Part V — Democracy Under Stress. The four chapters that follow this one address gerrymandering (Chapter 35), voting rights (Chapter 36), democratic erosion (Chapter 37), and proposed reforms (Chapter 38). All five chapters share a posture: empirically sharp on what is happening, scrupulously honest about what is contested, and refusing to translate either side's argument into the other side's vocabulary. If you are looking for a chapter that tells you whether Citizens United was rightly decided, this is not that chapter. If you are looking for a chapter that helps you read a Federal Election Commission disclosure, distinguish a super PAC from a 501(c)(4), follow a dollar from donor to ad, and engage in good faith with both the constitutional and the structural arguments at the heart of campaign-finance reform, you are in the right place.

We start with constitutional doctrine, because every empirical question in this domain is downstream of Buckley v. Valeo (1976) and Citizens United v. FEC (2010). We then walk through the regulatory architecture as it actually exists in 2026 — the contribution limits, the entity types, the disclosure regime — followed by the empirical magnitudes (how much money, from whom, to what end), the social-science evidence on what money does and doesn't buy, and the menu of reform proposals with serious objections from both sides. We close with a cross-national comparison, because most peer democracies have made very different choices, and understanding the alternatives is part of understanding the system we have.

By the end you should be able to (a) read a campaign-finance disclosure on the FEC website or OpenSecrets, (b) identify the legal entity through which any given dollar reached a campaign, (c) state the strongest case both for and against the Citizens United majority, (d) describe what the empirical research does and does not show, and (e) evaluate at least one reform proposal on its constitutional and empirical merits.

34.1 The constitutional framework: how money became speech

To understand the modern campaign-finance system, you need to understand a sequence of Supreme Court decisions, the most consequential of which is forty years older than most college students reading this chapter. The contemporary fight is downstream of Buckley v. Valeo (1976), and Citizens United extended a logic Buckley had already established. The choices the Court made in 1976 set the terms of every argument since.

Buckley v. Valeo (1976) — the foundational decision

In the wake of Watergate, Congress passed the Federal Election Campaign Act Amendments of 1974 — the most ambitious campaign-finance regulation in American history. The Act created the Federal Election Commission, imposed contribution limits, imposed expenditure limits on candidates and outside groups, and required disclosure. A coalition of plaintiffs led by Senator James Buckley (a New York Conservative) and Senator Eugene McCarthy (a Minnesota liberal) challenged the Act on First Amendment grounds. The case reached the Supreme Court in 1976.

The Court issued a per curiam opinion that runs nearly 300 pages and almost no one has read in full. Its central holdings have shaped every subsequent fight:

1. Money spent on political communication is speech for First Amendment purposes. Or, more carefully: money is the necessary instrument of mass political speech, because reaching a meaningful audience requires advertising, mailings, travel, staff, and venues, all of which cost money. To restrict spending is therefore to restrict the speech the spending makes possible. (This is the holding most critics of Buckley contest, and it remains the constitutional fulcrum.)

2. Contributions to candidates can be limited; expenditures by candidates and independent groups generally cannot. The Court drew a line. Contributions — money given to a campaign — can be capped because contribution limits target a "weighty governmental interest" in preventing corruption or its appearance, and because contribution limits leave the donor free to express support in other ways. Expenditures — money spent independently to advocate for or against a candidate — cannot be capped because they are core political speech. The contribution-expenditure distinction is the doctrinal hinge of the entire system.

3. The only constitutionally cognizable interest justifying campaign-finance restriction is the prevention of corruption or the appearance of corruption. Equality of political voice is not a sufficient interest. ("The concept that government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment.") This is the language Justice Anthony Kennedy would later quote in Citizens United — and the language reformers would later argue should be reconsidered.

4. Disclosure requirements are constitutional. Even where contribution and expenditure limits fail, the government can require that contributors and spenders be identified, because disclosure both deters corruption and informs voters.

5. Public financing of campaigns is constitutional, when accepted voluntarily as a condition of receiving funds.

A reader new to this material should pause on the structure of the holding. Buckley did not answer the question "should political money be regulated?" It answered "what kinds of regulations are constitutional?" The answer was: contribution limits, yes; expenditure limits, no; disclosure, yes; public financing on a voluntary basis, yes. Every subsequent Supreme Court decision in this domain has refined that framework, almost always in the direction of giving more constitutional protection to expenditures and reading the corruption interest more narrowly.

The contribution-expenditure distinction in plain English

Imagine you want to support a Senate candidate. Buckley allows the government to limit how much you can give to her campaign (a contribution), because a large direct gift creates the appearance of quid pro quo corruption — money for favors. But Buckley does not allow the government to limit how much you can spend on your own ads, mailers, or website supporting her, as long as you don't coordinate with her campaign (an expenditure). The theory: when you act independently, the candidate isn't taking your money; you're using your own money to express your views. Restricting that, the Court held, would be restricting speech itself.

Critics of the distinction (including Justice Stevens in his Citizens United dissent and many reform scholars) argue that it makes no functional sense. To a candidate or an officeholder, an independent expenditure of $5 million against her opponent is more valuable than a $3,500 contribution to her own campaign. The "independence" is a legal fiction; in practice campaigns and supportive super PACs share staff, vendors, polling, and strategic intent. The distinction, critics argue, has produced exactly the harms Buckley sought to prevent — large concentrations of money creating dependencies — while pretending they aren't dependencies because they technically aren't contributions.

Defenders of the distinction respond that the alternative — letting the government cap independent political spending — produces a worse problem. If the state can decide how much you may spend communicating your political views, it can decide who gets to be heard and how loudly. Historically, defenders note, campaign-finance laws have been used to suppress disfavored speakers: the original Tillman Act of 1907 was as much a tool against organized labor as against corporate power; the McCain-Feingold soft-money ban was repeatedly invoked to challenge issue advocacy by religious and ideological groups across the spectrum. The contribution-expenditure line is not perfect, defenders say, but the alternative is worse.

This dispute is the heart of the matter. It is unresolved, by serious people, and it will not be resolved by repeating slogans about "corporations are people" or "money is speech." The arguments need to be presented at full strength, which is what this chapter is trying to do.

Citizens United v. FEC (2010)

The case began in 2008. Citizens United, a conservative non-profit corporation, had produced a feature-length film called Hillary: The Movie, harshly critical of then-Senator Hillary Clinton (who was running for the Democratic presidential nomination). Citizens United wanted to make the film available on demand within thirty days of a primary. The Bipartisan Campaign Reform Act of 2002 (BCRA, popularly known as McCain-Feingold) prohibited corporations from funding "electioneering communications" — broadcast or cable communications that referenced a candidate within thirty days of a primary or sixty days of a general election. The FEC concluded that Hillary: The Movie was an electioneering communication, and that its corporate funding violated BCRA. Citizens United sued.

The case worked its way up. After an initial argument in March 2009, the Court took the unusual step of ordering re-argument the following September, asking the parties to brief whether Austin v. Michigan Chamber of Commerce (1990) — which had upheld a state ban on corporate independent expenditures — and the relevant portions of McConnell v. FEC (2003) should be overruled. That request signaled the direction the majority was moving.

The decision came down January 21, 2010, by a 5–4 vote. Justice Kennedy wrote the majority opinion, joined by Chief Justice Roberts and Justices Scalia, Alito, and Thomas. The majority held that:

• The First Amendment prohibits the government from restricting independent political expenditures by corporations, associations, or labor unions.
• Austin and the relevant portions of McConnell were overruled.
• The only legitimate state interest in restricting independent expenditures is the prevention of quid pro quo corruption — direct exchange of money for favors. Concerns about distortion of the political marketplace, undue influence, or aggregate concentration of speech are not sufficient.
• Disclosure requirements were upheld, 8–1 (only Justice Thomas dissented on disclosure).

Justice Stevens wrote a 90-page dissent, joined by Justices Ginsburg, Breyer, and Sotomayor. Stevens argued that:

• Corporations are not "We the People" and the historical record shows the Founders did not understand corporate political speech to be protected to the same degree as individual speech.
• The corruption interest, properly understood, includes more than direct quid pro quo. It includes the systemic distortion of representation that occurs when large concentrations of money create dependencies.
• Stare decisis required adherence to Austin and McConnell; the majority's willingness to overrule them on grounds the parties had not initially briefed reflected an "agenda-driven" methodology.
• The decision would produce "a corporate takeover of our democratic process" — language that has been quoted ever since by both critics (as prophecy) and defenders (as overheated).

Two further observations are worth making here. First, the holding of Citizens United is narrower than its political reputation suggests. The case concerned independent expenditures by corporations and unions; it did not strike down contribution limits, disclosure requirements, or the basic FEC regulatory framework. Second, the consequences of Citizens United — most notably the rise of super PACs — were a downstream effect of a separate D.C. Circuit decision, SpeechNow.org v. FEC (2010), which interpreted Citizens United's logic to mean that contribution limits to independent-expenditure-only PACs were also unconstitutional. SpeechNow is the case that actually built the super-PAC architecture. Citizens United is the case that gets the political attention.

McCutcheon v. FEC (2014)

In 2014, the Court (5–4 again, Roberts writing the plurality) struck down the aggregate contribution limits — the cap on total contributions a single donor could make across all federal candidates, parties, and committees in a two-year cycle. Individual contribution limits remained (the cap on what one donor could give to a single candidate). But the cap on the donor's total federal-cycle giving — which had been roughly $123,200 in 2014 — was eliminated. The result: a single donor could now write checks of the maximum individual amount to every congressional candidate, every state party, and every multicandidate committee on the same side, totaling potentially millions per cycle through entirely legal hard-dollar channels.

Other recent decisions to know

A handful of more recent decisions have continued the trajectory:

• Davis v. FEC (2008) struck the "Millionaire's Amendment," which had raised contribution limits for opponents of self-funded candidates.
• Arizona Free Enterprise Club v. Bennett (2011) struck a public-financing scheme that gave matching funds to publicly financed candidates whose privately financed opponents exceeded a spending threshold.
• FEC v. Cruz (2022) struck a BCRA provision limiting candidates' ability to use post-election contributions to repay personal loans they had made to their own campaigns.

The cumulative direction is consistent. The Roberts Court, by and large, reads the corruption interest narrowly and the speech interest broadly. Critics call this a constitutional revolution; defenders call it a faithful application of Buckley's long-standing framework. Both descriptions are partially true.

34.2 The campaign-finance regime as it actually exists (2026)

With the doctrinal frame in place, here is how the system works in practice. The architecture is layered, sometimes confusingly, and the layering is itself a consequence of the constitutional framework: where one channel was closed, another opened.

Hard money: contributions to candidates and parties

"Hard money" is the regulated, capped, disclosed money that flows directly to candidates and party committees. As of the 2025–26 cycle, the FEC limits are:

• Individual to candidate (per election): $3,500. Note that the *primary* and the *general* are separate elections, so the effective per-cycle individual cap to a single candidate is $7,000 ($3,500 primary + $3,500 general).
• Individual to national party committee (per year): $44,300.
• Individual to state, district, or local party committee (per year, for federal activity): $10,000 (combined limit).
• Individual to PAC (per year): $5,000.
• PAC to candidate (per election): $5,000 (multicandidate PACs).

These limits are indexed to inflation and adjusted in odd-numbered years. They are the regulated core. A reader checking an FEC filing will see most named donors well below these figures.

Political action committees (PACs)

A PAC is a formal entity registered with the FEC that raises money from individuals (or, for corporate or union PACs, from employees, members, and shareholders) and contributes to candidates. The traditional PAC, which has existed since the 1944 Smith-Connally Act made union political contributions illegal and prompted the formation of the CIO-PAC, is capped — both in what individuals can give to it and in what it can give to candidates.

PACs come in several flavors:

• Corporate PACs (formally, "separate segregated funds"): the corporation can pay administrative costs but cannot contribute corporate treasury money to the PAC. The PAC raises money from corporate executives and shareholders.
• Labor PACs: same structure for unions.
• Trade-association PACs: raise from member companies' executives.
• Non-connected PACs: ideological PACs not tied to a corporation or union.
• Leadership PACs: PACs formed by sitting members of Congress, used to give to other candidates and build influence within their party.

Super PACs

After Citizens United and SpeechNow, a new entity emerged: the independent-expenditure-only PAC, popularly called the "super PAC." A super PAC may raise unlimited amounts from individuals, corporations, labor unions, and other PACs — but it cannot contribute to candidates or coordinate with campaigns. Its money goes to independent expenditures: ads, mailers, digital, ground operations, all aimed at supporting or opposing candidates without the formal cooperation of any candidate's committee.

In practice the line between "independent" and "coordinated" has been porous. Super PACs are typically run by former campaign staff of the candidates they support, share polling and creative vendors, hold separate but parallel meetings on strategy, and increasingly post their internal media materials publicly so that the campaign's own ad-makers can adapt them. The FEC, deadlocked along partisan lines for most of the past fifteen years, has done little enforcement. We return to the coordination question below.

501(c)(4)s — the dark-money channel

A 501(c)(4) is a nonprofit "social welfare" organization under the Internal Revenue Code. It can engage in political activity as long as politics is not its "primary purpose" — a vague standard that the IRS has generally interpreted to mean less than 50% of expenditures. Critically, a 501(c)(4) is not required to disclose its donors. Money flows in anonymously; it can flow out through political ads, contributions to super PACs (where it appears as the (c)(4)'s name on the disclosure, but the underlying donors remain anonymous), or grants to other (c)(4)s.

This is the "dark money" channel. The amount of dark money in federal elections has grown substantially since Citizens United — from negligible amounts before 2010 to over a billion dollars across the 2024 cycle by some estimates. Both parties have major (c)(4) infrastructures. On the right, organizations such as Americans for Prosperity, the National Rifle Association's Institute for Legislative Action, and the Judicial Crisis Network. On the left, organizations such as the Sixteen Thirty Fund, Majority Forward, and the League of Conservation Voters. The largest single dark-money operators have been on the right historically; in recent cycles the partisan balance has narrowed, with organizations on the left raising and spending comparable amounts (Mayer; Vogel; OpenSecrets reporting through 2025).

501(c)(3)s, 527s, and joint fundraising committees

A few other entity types complete the picture:

• 501(c)(3) charities are tax-exempt and donations are tax-deductible, but they are prohibited from any "intervention in political campaigns." They can do issue education, voter education, and limited issue advocacy. They cannot endorse candidates or fund electioneering.
• 527 organizations are explicitly political organizations under section 527 of the IRC. They have broader political-spending latitude than (c)(3)s and (c)(4)s but face stronger disclosure requirements.
• Joint fundraising committees are partnerships among multiple committees (a candidate's principal campaign committee, a national party committee, several state parties, a leadership PAC) that solicit a single check from a donor and then split it according to pre-arranged ratios. Joint fundraising allows a donor to write a single very large check — sometimes hundreds of thousands of dollars — that is then distributed across many recipients each receiving an amount within the individual contribution limits. The structure is legal, increasingly common, and one reason the aggregate limits struck in McCutcheon mattered so much.
• Bundlers are individuals — often lawyers, lobbyists, or wealthy professionals — who solicit contributions from their networks and "bundle" them for delivery to a campaign. Bundlers are partially disclosed under post-2008 reforms (lobbyist-bundlers must be reported) but most non-lobbyist bundlers are not separately disclosed. Bundling is a major route to high-status access.

Coordination and the FEC

The legal prohibition on coordination between candidates and outside groups is the doctrinal seam holding the post-Citizens United system together. If coordination is real, the "independent" expenditure becomes a contribution and the contribution limits apply. The FEC has the authority to enforce this prohibition. In practice, since roughly 2010, the FEC has been deadlocked 3–3 along partisan lines on most enforcement matters, and coordination cases have rarely produced consequences. This is itself a contested question: defenders of the current system say enforcement is a matter of regulatory will and that egregious cases have been pursued; critics say the FEC's structural deadlock has effectively legalized de facto coordination.

34.3 The empirical magnitudes (2024 cycle)

How much money are we talking about? The numbers below come from FEC filings, OpenSecrets analysis, and Campaign Finance Institute data through the 2024 cycle (final post-cycle reporting available 2025).

• Total federal-election spending, 2024 cycle: approximately $15–20 billion across all sources (candidates, parties, PACs, super PACs, 527s, and disclosed (c)(4) political spending). This is the highest in modern history, in nominal and inflation-adjusted terms. The 2020 cycle was roughly $14 billion; 2016 was about $6.5 billion; 2012 was about $6.3 billion. The trend is sharply upward.
• Presidential campaigns: each major-party nominee's combined campaign-and-allied-spending operation in 2024 ran to roughly $2–3 billion or more, including super PACs and party committees on each side. This is comparable to the 2020 totals and substantially above 2016.
• Senate races: competitive Senate races in 2024 routinely cost $50–200 million each. The most expensive — Texas, Pennsylvania, Ohio, Wisconsin — in recent cycles have shattered prior records. The 2022 Pennsylvania Senate race (Fetterman v. Oz) topped $300 million combined; 2024 Ohio (Brown v. Moreno) reached comparable levels.
• House races: competitive House races typically cost $10–50+ million combined. The cost of running for an open House seat in a competitive district has roughly tripled since 2010 in real terms.
• Outside spending share: roughly 30–40% of competitive-race spending now comes from super PACs and disclosed dark-money groups, compared with under 10% before Citizens United. This is the most direct empirical signature of the post-2010 doctrinal shift.

Where does the money come from?

The donor base has shifted in two opposing directions in recent cycles. Both are real.

Small-donor revolution. Online platforms — ActBlue on the Democratic side (founded 2004), WinRed on the Republican side (launched 2019) — have democratized small-dollar fundraising. In 2024, contributions of less than $200 made up roughly 30–35% of candidate fundraising in major federal races, up from under 15% in the early 2000s. Some campaigns — including several presidential primary campaigns on both sides — raised the majority of their money from small donors. Senator Bernie Sanders's 2020 primary, Representative Marjorie Taylor Greene's House campaigns, and several recent Senate insurgent campaigns have demonstrated that small-dollar pipelines can sustain competitive operations.

Mega-donor concentration. At the top of the distribution, the share of total federal spending coming from a handful of donors has also grown. Through super PACs and (c)(4)s, individual donors writing single checks of $1 million or more accounted for an estimated 15–20% of total federal-election spending in 2024. Both parties have mega-donor classes. On the Democratic side, recent cycles have seen substantial spending by Michael Bloomberg, Tom Steyer, George Soros and his network, Reid Hoffman, Dustin Moskovitz, and others. On the Republican side, by the late Sheldon Adelson and his estate, Robert and Rebekah Mercer, Timothy Mellon, Elon Musk, Richard Uihlein, and others. Each party would prefer to discuss the other's mega-donors, but both have them, and the pattern is unmistakable on both sides.

Mid-major and "max" donors — those giving between $200 and the $3,500 max-out — make up another substantial slice, roughly 25% of candidate hard-dollar fundraising. The full cap-out group ($3,500) is a smaller share, perhaps 15%. Corporate, industry, and labor PAC giving has remained roughly stable as a share of candidate funds, around 25–30% of incumbent-member receipts depending on chamber.

The reader should hold these numbers in mind: small donors are a large and growing share of candidate committee receipts, mega-donors are a large and growing share of outside-group spending, and total spending has grown faster than either component. These can all be true at once, and they are.

Naming names: who the largest donors actually are

Generic talk about "billionaires" lets readers project whichever villain matches their politics. The honest move is to name names on both sides, paired, so the pattern is unmistakable. The following are individuals (or family operations) who in recent cycles wrote checks of $10 million or more, on a single side of the ledger, through super PACs and (c)(4)s. Numbers are approximate, drawn from FEC and OpenSecrets reporting through the 2024 cycle.

On the Democratic-aligned side: Michael Bloomberg (former NYC mayor, financial-data founder; multiple cycles, hundreds of millions across his self-funded 2020 presidential primary, his Independence USA PAC, and grants to allied groups), George Soros and his Open Society network (long-running funder of left-aligned (c)(4)s and the Democracy PAC family), Reid Hoffman (LinkedIn co-founder; major super-PAC donor since 2018), Dustin Moskovitz and Cari Tuna (Facebook co-founder; Future Forward USA Action), Tom Steyer (hedge-fund founder; NextGen Climate America and a self-funded 2020 presidential primary at $200+ million), Karla Jurvetson (physician and major individual donor on women's-empowerment and progressive infrastructure), Stephen Mandel (hedge-fund founder; Senate Majority PAC).

On the Republican-aligned side: Sheldon Adelson (casino magnate, deceased 2021; with his wife Miriam Adelson, the largest single Republican donor of the 2016 and 2020 cycles, with hundreds of millions to American Crossroads and other vehicles; Miriam continues at scale), Robert and Rebekah Mercer (hedge-fund family; major funders of the Make America Number 1 super PAC and a network of right-coded data and media ventures), Timothy Mellon (Mellon family heir; nine-figure giving to the MAGA Inc. super PAC in 2024), Elon Musk (Tesla/X founder; nine-figure giving to America PAC in 2024 and earlier cycles), Richard and Elizabeth Uihlein (Uline shipping-supply founders; long-time Republican Senate and ideological super-PAC funders), Ken Griffin (Citadel hedge-fund founder; major Republican donor through the 2022 and 2024 cycles), Jeff Yass (Susquehanna trading-firm founder; major individual donor to school-choice and Republican vehicles).

This list is not exhaustive and the precise totals shift cycle to cycle. The point is that mega-donor concentration is a feature of both parties' financial architecture, and any honest discussion of campaign-finance has to begin from that fact. There is real disagreement about whether the partisan balance of mega-donor giving is symmetric, asymmetric, and in which direction it has moved over time — credible analyses (Bonica's DIME data; OpenSecrets; CFI) suggest the partisan balance has moved closer to parity in recent cycles, with Republican-aligned mega-donor totals exceeding Democratic-aligned totals in some cycles and the reverse in others. What is not in dispute is that both parties depend on a small number of nine-figure-cycle donors at the top of their fundraising pyramids.

Small-donor pipelines: ActBlue and WinRed in detail

The small-donor revolution is not abstract. It runs through two specific platforms, each affiliated with one of the major parties, that have changed federal fundraising mechanics in a single decade. ActBlue, a 501(c)(4) launched in 2004, processes contributions to Democratic candidates, party committees, and aligned non-profits. By the 2024 cycle, ActBlue had cumulatively processed more than $13 billion across all its associated committees since launch, with a reported 90+ million unique contributions in 2024 alone. **WinRed**, the Republican-aligned counterpart, launched in 2019 and rapidly closed the structural fundraising gap; it processed several billion dollars in 2024 across Republican federal committees. Both platforms have made it possible for a candidate to raise a meaningful campaign in $25 increments from tens of thousands of donors, without ever holding a fundraiser. Both platforms have also been criticized — sometimes by their own party leadership — for fundraising practices that include aggressive recurring-donation defaults, alarmist email subject lines, and cross-promotion that bundles donations to many candidates at once. The platforms have transformed what is fundraisably possible; they have also transformed what fundraising looks like, in ways neither party is entirely comfortable with.

34.4 What the empirical research shows

This section is the place to be most careful, because the empirical questions are exactly where the political conversation is loudest and the evidence is most often misused. We try to follow the Balance Guide's discipline: state what the data shows, distinguish what is contested, and refuse to draw normative conclusions the data does not support.

Money buys access

The clearest, most replicated finding in the empirical campaign-finance literature is that money buys access to elected officials and their staff. The most cited recent evidence is a randomized field experiment by Kalla and Broockman (2016): congressional offices were significantly more likely to grant meetings to donor constituents than to non-donor constituents, with effects on the order of 3–4× higher meeting probability. Earlier observational work by Hall (2014) and others on lobbying activity points the same direction. The mechanism is mundane and human: members and staff have limited time, donors are people the office knows, and donors get scheduled.

This is a settled empirical claim. The normative interpretation is contested. Defenders of the current system point out that "access" is not "vote-buying," that constituents who self-identify by donating may be conveying more information than non-donors, and that meetings do not necessarily change votes. Critics respond that asymmetric access is itself a distortion of representation regardless of vote outcomes, because access shapes the agenda and the framing of policy.

Money does not reliably buy votes (but the picture is nuanced)

The harder claim — that members of Congress vote with their donors against the wishes of their constituents — is much weaker empirically. Decades of roll-call analysis (Snyder; Ansolabehere, de Figueiredo, and Snyder 2003; subsequent work) consistently finds that members vote with their constituency and party more than with their donors, and that PAC contributions correlate with rather than cause voting patterns. The Ansolabehere et al. paper "Why Is There So Little Money in U.S. Politics?" is the best-known statement of this finding: campaign contributions are economically small relative to what donors would pay if vote-buying were genuinely on offer.

The picture is nuanced, however. Some studies, particularly looking at narrow technical legislation or amendment-level votes (where constituency salience is low), find more substantial donor effects. Other work finds donor effects on legislative effort rather than final votes — bill introductions, committee speeches, regulatory comment letters. And work on the policy-responsiveness of Congress as a whole (most prominently Gilens and Page 2014, with subsequent debate including Branham, Soroka, and Wlezien 2017 and others) finds that bills aligned with the preferences of high-income Americans pass at substantially higher rates than bills aligned with median-voter preferences, controlling for partisan composition. Whether this differential reflects donor influence, agenda-setting, knowledge asymmetries, or the underlying composition of the donor class is contested.

A reasonable summary: money does not flip individual roll-call votes very often, but it has substantial influence on what gets onto the agenda, how bills are drafted at the technical level, what regulatory rules are written and enforced, and which candidates get the resources to compete in the first place. The empirical asymmetry in responsiveness is real, even where the mechanism through which money affects it is contested.

Citizens United — what changed and what didn't

The post-2010 effects of Citizens United turned out to be more nuanced than either side initially predicted. Some early predictions did not pan out:

• Direct corporate independent expenditures have been smaller than expected. Most large public corporations, mindful of brand risk and shareholder reaction, have not become major political spenders in their own names. Trade associations and (c)(4)s — which can mask corporate funding — have done much of the spending instead.
• The dark-money explosion has been substantial but is concentrated in a relatively small number of mega-donor operations rather than diffuse corporate spending.
• The rise of super PACs has been the most visible structural effect. The single biggest super PACs — Senate Majority PAC and Senate Leadership Fund (each chamber's party-aligned heavy spenders), the presidential-cycle vehicles, and ideological super PACs — have shifted a substantial share of campaign spending out of candidate committees and into independent operations.
• Coordination concerns have grown significantly. The effective merger of campaign-and-super-PAC strategy is widely visible in any competitive race.

What can be said with reasonable confidence: Citizens United did not by itself produce a corporate-money tsunami, but it (combined with SpeechNow) produced a structural shift toward outside groups and mega-donors in competitive federal elections. The agenda-setting and access effects of that shift are visible. Whether it has produced worse policy outcomes — a normative judgment that depends on what you think of the policies in question — is a separate question.

The agenda effect: who can run

One empirical effect is harder to quantify but very visible to political professionals. The cost of mounting a competitive federal campaign — particularly for the Senate or for a competitive House district — has risen so much that the candidate pool has narrowed. Self-funders with personal wealth, candidates with elite donor networks, and candidates who can attract national attention have a structural advantage over candidates who would need to build a fundraising operation from scratch. Both parties show this pattern. The result is a candidate class that is, on average, wealthier and more nationally networked than the median citizen — a representational effect of the campaign-finance system that operates upstream of any vote.

Legislative drafting and bureaucratic implementation

A finding that does not get the headlines but matters at least as much as roll-call analysis: large concentrations of campaign money correlate with substantive influence on the technical drafting of legislation and on the implementation of regulation by the bureaucracy. The mechanism is straightforward and almost mundane. A bill of any complexity is drafted by congressional staff working from input by lobbyists, industry experts, and policy advocates. Whose input gets in early — when the legislation is still malleable — and whose input arrives at markup, when the structure is set, materially shapes the final product. Donor-funded organizations and the lobbyists they support routinely place themselves in the early-input position; non-donors typically do not. The same pattern repeats at the regulatory stage. When an agency opens a notice-and-comment rulemaking, comment-letter quality is a function of legal and technical resources, and donor-class organizations bring those resources at scale. The result is a regulatory apparatus that is procedurally open to all but substantively responsive to those with the staff to engage. None of this is illegal; much of it is sensible policymaking; all of it advantages well-resourced actors. The campaign-finance literature increasingly studies these downstream effects rather than the up-or-down vote, because the up-or-down vote captures the smallest part of what donors get from their giving.

Smaller-than-predicted Citizens United effects, more carefully

It is worth being precise here, because the political mythology around Citizens United has run well ahead of the data. Several common claims do not survive close empirical examination. Citizens United did not flip a single Supreme Court election to corporate-funded candidates; partisan polarization at the state-judicial level had been rising before 2010 and continued at roughly the same trajectory after. Citizens United did not produce dominant corporate spending in House and Senate races; the largest super-PAC spenders are typically partisan party-aligned vehicles funded by individual mega-donors, not Fortune 500 companies. Citizens United did not, by itself, transform American policy in a corporate-favorable direction; the share of federal legislation aligned with Chamber-of-Commerce priorities is not visibly different post-2010 than pre-2010 once partisan composition of Congress is controlled for. What Citizens United (and SpeechNow) did accomplish is a structural change: a shift in where the money sits and a parallel campaign apparatus visible in every competitive race. The effects are real and worth taking seriously, but they are not the corporate takeover Justice Stevens warned about — and not the irrelevant tempest its defenders sometimes claim either. They are something in between, and getting the description right matters for the policy debate.

34.5 Steel-manning both sides

This is the section where it would be easiest to slip. We will try to give each position the strongest possible articulation, in the language each side actually uses, without translation into the other side's terms. Readers should be able to finish each subsection and feel — even if they disagree — that they understand why a thoughtful person might hold the view.

"Money in politics is a serious problem" — the strongest case

Adherents include Larry Lessig (Harvard Law), the Brennan Center for Justice, RepresentUs, Common Cause, the Campaign Legal Center, Public Citizen, and a range of scholars across the social sciences.

The argument runs roughly as follows. Representative democracy depends on responsiveness — elected officials taking the preferences of their constituents into account when they govern. The empirical evidence shows that responsiveness is asymmetrically distributed across the income distribution, with the policy preferences of the top deciles much more likely to translate into law than those of the median voter (Gilens and Page; Bartels). The mechanism is over-determined: donors get more access, donor-class issues dominate the agenda, donor-acceptable candidates get the resources to compete, donor-staffed think tanks shape the policy menu, and donors' preferences are over-represented in the technical drafting of legislation and regulation. Citizens United and SpeechNow did not create this asymmetry, but they widened it by allowing unlimited super-PAC and dark-money spending.

The harm is not primarily quid pro quo corruption — though that exists. The harm is systemic corruption: a slow distortion of the representational relationship that produces policies persistently slanted toward the preferences of those who fund campaigns, whatever their party. The public perceives this distortion: poll after poll finds large bipartisan majorities believing that political donors have too much influence, that the system is rigged in favor of the wealthy, and that meaningful reform is necessary.

The constitutional framework that produced this state of affairs is, on this view, mistaken. The Buckley equation of money with speech is a category error: money is the means to speech, but the means is not the speech itself, and other means (food, transport, equipment) are routinely regulated without being treated as speech. The corruption interest, as the Court has narrowed it, leaves no room for the most important harms. And the contribution-expenditure distinction is increasingly fictional in a world of fully integrated super-PAC operations. The structural cure — reformers argue — is some combination of public financing, contribution limits, expenditure limits, robust disclosure, and (more ambitiously) constitutional amendment to overrule Citizens United.

"The current system is constitutional and reform is dangerous" — the strongest case

Adherents include Bradley Smith (former FEC chair, founder of the Institute for Free Speech), the Cato Institute, parts of the ACLU on speech grounds (the ACLU has long opposed broad campaign-finance restrictions, though its positions have shifted somewhat), the Federalist Society, and a range of First Amendment scholars across the spectrum (Kathleen Sullivan, Lillian BeVier, Floyd Abrams, others).

The argument is, at its core, a First Amendment argument. Political speech is the speech the First Amendment most clearly protects; criticism of incumbent officeholders is its core function. Mass political speech requires money, because reaching a meaningful audience in a country of 330 million people requires advertising, mailings, travel, staff, and venues. To restrict campaign spending is to restrict the speech the spending makes possible. The state's interest in preventing quid pro quo corruption is real but narrow; the state's interest in equalizing political voice is, the Court has rightly held, no interest at all under the First Amendment. Once the state can decide that some voices are too loud and must be quieted, no principle constrains that authority. The McCain-Feingold soft-money ban was used to chill issue advocacy by religious organizations (some left-coded, some right-coded), and the original Tillman Act of 1907 was as much an instrument against organized labor as against corporate power. Campaign-finance restrictions historically have been used to suppress disfavored speakers; the bias of the regulator determines who gets suppressed.

The empirical case for reform, defenders argue, is also weaker than reformers claim. Citizens United did not produce the corporate-money flood predicted; small-donor fundraising is ascendant; the mega-donors who have emerged exist on both sides and offset each other; the access asymmetries reformers identify exist in every democratic system, regardless of campaign-finance rules. Public financing schemes, where tried, have produced more diverse candidate pools but also greater taxpayer expense and a number of unintended consequences (incumbent advantage in some designs, vulnerability to gaming in others). Disclosure regimes work to some degree but have been weaponized against donors — the AFL-CIO v. FEC and NAACP v. Alabama doctrines exist for a reason, and the modern social-media environment makes disclosure-as-harassment a real risk for many donors.

The structural cure, on this view, is the system we have, with whatever marginal adjustments fit within the existing constitutional framework: aggressive enforcement of the contribution limits that already exist, robust disclosure within the bounds of the NAACP v. Alabama tradition, voluntary public financing for candidates who choose to participate, and confidence that an open speech environment ultimately self-corrects more reliably than a regulated one.

What both sides agree on

The genuine areas of agreement are larger than the political conversation suggests:

• Disclosure is generally good. Both sides support some form of disclosure of large contributors. The disagreement is at the margins (how much disclosure of small donors, what counts as electioneering, when disclosure produces harassment).
• Quid-pro-quo corruption is bad and should be illegal. Both sides agree on the constitutional core of Buckley's anti-corruption interest.
• Voluntary public financing is constitutional. Both sides agree, since Buckley. The disagreement is whether public financing is a good idea — does it work, what does it cost, what are the unintended effects.
• Coordination prohibitions are legitimate, in principle. Both sides agree that if the law allows independent expenditures while regulating contributions, the line between independent and coordinated must be enforced. The disagreement is whether the FEC's failure to enforce reflects a regulatory problem or a fundamental design problem.
• The current system is unsatisfying to most Americans. Both sides agree on this descriptive fact. They disagree on what to do about it.

34.6 The reform menu

This section walks through the major reform proposals on the table as of 2026. For each, we present the strongest case for, the strongest case against, and the empirical evidence on what works.

Constitutional amendment to overrule Citizens United

The most ambitious reform. Versions have been introduced in Congress repeatedly since 2010 (the most prominent being the "Democracy for All" amendment family). The amendment would establish that Congress and the states have authority to regulate campaign spending and that constitutional protections of speech and the press do not apply to corporate political expenditure. Variants differ on whether they reach independent expenditures only, contributions and expenditures, or also media corporations.

Strongest case for: Would restore congressional authority to regulate campaign finance free of the Buckley / Citizens United framework. Would allow expenditure limits, corporate spending bans, and aggregate limits without ongoing constitutional challenge. Polls show majority support, including substantial Republican support among voters (though not Republican legislators).

Strongest case against: Would carve out a substantial speech exception in the First Amendment for the first time. Would make the scope of protected speech contingent on which government regulator — federal or state — was wielding the new authority, and history shows campaign-finance regulators have frequently used their authority against speakers they disfavor. The "media corporations" carveout (essentially every proposed version exempts media corporations because applying spending limits to the New York Times would clearly be unconstitutional under any reading) creates an awkward asymmetry: the New York Times Company can spend unlimited corporate dollars on political speech, while Citizens United (or its left-coded equivalent) cannot. Critics argue this codifies a preference for institutional press over civic-association speech.

Realistic prospects: Constitutional amendment requires two-thirds of both houses of Congress and three-fourths of the states. No campaign-finance amendment has come close to that threshold since 2010. As a practical matter, this is not a near-term reform path; it is a long-horizon political project.

Public financing

The most empirically tested category of reform. Several jurisdictions have substantial track records:

• New York City has run a public matching-funds program since 1988. Small donations from residents are matched at up to 8-to-1, capped, in exchange for participating candidates accepting expenditure limits. Empirical analysis (Brennan Center; CFI) shows the program has produced more diverse candidate pools, higher small-donor participation, and more competitive primaries.
• Maine runs the Maine Clean Election Act, a voluntary full-public-financing program for state legislative and gubernatorial candidates. Participation rates fluctuate but have been substantial; the program has been credited with reducing legislator dependence on lobbyist fundraising, though enforcement disputes have repeatedly returned to court.
• Connecticut has a similar program for state offices.
• Seattle runs a "democracy voucher" program, distributing $100 in vouchers to each registered voter to direct to participating candidates.
• Federal proposals have included matching-funds programs (the For the People Act variants), the "Government By the People" Act, and tax-credit approaches.

Strongest case for: Empirically tested at scale; works to broaden candidate pools and reduce dependence on large donors; constitutional under Buckley (voluntary participation conditioned on accepting limits); cost is modest in federal-budget terms; popular with the public.

Strongest case against: Cost (real, even if modest); incumbent advantages can be subtle (incumbents already have name recognition; matching small-dollar fundraising amplifies what they already have); programs can be gamed (straw donors, false residency claims) and require enforcement infrastructure; in the worst case, public financing can subsidize fringe candidates whose campaigns the public did not freely choose to fund. Conservative critics also raise an objection of conscience: taxpayers should not be forced to fund political speech they oppose.

Disclosure expansion

Multiple proposals (the DISCLOSE Act in various forms, the Honest Ads Act, state-level proposals) would extend disclosure requirements to (c)(4) political spending, online political advertising, and the actual humans behind anonymous political donations.

Strongest case for: Disclosure is the least restrictive form of regulation, the form the Court has been most willing to uphold (Citizens United upheld disclosure 8–1), and the form with the strongest empirical record. Voters can use disclosure information; reporters can investigate; opposition campaigns can respond. Disclosure does not silence anyone; it just tells the public who is speaking.

Strongest case against: Disclosure has empirical limits (large amounts of dark money still flow through (c)(4)s), can be evaded by routing money through multiple anonymous LLCs, and can be weaponized against donors whose views are unpopular. The NAACP v. Alabama line of cases recognizes that compelled disclosure of association membership can chill association rights, and the modern social-media environment makes that risk acute. A disclosure requirement that produces harassment of donors is itself a free-speech harm.

Restoring aggregate contribution limits

Would require either statutory revision (likely struck under McCutcheon) or a constitutional amendment.

Anti-coordination enforcement

Less of a reform than a call to actually use existing law. The FEC has authority to enforce the prohibition on coordination between super PACs and campaigns. As of 2026, that enforcement is essentially dormant. Reform proposals in this category include restructuring the FEC (eliminating the 3–3 deadlock by making it a five-member commission with no party majority), strengthening the coordination rules to address common workarounds (parallel digital infrastructure, shared vendors, "redboxing" of strategic instructions on candidate websites), and creating a private right of action so that opponents can sue rather than depending on FEC enforcement.

Strongest case for: The current de facto regime has hollowed out the formal structure. If contribution limits exist but coordination is unregulated, the limits are illusory.

Strongest case against: Coordination rules are vague and risk chilling First Amendment-protected speech. Aggressive enforcement can become political weaponization. The FEC's deadlock, in some readings, is a feature: it prevents regulatory overreach.

What the empirical record on reform shows

The reader should not move past this section without absorbing one important fact: the public-financing reforms above have actually been tested at scale, in real American jurisdictions, with measurable outcomes. The conversation about campaign-finance reform is often conducted as if it were purely speculative, but the empirical record is real and worth examining.

New York City's 8-to-1 match. Brennan Center analysis of the 2021 city-council cycle found that program-participating candidates raised a substantially larger share of their funds from small donors and from a more residentially diverse donor pool than non-participating candidates did. Candidates of color and women candidates disproportionately participated. The total cost to the city per cycle is on the order of $40 million — a non-trivial sum but small relative to the city's $100+ billion annual budget.

Maine's Clean Elections. Mixed evidence. Participation rates have fluctuated with the program's funding levels and with court challenges to the matching-funds component. When the program has been fully funded, participation has reached majorities in both legislative chambers. Surveyed legislators consistently report less time spent fundraising and more time spent on legislative work. Critics point to a small number of fringe candidates whose campaigns benefited from public funds.

Seattle's democracy vouchers. Implemented for the 2017 cycle and renewed since. Empirical analyses (CFI; University of Washington political science) find an expanded donor pool and a meaningful increase in lower-income participation. Critics note a relatively low voucher-redemption rate — most distributed vouchers are not used — and questions about long-term sustainability.

Connecticut's Citizens' Election Program. Operating since 2008. Studies show majority participation in legislative races and reduced contributions from registered lobbyists.

The aggregate finding from this empirical record is consistent: public financing, where well-designed, broadens the candidate pool, increases small-donor participation, and modestly reduces dependence on large donors. It does not eliminate big money — outside groups continue to spend on participating-candidate races, and the structural incentives toward outside spending remain. But the marginal effect on candidate diversity and donor base is real and replicable. Whether that marginal effect is worth the cost is a normative question on which reasonable people differ; the empirical question is increasingly settled.

34.7 Cross-national comparison

A brief comparison with peer democracies clarifies how distinctive the American system is. Most democracies of comparable size and economic development have substantially stricter campaign-finance regimes — including hard expenditure caps that would be unconstitutional in the United States under the Buckley framework.

• The United Kingdom has near-total expenditure caps. Parliamentary candidates may spend a fixed amount during the regulated campaign period (a few thousand pounds, indexed to constituency size). National party spending is capped at a few tens of millions of pounds across the entire general-election period. Television and radio advertising for parties is banned; parties receive free broadcast slots ("party election broadcasts") allocated by formula.
• Germany has spending caps, public financing of parties (proportional to vote share, with caps), strong donor disclosure, and bans on certain corporate spending categories.
• Canada has spending caps for both candidates and parties, third-party-spending limits, and partial public financing.
• Australia has disclosure requirements, partial public financing, and recent moves toward expenditure caps.
• France has full bans on certain forms of campaign advertising, expenditure caps, and a system of partial public reimbursement for candidates exceeding a vote-share threshold.

The empirical picture in those systems is mixed but instructive. The UK's strict caps produce relatively short, low-spending campaigns by US standards — but they also produce a media environment dominated by a small number of national newspapers and broadcasters, with corresponding concerns about influence asymmetry of a different kind. Germany's public financing has been credited with the political stability of the postwar period but has also been criticized for entrenching the established parties against insurgent challengers. Canada's caps are popular and seem to function reasonably well; some scholars argue they have contributed to lower polarization and shorter election seasons.

The point of the comparison is not that the US should adopt the UK system. Each country's regime fits its constitutional traditions, its media environment, and its underlying political culture. The First Amendment, as currently interpreted, makes much of the European campaign-finance toolkit unavailable in the United States. The cross-national comparison clarifies what is at stake in the Buckley framework: not just the rules within a system, but the choice of a system. American distinctiveness is not an accident; it is a constitutional commitment, and the constitutional commitment has both costs and benefits that any honest reform conversation must weigh.

It is also worth noting that the apparent simplicity of European caps is partly an artifact of shorter campaign seasons and parliamentary rather than presidential systems. A UK general election unfolds in about six weeks of regulated campaigning; a US presidential cycle runs effectively two years. Caps that work in a six-week sprint would be operationally different in a two-year marathon, even if the constitutional framework permitted them. Reform proposals that import European mechanisms have to grapple with this institutional context — a single national legislative campaign with a fixed start date and a single regulated ballot is structurally different from fifty state-administered elections feeding a single national contest with primaries that begin a year before the general. None of this is a reason to abandon comparative analysis; it is a reason to read it carefully.

34.8 What this means for the citizen

We close where every chapter in this book closes: with what to do about it.

The American campaign-finance system is one of the most documented and disclosable political systems in the world. Almost every dollar in a federal campaign is, eventually, traceable through public records: FEC filings, IRS Form 990s for non-profits, state reporting in many cases, joint-fundraising-committee disclosures, and the meticulous reporting work of organizations like OpenSecrets, the Campaign Finance Institute, the Brennan Center for Justice, the Institute for Free Speech (yes, it does data work too), and the data-and-investigative journalism arms of national news organizations. The notional opacity of "dark money" applies to the donors of (c)(4)s, but the spending of (c)(4)s on disclosed political ads is reported, and the chains of money between entities are largely traceable.

What you can do:

1. Read the disclosures for your own representative. Both her campaign committee and the major super PACs spending in her district. Look at top donors. Look at industry concentration. Look at small-dollar share. This is part of the Democracy Audit (Chapter 1) and one of the exercises at the end of this chapter.
2. Form your own view of the doctrinal question. Read the Citizens United majority and the Stevens dissent; read the Buckley opinion; read at least one serious work from each side (Bradley Smith and Larry Lessig are recommended starting points). Steel-man the position you reject. Then form a view.
3. Distinguish empirical claims from normative claims. When a candidate or commentator says "money buys elections," ask what specifically they mean and what evidence supports the claim. When they say "Citizens United is corruption," ask whether they mean corruption in the legal sense (quid pro quo) or in the structural sense (systemic distortion). The two are different, and the policy implications are different.
4. Vote on this issue. Campaign-finance reform is on the legislative agenda continuously, both in Congress and in state legislatures. Candidates take positions. Ballot measures appear regularly. If this is an issue you care about, your vote on it is one of the few direct levers you have.
5. Donate, if you choose. If your view is that political money is bad, you can refuse to give. If your view is that the asymmetry of giving is itself the problem, you can give in an attempt to balance it. If you give nothing because you believe nothing matters, the result is exactly the asymmetry the system already exhibits. Power flows to those who show up, and in the American campaign-finance system, donors are people who show up.

The system is what it is. It produces both the Bernie Sanders small-dollar revolution and the Adelson super PAC, both the Tom Steyer climate operation and the Mercer political-data network, both the most expensive elections in human history and the most elaborate transparency infrastructure of any democratic system. Whether democracy is for sale, in the American system, is not a question with a single answer. It is a question whose answer depends on which dollars you look at, which races you study, what you mean by "for sale," and what you think the alternatives are.

Reasonable Americans differ. The two case studies that follow walk through Citizens United itself end-to-end (Case Study 01) and the 2024 election as a money-in-politics laboratory (Case Study 02), so that you can see the doctrinal and empirical material in concrete form. The exercises at the end of this chapter ask you to do your own analysis of the financing of a real congressional race. The further-reading list spans both sides of the doctrinal debate. By the time you finish, you should be able to argue this question well, in either direction, with the data and the doctrine in hand.

That is what civic literacy on this topic looks like — and it is harder than the partisan version, on either side.

A final note on humility. Campaign-finance is one of those domains where the smartest scholars on each side have been wrong about predicted effects, and where confident forecasts about reform consequences should be treated with caution. The post-Citizens United world looks different from what either the majority or the dissent expected. Reform programs in Maine, NYC, Seattle, and Connecticut have taught us things we did not know in advance about what works and what does not. The next decade will teach us more — about AI-generated political ads, about cryptocurrency contributions, about the legal status of foreign-influenced (c)(4)s, about the post-2024 enforcement environment. A reader who finishes this chapter ready to argue both sides, holding the empirical record honestly, and willing to update on evidence has done what the chapter is for. Certainty would be a sign you read it wrong.