The Decisions Nobody Gets to Vote On

The meeting is open to the public. The important decision is already missing from the agenda.

You can comment on the landscaping. You can ask about traffic. You can express a preference between two versions of a plan. The question of whether this is the right plan, or whether the benefits and risks should be distributed differently, has been settled somewhere else.

The invitation is genuine. So is its limit.

A great deal of power operates this way. It determines which matters become public questions, which alternatives are described as serious and which conditions are presented as unavoidable. By the time a vote is counted, the field of choice may have been narrowed considerably.

This does not make voting useless. It makes the work before and after voting indispensable.

The person who selects the menu has exercised power before anyone orders. In democratic life we need to know who has that role, who can challenge the selection and whether the people living with the consequences can add something that was left out.

An issue must become somebody’s job

Suppose thousands of workers face unpredictable schedules. Each household experiences the problem privately: a childcare arrangement fails, a second job becomes impossible, a medical appointment is postponed. People may blame an individual supervisor or themselves for being unable to cope.

For the problem to become a public issue, somebody has to recognize the pattern. Somebody must gather accounts without exaggerating them, identify the employers or rules involved, explain a remedy and persuade an institution with authority to act.

Each step requires labor. Often unpaid labor.

An established industry association can perform similar work as part of its routine operation. It can notice a proposed rule, commission an analysis, arrange meetings and stay engaged through revisions. The public may encounter the subject only when an announcement says consultation has ended.

That difference in continuity matters. Organized interests can hold a conversation with government over years. A dispersed public can be left reacting to isolated moments.

We should not romanticize disorganization as purity. If ordinary people want influence, they need institutions capable of remembering, learning, negotiating and returning. A campaign that begins from zero every time will spend much of its energy rediscovering what the other side has kept in a file.

This is one reason unions, civic associations, public-interest legal organizations and independent local journalism matter. They can turn individual experiences into a sustained capacity. They do not always do it well. Their leadership and incentives require scrutiny too. But their absence leaves a space that privately financed organizations are well equipped to occupy.

What the famous study found

Martin Gilens and Benjamin Page’s 2014 study examined 1,779 proposed federal policy changes for which surveys were available from 1981 through 2002. They compared estimated preferences at the fiftieth and ninetieth income percentiles, along with organized-interest positions, with policy outcomes over the following four years. In their joint statistical models, affluent preferences and organized business positions had substantial independent associations with outcomes; average-citizen preferences had little independent association.¹

That is serious evidence of unequal responsiveness. It is not an experiment assigning political power, a complete census of public decisions or a measurement of contemporary billionaire control. The affluent proxy was the ninetieth income percentile. The policy set and historical period matter. Correlated preferences make interpretation harder.

The finding deserves attention without being inflated into a declaration that ordinary voters have no influence whatever. Such a declaration claims more than the design can establish and encourages exactly the withdrawal that organized advantage would welcome.

A later reanalysis by J. Alexander Branham, Stuart Soroka and Christopher Wlezien found that middle-income and affluent majorities agreed in 89.6 percent of the same cases. Among 185 disagreements, affluent preferences prevailed 53 percent of the time and middle-income preferences 47 percent, a difference that was not statistically significant. Their analysis also found a clearer affluent advantage when distinguishing policy passage from blocking change, and poorer respondents fared worse.²

These results do not simply cancel each other. Getting an outcome you favor is different from having independent influence over the decision. If two groups agree, the outcome alone cannot reveal which was heard. Equally, a statistical model does not license ignoring ordinary citizens’ actual victories or the limits of the data.

We need a politics mature enough to hold those distinctions. Unequal influence is troubling even when it is incomplete. Democratic agency is valuable even when its distribution is unfair.

The door before the decision

A smaller experiment examined an earlier point in the process: access.

Joshua Kalla and David Broockman studied meeting requests sent by an advocacy organization to 191 House offices of one party in 2013. The requests randomly revealed or omitted the fact that prospective attendees were campaign donors. Senior officials were made available in 18.8 percent of the donor-disclosure condition, compared with 5.5 percent when participants were described simply as constituents.³

This randomization concerned disclosure of existing donor status, not randomly assigned donations. The outcome was access, not a vote or a policy change. The setting was specific. Within that setting, the difference is difficult to dismiss as merely a reflection of constituents having different policy preferences.

The experiment helps locate an institutional advantage: some people are more likely to reach someone senior when their financial support becomes visible.

What happens after a meeting remains an open question. A politician may listen and disagree. Staff may learn useful information. A constituent without donor status may be heard through another route. None of that makes unequal access unimportant.

A democracy cannot guarantee that every request succeeds. It can ask whether access to attention is being allocated according to a standard compatible with political equality.

Attention is a scarce resource in government. That makes its distribution a constitutional concern in the broad democratic sense, even when a particular allocation is lawful.

The power to prevent a question

There are decisions we can observe and possibilities that disappear before they become decisions. The second category is more difficult to measure. Peter Bachrach and Morton Baratz made this distinction central to their 1962 argument about the two faces of power: studying visible contests alone can miss the ability to keep grievances from becoming issues.⁴

A proposal may never be drafted because a committee chair signals hostility. An official may exclude an option they believe would provoke capital flight. A researcher may avoid a question unlikely to attract funding. A local organization may soften a demand because its largest donor employs many of its members.

These are mechanisms to investigate, not facts to assume in every case. The absence of a proposal does not prove someone suppressed it. People have limited time. Bad ideas also fail to gain attention. Institutional agendas must select among competing needs.

The practical test is to look for traces. Who proposed the option? What reasons were offered for excluding it? Were alternatives evaluated under comparable assumptions? Which participants could see the evidence? Did the institution publish its reasoning? Could a challenge reopen the question?

This is slower than declaring that everything is rigged. It is also more likely to reveal something that can be changed.

An agenda should be treated as a decision with authors, reasons and an appeal route. Public participation becomes more substantial when people can challenge the frame, rather than merely decorate an outcome chosen elsewhere.

For a major development proposal, that might mean requiring early publication of alternatives and distributional effects before contracts narrow the options. For a workplace, it might mean giving elected worker representatives the right to place scheduling or safety issues on a formal bargaining agenda. For a public agency, it might mean a clear procedure for reviewing patterns of complaints rather than treating every complaint as an isolated file.

The principle is straightforward: those affected should have a credible way to make an ignored problem someone’s responsibility.

When investment becomes an argument

A government needs employment, tax revenue and useful production. Private investors can decide where and whether to commit resources. That creates an important source of influence even before anyone makes a political contribution.

An official considering a labor or environmental rule may anticipate an investment response. Sometimes that concern is well founded. A badly designed policy can drive activity away or prevent a useful project. Sometimes the threat is overstated. A firm may have strong reasons to remain, or may be seeking concessions it does not need to proceed.

The word investment does not resolve the dispute. It tells us where the investigation should begin.

What is the business actually able to move? Over what period? What productive assets, workers, suppliers, customers or infrastructure anchor it to the location? What public support has it received? What are the costs of departure to each side? Is the proposed concession purchasing additional activity, or rewarding activity that would have occurred anyway?

Those questions protect public bargaining. A government that never asks them can become an anxious host, continually lowering its demands to keep a guest who may have nowhere better to go.

At the same time, bluffing about public leverage can hurt the people a reform is meant to help. A plant can close. A financing arrangement can fail. A town with one large employer may have genuine exposure. An honest working-class politics must take the consequences seriously without treating vulnerability as a permanent obligation to submit.

The long-term answer is to improve alternatives. Diversify the local economy. Invest in transferable skills and reliable infrastructure. Avoid dependence on one owner’s promise. Coordinate with other jurisdictions where competition for investment is being used to weaken common standards.

Public capacity changes the negotiation. The more a community can provide, understand and organize for itself, the less it must accept a private forecast as a command.

Expertise can become a gate

Technical knowledge is indispensable. A rule about electricity cannot be designed well by ignoring engineering. Financial supervision requires people who understand balance sheets. Healthcare decisions need clinical expertise.

But an expert’s authority over a technical question should not quietly become exclusive authority over the values at stake.

An engineer can estimate risks under specified conditions. The acceptable distribution of those risks involves public judgment. An economist can model responses to a tax. The weight placed on equality, security and public provision cannot be read straight off the model. A clinician can explain likely benefits and harms. A patient’s informed priorities still matter.

The difficulty is that technical and political judgments are often mixed together. Assumptions about which costs count, whose time is valuable and what alternatives are feasible can enter a model long before its conclusion is presented as objective.

We should ask experts to expose those assumptions. We should also pay for independent expertise that people without private resources can use. Telling a community it may challenge a technical report is hollow if only the project sponsor can afford someone capable of reading it.

A public-interest technical office, an adequately funded regulator or a university partnership with genuine independence can change that imbalance. Such institutions need transparent methods and room for disagreement. Expertise financed by the public should not become a priesthood protected from questions.

The aim is informed democracy. Neither rule by credential nor rule by whoever distrusts credentials most loudly will deliver it.

The comfort of total cynicism

There is an emotional attraction to believing that nothing can be changed. It releases us from the possibility of trying and losing. If every election is theatre, every official is bought and every institution is a puppet, then withdrawal can feel like superior understanding.

But a theory that explains every result in advance is difficult to test and easy to use against us. If a reform loses, the system is all-powerful. If it wins, the victory must have been permitted as a distraction. No evidence can alter the conclusion.

This is not a demanding critique of power. It is a closed room.

Real institutions contain conflicts. Businesses disagree. Officials have professional commitments as well as ambitions. Workers possess skills their employers need. Public opinion can impose costs. Rules can create opportunities their designers did not anticipate. A movement can learn, split, recover and win something partial that changes the next contest.

None of this requires sentimental faith. It requires attention to the actual conditions of a struggle.

The useful question is seldom whether ordinary people possess power in the abstract. It is which people, organized how, around what demand, facing which institution, with what alternatives and what time horizon.

That question produces work. It gives despair less room to impersonate intelligence.

A public agenda worthy of the name

A democratic agenda needs more than an open microphone. It needs pathways by which experience becomes evidence and evidence can force a response.

Imagine a city where a documented pattern of wage theft or unsafe housing can trigger an investigation without requiring each affected person to hire a lawyer. Imagine public committees required to explain why a widely supported proposal will not receive a hearing. Imagine labor representatives participating early in decisions about a major workplace reorganization, while alternatives still exist.

Each proposal raises design questions. A trigger can be abused. Hearings can become performative. Representatives can become detached from their constituents. Safeguards should be built around those risks. The existence of a possible failure is a reason to design well, not a universal veto on reform.

We should measure whether the process changes decisions, how much effort participation costs and whether the least resourced people can use it. A thousand submissions are not a democratic achievement if nobody can discover what they changed.

There is a further obligation. Institutions should make their limits visible. If a local board cannot decide a matter, it should identify who can. If a public consultation cannot reopen a signed contract, that fact should be clear before residents spend their evenings preparing arguments.

Respect begins with telling people where authority actually sits.

The worker attending the meeting after a long shift does not need to be thanked for their passion. They need a process capable of carrying their reason to a decision-maker who must answer it.

That is a modest-sounding demand with radical consequences. It challenges the habit of treating public participation as a source of legitimacy for institutions that retain the exclusive right to define what participation may accomplish.

A democracy should be able to hear an inconvenient question before the answer has become irreversible.