BY Table of Contents
Democracy has two filters.
The first is suffrage: legal eligibility to vote. Who is allowed to cast a ballot. The franchise.
The second is turnout: who actually exercises that eligibility on election day. The people who show up.
The effective electorate is what you get when you multiply the two. Eligible voters times actual participation. The slice of the citizen body that walks through the door. That is the group politicians respond to. Not “the people.” Not the eligible population. The slice that shows up.
Almost all modern democratic reform energy targets the first filter. Almost none of it targets the second.
That is not an accident, and it is not neutral. The second filter is doing more of the work in determining outcomes. The people who benefit from the current second-filter design are the same people who write the reform agenda.
The First Filter: Who Is Allowed to Vote
Suffrage debates are easy to identify. The standard reform menu looks like this.
Voter ID requirements: does showing ID at the polls suppress turnout among eligible voters, or does it protect the integrity of the franchise? First filter.
Automatic voter registration: should eligible citizens be registered by default, or should registration require an affirmative act? First filter.
Felony re-enfranchisement: should people with criminal records have their voting rights restored after completing their sentences? First filter.
Non-citizen voting in local elections: should legal permanent residents vote in school board and municipal elections? First filter.
Age 16 voting proposals: should the franchise extend to 16- and 17-year-olds? First filter.
All of these are real debates. The companion entry on Qualified Democracy takes the first filter seriously and argues it has a right answer: the franchise should track a genuine, durable stake in the polity, not mere presence or legal residency. This piece is not going to re-run that argument.
The structural point is simple: every reform on that list changes who is legally eligible. None of them changes the conditions under which eligible voters actually participate.
The Second Filter: Who Actually Votes
The turnout picture in the United States is a cliff.
Presidential elections: roughly 66-67% of the eligible voting population.
Midterms: roughly 50%.
Municipal elections: below 15% in most jurisdictions.
School board elections: 5-10% in many places.
Off-cycle primaries: often below 10%.
Every step down the ballot is a step toward an electorate that looks less like the eligible population.
Off-cycle electorates skew older, wealthier, and more heavily organized than the eligible population they are drawn from. The people voting in school board elections are not a cross-section of the parents, taxpayers, and residents whose lives depend on those schools. They are whoever showed up: older homeowners, public-sector employees with a direct financial stake in the outcome, organized groups that operate at full strength whether turnout is 7% or 70%.
At 7% turnout, the organized group does not have to be large. It just has to show up.
The companion entry on Off-Cycle Elections works through the mechanics in detail. The key point here is that the second filter is not random. It is structural. And the structure was, in many cases, deliberately designed.
What First-Filter Reforms Actually Produce
Take the 26th Amendment. Ratified in 1971, it lowered the voting age from 21 to 18. The argument was Vietnam: old enough to fight, old enough to vote.
The prediction was that enfranchising 18- to 20-year-olds would drive a youth-powered electoral politics. It did not. The 26th Amendment added approximately 11 million new eligible voters to the rolls. The effective electorate did not shift to match that expansion, because most of the newly eligible did not show up.
The pattern holds across first-filter expansions. Automatic voter registration bumps registration rates. It does not correspondingly bump turnout. The eligible pool grows; the effective electorate grows less.
The first filter is real. It matters. A franchise that excludes whole categories of citizens by race, class, or arbitrary status is indefensible. That point is not in dispute.
But the second filter absorbs most of what the first filter produces. You can expand the eligible pool at the margins and see almost no change in actual outcomes if the second filter stays broken. The effective electorate will still be the subset that can be reliably mobilized, and the most reliably mobilizable groups are not the ones that suffrage expansions are designed to include.
What the Second Filter Is Actually Producing
The Baltimore experiment.
Baltimore moved its municipal elections to align with the presidential-year cycle in 2016. Turnout went from 13% to 60%. The same eligible electorate. A different effective electorate. Different politics.
California passed SB 415 in 2015, requiring cities and special districts with chronically low off-cycle turnout to consolidate their elections onto even-year November ballots. The results tracked what the research predicted: higher turnout, a broader effective electorate, measurable shifts in the policies those local governments produced.
The academic spine here is Sarah Anzia’s Timing and Turnout: How Off-Cycle Elections Favor Organized Groups (Cambridge University Press, 2014). Anzia is not a polemicist. She is a political scientist with a data set. Her finding is unambiguous: off-cycle election timing systematically advantages organized groups, most prominently public-sector unions in school board and municipal elections.
The effect is not marginal. The structure does the work.
If you want to understand why teachers’ union contracts look the way they do in cities that spend heavily on education and still produce weak outcomes, start there. The answer is not corruption in any simple sense. The answer is who votes in school board elections and when.
The Progressive-era reformers who introduced off-cycle municipal elections understood the logic. Some of them preferred what off-cycle elections produced: a small, manageable electorate of civic-minded property owners, insulated from immigrant working-class politics at the state and federal level. The civic-hygiene rhetoric sold the reform. The class interest drove it.
The pattern persists. Public-sector unions, real estate interests, and incumbent political machines benefit from a small, predictable electorate. They have every reason to keep elections off-cycle and off the radar of the broad eligible population.
Why Reform Energy Goes to the Wrong Filter
Here is the diagnosis.
The people who win under the current second-filter design are the same people who staff the institutions that produce democratic reform proposals: public-sector unions, the credentialed professional class, university-trained policy advocates, foundations funded by the gentry. They are not, as a group, ignorant of how election timing works. Anzia’s book has been out since 2014. It is not difficult to read.
They direct reform energy toward the first filter because changes to the first filter do not threaten the coalitions that benefit from the second-filter design.
Expanding the franchise to felony voters or 16-year-olds produces real political debate. It signals democratic commitment. It does not change who controls school board elections in a jurisdiction with 8% off-cycle turnout. The organized groups that run those elections at 8% will still run them at 9% after re-enfranchisement expands the eligible pool.
Consolidating elections onto the on-cycle calendar is a different kind of threat. It brings 60% participation to contests that organized rent-seekers have been winning at 13%. That relocates the effective electorate from the reliable mobilization base of the professional-managerial coalition to something closer to the eligible population at large.
The first filter is the safe surface to argue about.
The second filter is the live wire.
This is not a conspiracy. It does not require bad faith from reformers who focus on the first filter. It is a structural alignment of interests, and structural alignments do not need to be coordinated to produce consistent outcomes. The gentry-class professional who advocates for 16-year-old voting rights has genuinely convinced herself that the franchise expansion is the more important question. The institutional incentives that produced that conviction are real. A professional-class ecosystem that rewards first-filter advocacy produces no career incentive to threaten the second-filter design.
Reform energy gets redirected. The diagnostic work that would disturb the right people never gets done at scale.
A Note on Gerrymandering
The two filters produce an effective electorate. Once that electorate exists, its votes have to be translated into representation. Gerrymandering is where that translation gets distorted: district lines drawn to produce predictable outcomes for incumbents and organized interests regardless of how the broad electorate votes.
Gerrymandering is not a third filter. It is a downstream distortion. All three are part of the same architecture. (See the companion entry on Gerrymandering for the full treatment.) Fix both filters and you substantially reduce gerrymandering’s leverage, because the gerrymanderer’s power depends on a predictable and limited electorate. A broad, on-cycle effective electorate is much harder to carve into reliable blocs than a narrow, off-cycle, organizationally captured one.
What Both Filters Working Looks Like
The affirmative case is the more important one.
Get the first filter right, as the Qualified Democracy entry argues, and the eligible electorate represents people who hold a genuine stake in the polity: citizens with demonstrated commitment to the place, old enough to have contact with consequential decisions, contributing to the fiscal community they are voting to govern. Not a perfect electorate. A more serious one.
Get the second filter right, as the Off-Cycle Elections entry argues, and the effective electorate tracks the eligible electorate rather than the slice that organized rent-seekers can mobilize. On-cycle consolidated elections bring the broad population of genuine stakeholders into the room for the decisions that most directly affect their lives: the school board, the county commission, the state assembly primary.
Two filters working in sequence produce something that has not been the norm in American local governance for most of a century: politics that can actually be answered to.
The school board answers to the parents and neighbors whose children will be in those classrooms in ten years.
The city council answers to the taxpayers who will be paying the pension obligations it votes to extend.
The state assembly primary answers to the registered party members who will live under the policy agenda the nominee pursues.
This is not romance. It is structure. The man who owns a house on the block, has children in the district school, and has been paying city taxes for fifteen years is not politically alienated by nature. He is politically alienated by a second-filter design that makes his participation in the elections most relevant to his life nearly invisible and structurally inconsequential. The firefighters’ union that mobilizes four hundred votes and decides the county commission seat is not acting in bad faith. It is acting in its interest. The question is whether the design gives anyone else an equal shot.
Fix the structure. And he will show up.
Final Thoughts
For the scale layer of the same argument, see the companion entry on Jefferson’s Hundreds.
For the upstream question of whether a polity has the substrate to sustain self-government in the first place, see the companion entry on The National Question.
If you read one book on this, read Sarah Anzia’s Timing and Turnout: How Off-Cycle Elections Favor Organized Groups (Cambridge University Press, 2014). It is short, precise, and empirically damning. Anzia is not a partisan. Her data is.
For the structural critique of why unlimited democracy produces the second-filter conditions she documents, Hans-Hermann Hoppe’s Democracy: The God That Failed (Mises Institute, 2001) is the necessary companion. Hoppe argues that universal suffrage, unconstrained by any stake qualification, generates high-time-preference governance because the people making decisions do not bear the full costs and will not be around when the bill arrives. Anzia shows you what that looks like at the local level. Hoppe explains why it is not a bug.
Pair both with the companion entries: Qualified Democracy for the first filter, Off-Cycle Elections for the second, and Gerrymandering for the downstream translation problem. The three entries form a coherent architecture.