The Infill Strategy

The economics are stark. A 2013 study by the Halifax Regional Municipality found that low-density suburban development cost the city $3,462 per capita annually in infrastructure and services, while urban core neighborhoods cost only $1,416 — less than half. The difference compounds over decades. Every new subdivision requires new water mains, new sewer lines, new roads, new bus routes, new schools — while existing infrastructure in older neighborhoods sits underutilized as populations decline.

Edmonton itself has hundreds of hectares of vacant or underused land within its mature neighborhoods. The City's 2020 Infill Roadmap identified over 12,000 potential infill sites — lots where buildings once stood, parking lots that serve dead shopping centers, industrial land rezoned but never redeveloped. Meanwhile, the city continues to extend services to new suburbs at Windermere, Glenridding, and The Orchards, each requiring twenty-five-year infrastructure financing that future residents will inherit.

Alexander understood this pattern of urban holes. In *City Country Fingers* (3), he argued that cities should not sprawl evenly but should grow in concentrated bands with green fingers between them. The infill strategy honors this by growing inward rather than outward — filling the gaps before extending the fingers. It recognizes that a neighborhood with vacant lots is not complete; it is waiting. The empty corner lot where a grocery once stood, the gravel parking lot behind the hardware store, the fire-damaged house that was demolished and never rebuilt — these are not neutral spaces. They are wounds in the urban fabric, places where eyes have left the street, where snow accumulates uncleared, where weeds grow through cracked asphalt.

Portland, Oregon, has pursued infill systematically since the 1990s, adopting design standards in 2008 that govern how new buildings on vacant lots must relate to their neighbors — matching setbacks, limiting height differences, requiring ground-floor windows. The result is visible: a skinny house on a half-lot between two older homes, a small apartment building on what was a used-car lot, a row of townhouses where a warehouse burned. The neighborhood density increased by 12% between 2000 and 2015 without the demolition of existing homes, and the city avoided $1.1 billion in infrastructure costs it would have spent extending services to equivalent greenfield development.

The pattern also supports Incremental Development (100) by keeping the scale small. One lot at a time, one building at a time, by different builders over years. The thirty-foot-wide vacant lot gets a narrow house. The corner parking lot becomes four townhomes. The old motel site becomes a courtyard apartment building. No single project transforms the neighborhood overnight; each one fills a gap and the neighborhood becomes, over time, complete.