Photo by Gayatri Malhotra on Unsplash
WORKING PAPER NO. OPM-2026-001 Visibility & Fundraising Correlates Unit Classification: Unclassified // For Distribution to Anyone Still Reading Platforms
TO: Members of Congress, Their Staff, and the Consultants Who Actually Read This
FROM: Dr. Beatrix Solano-Fairweather, Director, Visibility & Fundraising Correlates Unit
RE: The Third Model: How Persuasion Politics Displaced Representation in America
DATE: [current reporting period]
This Office was chartered to measure the correlation between a representative’s public visibility and a representative’s fundraising yield, on the reasonable assumption that this would be a small, tidy dataset suitable for a quarterly memo. It has not been a small, tidy dataset. It has been, instead, the entire modern job description.
Political theory offers two classical accounts of what a representative owes a constituency: the delegate, who executes an expressed mandate, and the trustee, who exercises independent judgment on the constituency’s behalf. Neither model, as this working paper documents at length, anticipated a third activity — persuasion, meaning effort spent shaping what constituents want rather than either executing or judging on their behalf — becoming the dominant use of a representative’s clock. This Office regrets to report that our visibility-to-fundraising correlation coefficient is strong enough to explain most of it, and that the explanation gets considerably less comfortable in Section IV, where persuasion stops being a scheduling choice and starts being what authority does once it has already acted and needs consent to catch up.
We publish the full assessment below without abridgment, because abridging it would itself be an act of persuasion, and this Office would like, for one document, to not be the joke it is measuring.
Democratic theory has never fully settled what a representative is for, but for most of its history the argument has run between two poles.
The delegate model holds that a representative is an agent of a constituency’s expressed will — sent to enact instructions, not to originate them. The representative’s judgment is subordinate to the mandate; discretion is a failure mode, not a feature. This is the logic behind binding platforms, recall elections, and instructed convention delegates.
The trustee model, canonically stated by Edmund Burke in his 1774 address to the electors of Bristol, holds the opposite: a representative owes constituents not obedience but judgment. Burke’s claim was blunt — a representative who “sacrifices” independent judgment to public opinion “betrays” rather than serves the constituency, because the representative has time, information, and deliberative distance the public does not. The trustee is elected precisely because direct, unmediated public preference is a poor substitute for informed decision applied to specific and shifting circumstances.
Madison’s design for the American republic leaned trustee. The point of representation, in Federalist 10, was to “refine and enlarge the public views by passing them through the medium of a chosen body of citizens,” on the theory that the refined judgment of elected intermediaries would better serve the public interest than the public’s raw, unmediated preferences — particularly preferences vulnerable to faction and passion.
Both models share an assumption worth naming: that the representative’s time is a scarce resource being spent on one of two activities — receiving and executing instruction, or exercising independent judgment on the public’s behalf. Whichever model you prefer, the work is recognizably legislative: drafting, negotiating, evaluating, deciding.
Neither model fully anticipated persuasion becoming the dominant use of a representative’s time. Burke assumed representatives would need to explain and defend their judgment to constituents; Madison’s refining function presupposes some degree of public address. The claim here is not that political theory overlooked persuasion — rhetoric has always been part of representation — but that neither model treated persuasion as capable of displacing delegation and trusteeship as the primary occupation of the office itself.
Define the term precisely, since it carries most of the essay’s weight: persuasion is representative effort directed at altering public opinion rather than altering law — shaping the electorate’s disposition rather than converting an existing disposition into legislative outcome. Campaigning, fundraising appeals, branding, and media performance fall inside this category. Constituent casework and coalition-building around specific statutory text do not, even though both involve communication. The line is what the effort is for.
It is also worth flagging that this makes persuasion a different kind of category than its two counterparts. Delegate and trustee are normative theories of where legitimate authority should reside — accounts of how a representative ought to act. Persuasion, as used here, is not a competing normative ideal. It is a behavioral description of what representatives are increasingly incentivized to do, regardless of which normative theory they claim to follow. Call the resulting pattern the persuasion model, with that caveat built in. It is not trusteeship, because the representative is not primarily applying independent judgment to legislative substance — floor speeches and cable appearances are optimized for resonance, not for the hard, often boring work of building coalitions around statutory text. It is not delegation either, because there is frequently no clear antecedent mandate being executed — the representative is not carrying out a platform so much as producing public opinion in real time, then citing that opinion as license.
The persuasion model is parasitic on the legitimacy of the other two. It borrows the delegate’s rhetoric (“the people want”) without the delegate’s actual mechanism of accountability — a specific, votable platform against which performance can be measured. It borrows the trustee’s rhetoric of judgment and expertise without trusteeship’s obligation to actually decide things — persuasion can be perpetually deferred, since there is always another news cycle to shape opinion for, and no moment at which the representative is required to convert opinion into legislative outcome.
The apparent exception — the populist claim to speak for “the people” against an establishment — is worth pausing on, because it looks at first like a genuine mandate. It isn’t. A real mandate is prospective: a platform is settled before the representative acts, and constrains the action that follows. The populist invocation is retrospective and self-authored: the representative first shapes the opinion, then cites that same opinion, after the fact, as license for whatever was already decided. It functions less as an instruction the representative is bound by than as a shield deployed after the fact — a manufactured mandate, produced and consumed by the same actor, rather than a prior constraint imposed from outside.
This is not a claim that persuasion is new to politics — rhetoric has always been part of representation. The claim is narrower: that persuasion has shifted from being an instrument used in service of legislative or delegative goals to being the dominant allocation of a representative’s time, because the electoral market rewards it more reliably and more immediately than either of the other two activities does.
A representative’s hours are finite and split, roughly, between two competing uses: legislative craft (bill text, committee work, negotiation, oversight, budget detail) and public-facing activity (media appearances, fundraising, digital messaging, constituent-facing signaling). These are not complements drawing on different resources; they are substitutes drawing on the same clock.
The account that follows assumes, as most institutional analyses of Congress do, that representatives are predominantly reelection-seeking actors — not exclusively, but reliably enough to shape the modal allocation of effort. The causal chain is short. Representatives want reelection. Reelection increasingly depends on visibility rather than legislative output. Visibility is produced by persuasion, not by craft. Effort reallocates accordingly. Three structural features make that reallocation close to inevitable rather than merely likely:
Visibility asymmetry. Legislative craft is slow, technical, and largely invisible to the public and to primary electorates. A markup session or an appropriations rider does not travel; a viral clip does. The electoral market prices visibility, not legislative competence, and the two are only loosely correlated.
Primary-election exposure. In a polarized, low-turnout primary system, the median voter a representative must satisfy to survive is not the median voter of the general electorate but the median voter of an intense, ideologically sorted primary base — a base for which cultural signaling and identity performance are often more legible and more mobilizing than legislative output.
Fundraising structure. Small-dollar donor bases and national donor networks respond to affective, identity-coded messaging far more reliably than to legislative substance. A representative’s fundraising email about a culture-war flashpoint outperforms one about a subcommittee markup, which means the fundraising incentive and the persuasion incentive point the same direction.
None of this requires bad faith on the part of individual legislators. It is a description of an incentive gradient that a rational, career-motivated actor would follow regardless of personal conviction — which is precisely why the pattern is systemic rather than a matter of individual character, and why exhortations to “just legislate more” have had limited effect.
Section III explains persuasion as an efficient use of a representative’s time under electoral incentive — a rational response to what pays. But that account still treats persuasion as optional: a representative could, in principle, spend the marginal hour on legislative craft instead, at some cost to reelection odds. There is a stronger version of the claim, in which persuasion is not optional at all but structurally required — a mechanism institutions reach for once the ordinary channels of legitimacy have already been bypassed.
Delegation and trusteeship share a temporal structure that is easy to overlook because it is so basic: both locate legitimacy before the exercise of authority. A delegate acts because a prior mandate authorized the action. A trustee acts because a visible, accountable process of judgment preceded the decision and can be evaluated against outcomes afterward. In both cases, legitimacy is a precondition of the act, not a consequence claimed for it after the act has already occurred.
A great deal of contemporary governance no longer works this way. Executive orders issued to bypass a gridlocked legislature, emergency and crisis powers invoked under time pressure, agency rulemaking that fills the space Congress declined to occupy, judicial reinterpretation of long-standing statutes — these share a structural feature. The action occurs, or is understood by the actors involved to be a foregone conclusion, before anything resembling a mandate has been secured, and often without any visible deliberative process that a trustee-style defense could point to. Authority has, in the relevant sense, already been spent before the question of its democratic warrant has been settled.
This is the condition under which persuasion stops being a discretionary use of time and becomes a legitimizing necessity. The action cannot be justified as delegation, because there was no antecedent instruction to point to. It cannot be justified as trusteeship, because trusteeship requires a visible exercise of judgment the public can inspect and hold accountable, and reactive, crisis-driven, or administratively insulated decisions rarely offer that. What remains is the third option: change the public’s interpretation of the action after the fact, so that consent appears to have preceded a decision that in fact preceded consent. The population is not asked to authorize the action prospectively. It is managed into recognizing the action as having been theirs all along.
This is a sharper version of the manufactured mandate described in Section II. There, persuasion was one representative’s rhetorical strategy — an individual borrowing legitimacy language without its substance. Here, persuasion becomes the default mode by which a reactive institutional order sustains its own legitimacy at all, because the alternative — conceding that the action was never authorized in either the delegate or the trustee sense — would corrode the authority of the office itself. The population’s interpretation has to be brought into line with the decision, because the decision was never going to wait for the population’s interpretation to form first. Authority exercised this way does not seek agreement in order to act. It acts, and then requires agreement in order to keep acting.
Read this way, the persuasion model is not merely a description of idle hours between votes. It is what authority does once it has outrun its own sources of legitimacy and needs the appearance of consent to keep functioning. This is the deeper answer to the essay’s opening question about political education: politicians increasingly cast themselves as guides and teachers of the population not because civic instruction is the job, but because manufacturing retroactive agreement has become the job, in the specific cases where authority was never going to wait for agreement to precede it.
Consider, as a test case rather than a proposal, a party organized around the “responsible party government” model American political science briefly took seriously — most explicitly in the APSA’s 1950 report, and closer to how party systems with binding manifestos and disciplined voting actually operate in the UK, Germany, and elsewhere. Persuasion and deliberation occur upstream, concentrated at platform formation — conference votes, primaries, membership referenda. Once elected, the representative converts from open-ended trustee or freelance persuader into something closer to a pure delegate: an agent authorized to execute a platform that was itself the product of a prior, bounded, comparatively transparent deliberative process.
This does not eliminate persuasion. It relocates it, from continuous and diffuse to periodic and bounded.
It is worth being honest about why this is a test case and not a live option. Parliamentary party discipline works because it has a constitutional home: the executive is drawn from and answerable to the legislative majority, so a platform can be enforced through confidence votes and fused executive-legislative incentives. The American system has neither lever. Separation of powers, staggered terms, decentralized primaries, and independently financed campaigns mean no party center can bind a member the way a parliamentary whip can. The U.S. Constitution was designed, in part, to prevent exactly this kind of concentrated party control — so importing the model would mean rebuilding the incentive structure the Constitution deliberately fragmented, not merely adopting a platform.
The gain this promises is legibility. “Did the party deliver on the platform it was elected to enact?” is a question with a determinate answer. “Did my representative exercise good judgment on my behalf across hundreds of votes I did not follow?” is not — it collapses, in practice, into a referendum on identity and affect, which is exactly the terrain persuasion politics is optimized to compete on.
The platform-party model is not a free repair, and its distortions are structural rather than incidental.
Discretion relocates rather than disappears. Platforms are drafted under one set of facts and must be executed under another, often years later, under circumstances the platform’s authors did not anticipate. Events eventually outrun the mandate’s specificity. When they do, someone still exercises judgment about what the mandate implies. If the elected representative has no authorized discretion, that judgment migrates to whoever is left holding it: party leadership, bureaucratic implementers, or courts asked to interpret statutory intent. None of these actors is more accountable to voters than an elected representative was. The trustee function is not eliminated by a platform system. It is displaced downward and rendered less visible — a worse outcome for accountability, not a better one.
Platform authorship becomes the real site of power — and a narrower one. Whoever controls the platform-drafting process — activist factions, party apparatus, an intensely engaged primary electorate — exercises outsized influence relative to the diffuse general electorate the representative nominally serves. Internal party life becomes the substantive politics, and internal party life is more, not less, capturable by an organized, intense minority than a general election is. Persuasion-driven representatives responsive to donors and primary bases would simply give way to platforms drafted by whoever controls party conference. That is a substitution, not a solution.
Legislative flexibility narrows. A delegate authorized only to execute a fixed platform has little room to logroll, trade, or compromise across party lines without appearing to betray the mandate that elected them. This is the substance of Lord Hailsham’s “elective dictatorship” critique of the British system: majority-plus-manifesto-plus-party-discipline produces minimal deliberation once in office, because deliberation was supposed to have already happened, upstream, at the manifesto stage. Whether this is a bug or a feature depends on how much one trusts the platform-formation process to have anticipated the actual governing problem — which is precisely the discretion question restated.
The persuasion model’s most corrosive long-run effect may not be on legislation directly but on the public’s model of its own civic role. Section IV described this in its sharpest form: a population trained to backfill consent for decisions already made. What follows is the slower, more diffuse version of the same pattern.
If representatives spend their time persuading rather than either executing mandates or visibly exercising judgment on discrete legislative questions, citizens are trained into a spectator relationship with governance. The job of “being informed” comes to mean tracking a representative’s messaging — their media performances, their cultural stances, their fights — rather than tracking legislative outcomes against any stable standard, whether a platform or a record of judgment. Political scientists Christopher Achen and Larry Bartels have argued, in Democracy for Realists, that voters already reason primarily through group identity and retrospective, low-information heuristics rather than policy substance. A persuasion-dominant system does not create this tendency. But it entrenches and monetizes it, since identity-coded messaging is the persuasion model’s most efficient product.
Over time this produces a public that is civically active — attentive, opinionated, emotionally invested — while being civically disempowered in the specific sense that matters. It has few reliable tools for holding representatives accountable to anything more concrete than vibes and tribal loyalty. Engagement becomes intense and spiky, mobilized around cultural flashpoints, elections, and media moments, rather than steady and substantive tracking of legislative text, oversight outcomes, and budget detail. The public comes to expect political leaders to function as moral and cultural instructors — the “guidance” function named at the outset — not because that is what representation requires, but because that is the only function persuasion-oriented representatives are, in practice, spending their time performing. Expectations adapt to supply. The population is taught to wait for its opinions to be shaped, because shaping opinion is the one thing it can reliably watch its representatives do.
A platform-party system, for all its own distortions, at least relocates the public’s civic labor to a moment — platform formation — where engagement can translate directly into outcome. Its risk is converting citizenship into a once-per-cycle referendum on brand loyalty rather than an ongoing audit of an agent’s judgment. The persuasion model’s risk is worse: it offers no moment at which public engagement reliably converts into anything, because there is no fixed standard — mandate or judgment record — against which “delivery” can be measured at all. What looks like a uniquely engaged public may, on this view, be a public that has been handed an unusually large amount of political stimulation in exchange for an unusually small amount of political leverage.
The delegate/trustee debate has always been an argument about where in the system judgment should legitimately live — with the electorate directly, or with an elected agent trusted to exercise it. Persuasion politics does not resolve that argument; it evades it, by locating enormous representative effort in a third activity — opinion formation — that is accountable to neither a mandate nor a visible standard of judgment.
The sharper version of the original question, then, is not “should representatives execute or judge,” but: where should persuasion legitimately live in a democratic system, and what happens when it lives everywhere, diffusely, with no clear endpoint or accountability for whether it was conducted honestly — and worse, when it is called on not merely to win future consent but to manufacture consent for what has already been decided? Right now, in the American case, persuasion lives in the daily operation of nearly every elected office, competes directly with the hours that might otherwise go to delegation or trusteeship, and increasingly does the retroactive work of legitimizing authority that outran its own warrant before anyone was asked. Concentrating persuasion at a bounded moment — platform formation — would not fix that second failure on its own. But it would at least restore the correct sequence: agreement before action, rather than action followed by the manufacture of agreement to match it. That sequencing, more than any specific institutional design, is the thing worth defending.
Dr. Beatrix Solano-Fairweather
Director, Visibility & Fundraising Correlates Unit
Office of Persuasion Metrics
This Office measures persuasion. It does not, on current staffing, practice it — a distinction we ask readers to take on faith, which is, we recognize, exactly the kind of request this paper is about.