The Gardens of Grey Nax

Filtering Theory and Affordability Mechanisms

8 min read

Filtering Theory and Affordability Mechanisms

Concept note — engaging the mainstream housing supply argument honestly, and comparing how different affordability mechanisms work. The vault’s most important opposition engagement.

The Supply-Side Argument

Filtering theory is the most common objection to non-market housing: “just build more market-rate housing and affordability will take care of itself.” The argument goes:

New market-rate construction attracts higher-income households. Those households vacate their previous units, which become available to moderate-income households. Those households vacate their previous units, which become available to lower-income households. Over time, the housing stock “filters down” as it ages and depreciates, becoming more affordable at each step. More supply at the top eventually produces affordability at the bottom.

This is the intellectual backbone of YIMBY (Yes In My Backyard) advocacy and supply-side housing policy. Its strongest academic proponent is Edward Glaeser (Triumph of the City, 2011), who argues that density, deregulation, and permissive zoning are the primary tools for housing affordability.

Where Filtering Works

The vault cannot credibly dismiss filtering entirely. The evidence supports it at certain scales:

Regional supply elasticity. At the metropolitan level, cities that build more housing generally have lower price growth than cities that don’t. Tokyo, which permits significantly more housing construction than most American cities, has lower real rent growth than housing-constrained markets like San Francisco or Boston. The aggregate relationship between supply and price is real.

Long-term price moderation. Over multi-decade timescales, new construction does moderate prices in aggregate. The housing that was market-rate in 1960 is often naturally affordable in 2025 — not because of policy but because of age and depreciation. Filtering is a real phenomenon in the long run.

Vacancy chain effects. Research (notably Mast 2021, Pennington 2021) has documented measurable vacancy chain effects from new market-rate construction — new luxury units do produce downstream moves that free up units at lower price points, even in the short term.

Granting these points is not a concession. It’s a precondition for being taken seriously. Any argument that begins “markets don’t work” will be dismissed by the people the project needs to persuade. The argument needs to begin “markets work for X but fail for Y, and here’s why Y matters.”

Where Filtering Fails

Neighborhood-level displacement. Filtering operates at aggregate scale. At the neighborhood level, the mechanism often runs in reverse: new market-rate construction signals neighborhood desirability, attracts investment, and accelerates gentrification. The aggregate regional effect (more supply → lower prices) can coexist with the local effect (new construction → displacement of existing residents). A family displaced from their neighborhood by rising rents is not helped by the fact that a cheaper unit exists forty-five minutes away by bus. Housing is place-specific.

Time horizon mismatch. Filtering takes decades. A family that needs affordable housing now cannot wait thirty years for today’s luxury apartments to depreciate into affordability. The time horizon of filtering is measured in housing generations; the time horizon of housing need is measured in months. This mismatch is fatal for the claim that supply alone solves affordability.

Non-substitutability of community. Filtering assumes housing units are interchangeable — a two-bedroom apartment in one neighborhood is a substitute for a two-bedroom apartment in another. But housing is not just shelter. It’s social infrastructure. The relationships, the school connections, the mutual aid networks, the familiarity — these are non-substitutable. When a family is displaced from a functioning community to a cheaper unit elsewhere, they lose capabilities (in Sen’s sense) that the cheaper rent cannot replace. See Social Infrastructure, Incidental Contact.

Filtering doesn’t work in reverse for land. The supply argument applies to buildings, which depreciate. Land doesn’t depreciate — it appreciates, driven by community investment and scarcity. Building more housing doesn’t reduce land costs in desirable locations. This is Henry George’s insight (see Henry George — Progress and Poverty, Land Value Capture): the land question is categorically different from the building question, and filtering theory doesn’t distinguish between them.

Upward filtering. In hot markets, filtering can operate upward: older housing stock is renovated, repositioned, and repriced at market rate. The naturally affordable housing that filtering theory counts on disappears through rehabilitation, demolition, and conversion. This is particularly acute in cities like Durham where older housing stock is being purchased by investors and upgraded for higher rents.

The Affordability Mechanism Comparison

Different approaches to housing affordability operate on different logics. Understanding the mechanism clarifies what each can and cannot do:

Market-rate supply (filtering). Mechanism: build enough that aggregate prices moderate. Strength: politically achievable, leverages private capital, no ongoing public cost. Weakness: slow, geographically imprecise, doesn’t protect existing communities, doesn’t work for land. Affordability is a byproduct, not a guarantee.

Housing vouchers (Section 8 / Housing Choice Vouchers). Mechanism: subsidize demand by paying the gap between what a household can afford and what the market charges. Strength: portable, lets residents choose location, doesn’t require new construction. Weakness: depends on continuous appropriation (funding can be cut), inflates rents in supply-constrained markets (landlords capture the subsidy), relies on landlord willingness to accept vouchers (many don’t), and affordability vanishes the moment the voucher does.

Low Income Housing Tax Credit (LIHTC). Mechanism: federal tax credits incentivize private developers to build affordable units. The dominant production program for affordable housing in the U.S. Strength: produces actual units, leverages private investment, has built millions of affordable apartments since 1986. Weakness: affordability is temporary — compliance periods run 15–30 years, after which properties can convert to market-rate. The affordable units LIHTC built in the 1990s are beginning to expire now. LIHTC also depends on investor appetite for tax credits, which fluctuates with tax policy. And the subsidy flows primarily to developers and investors, not to residents.

Inclusionary zoning. Mechanism: require or incentivize developers to include affordable units within market-rate projects. Strength: produces affordable units without direct public expenditure, integrates income levels within buildings. Weakness: typically captures only 10–20% of units, affordability may be time-limited, can be traded for in-lieu fees that fund affordable housing elsewhere (breaking the integration promise), and developers resist or avoid jurisdictions with strong requirements.

Public housing. Mechanism: government builds and operates housing directly. Strength: permanent affordability, no expiration date, direct public control. Weakness: in the U.S., severely underfunded, stigmatized, and often paternally managed in ways that undermine residents’ capabilities (see Nussbaum — Creating Capabilities — bodily integrity, practical reason, and control over environment are all compromised by conventional public housing management). The model works well in Vienna and Singapore; it has been deliberately starved in the U.S.

Community Land Trust. Mechanism: remove land from the market permanently; residents own improvements on leased land with capped resale. Strength: affordability is permanent (not time-limited), residents build modest equity, democratic governance, no ongoing subsidy required after capitalization, resistant to market cycles. Weakness: requires initial capitalization (land acquisition + development), operates at small scale (individual projects, not citywide), and the limited equity model builds less wealth than market homeownership (which is the point, but it’s a real tradeoff for individual residents).

The Vault’s Position

The vault’s argument is not “filtering doesn’t work.” It’s: “filtering addresses the aggregate supply question but cannot address the community question — and the community question is what matters for the specific kind of flourishing Wellspring is building.”

More precisely:

  1. Supply-side solutions and non-market solutions are not in opposition. Durham needs more market-rate housing and permanently affordable non-market housing. Filtering helps at the metro level; CLTs help at the community level. These are complementary strategies, not competitors.

  2. The relevant unit of analysis matters. If you evaluate affordability at the metropolitan level, filtering looks adequate. If you evaluate it at the neighborhood level — where people actually live, form relationships, build mutual aid networks, and send their kids to school — filtering is insufficient. Wellspring operates at neighborhood scale.

  3. Permanence is a design feature, not a luxury. LIHTC expires. Vouchers can be cut. Inclusionary zoning units can be bought out. The CLT ground lease is permanent. The question isn’t “what makes housing affordable today?” — it’s “what keeps housing affordable in 2055?” Only structural removal of land from the market answers that question.

  4. Affordability without capabilities is insufficient. A cheap apartment in a car-dependent suburb far from jobs, transit, and social infrastructure is “affordable” by the numbers but fails the capabilities test. Wellspring’s claim isn’t just “cheaper” — it’s “cheaper and embedded in social infrastructure that expands what residents can do and be.” See Sen — Development as Freedom, Nussbaum — Creating Capabilities.

Graph View

Backlinks