Vacancy Surveys and Counting Empty Buildings

Why a simple count is not simple

“How many empty buildings are there?” sounds like a question with one answer. In practice, the answer depends on what counts as empty, what counts as a building, who is doing the counting and what evidence they use. Two careful counters can examine the same neighborhood in the same month and report different totals, and neither need be wrong.

Vacancy numbers matter because they shape decisions: where to direct inspection effort, which properties might be considered for reuse, how to describe a downtown or a neighborhood to the public, and whether a market is loosening or tightening. A number that appears precise but rests on a loose definition can mislead those decisions.

Defining vacancy

The first step is to settle on terms. Some common distinctions:

• Vacant lot or vacant building. A lot with no structure is vacant land. A lot with a structure that no one uses is a vacant building. Surveys that merge the two can inflate a count or confuse its meaning.
• Unit or building. An apartment building with 30 units might have 3 empty. Is the building vacant? Are 3 units? Counting units gives a rate of vacancy within the housing stock, while counting whole buildings flags structures with no use at all.
• Residential or commercial. Housing vacancy and storefront vacancy behave differently and are measured differently.
• For sale or rent, held off the market, or abandoned. An apartment between tenants, a seasonal home, a property in an estate dispute and a building left to decay are all “unoccupied,” but they call for very different responses.
• Short-term versus long-term. A unit empty for a few weeks between renters is part of normal turnover. One empty for several years is of more concern. A survey that gives no time dimension cannot distinguish them.

A written definition, ideally stating the minimum duration of emptiness used, is the most important part of any count.

Common ways to count

Windshield or walking surveys. Surveyors go block by block and record whether each property looks occupied. Signs of vacancy include boarded or broken windows, overgrown yards, accumulated mail or flyers, no curtains, no utility meter activity visible, and for storefronts, empty display windows or “for lease” signs. This method sees what is physically there. But it relies on judgment, can misjudge a quiet but occupied home, and cannot see inside buildings.

Postal data. In some countries, postal services report addresses from which mail has gone undelivered for a period. Such data covers large areas cheaply and is updated regularly. Its weakness is that it reflects delivery status, not the building’s actual condition, and the rules and availability differ by place.

Utility records. Water, electric or gas accounts with no usage for an extended time may point to vacancy. This can be precise for units but is not always shared with outside groups, and a unit may be empty yet billed, or occupied by someone using no service in their own name.

Census or survey estimates. Statistical surveys of households estimate vacancy rates for housing units across areas. They are well documented and consistent over time, but they are based on samples, so small areas carry larger margins of error, and they report housing units rather than the condition of individual buildings.

Local government files. Code enforcement records, vacant building registries, tax delinquency lists and demolition records each capture a slice. A registry, for example, only includes properties whose owners registered or whom the city has identified. Properties heading toward tax foreclosure often appear as delinquent long before they look empty, and those lists give a different picture than a street survey.

Resident reports. People who live nearby often know which buildings are really empty. Their knowledge is valuable but is uneven and hard to compare across areas without a consistent process.

A small hypothetical example

Suppose a single block has 40 buildings. A street survey in March finds six that look empty: boarded or neglected. That is 6 of 40, or 15 percent. A second counter using utility records for the same block finds ten with no active electric accounts over the past year: 10 of 40, or 25 percent.

Why the gap? Possibly four buildings have no electric service but look occupied (perhaps they are used for storage or occupied by someone who is not on a utility account, or have an account in an unexpected name). Or maybe two of the six boarded buildings are actually in use for other purposes. Or the street survey missed buildings that look fine from the front but are empty. The figures here are invented for illustration and not drawn from any place, but the pattern is common: different sources capture different signals, and the overlap is partial.

A good report would show both numbers, explain the methods, and say which buildings appear on both lists, which appear on only one, and why.

Why numbers differ between counters

• Different definitions of vacancy, duration and building type.
• Different dates. Properties change status quickly, particularly where sales, renovations and demolitions are happening.
• Different geography. Boundaries of the study area may not match.
• Different evidence. Visual cues, records and reports each have blind spots.
• Different inclusion rules. Whether to count partially vacant buildings, upper floors over occupied shops, garages or accessory buildings varies.
• Observer judgment. Without training and written guidance, surveyors vary in how strongly they interpret a sign.

Making a count more reliable

1. Write the definitions and the evidence rules down, with photographs of examples.
2. Train surveyors together, then have two of them independently survey the same sample and compare.
3. Combine at least two sources and record how each building was classified.
4. Use categories beyond yes and no, such as “apparently occupied,” “apparently vacant,” “uncertain” and “under renovation.”
5. Date everything, and repeat the count on a regular schedule so that changes can be seen.
6. Verify borderline cases, for instance by checking with the owner, neighbors or records.
7. Report limits openly, including the share of uncertain cases.

Reading vacancy figures with care

A high vacancy figure does not by itself explain cause, nor does a low one prove health. Some vacancy is normal in an active market; units must be empty sometimes for turnover and repair. A figure for a commercial strip such as downtown vacancy also depends heavily on whether upper floors and small spaces are counted. When comparing places or years, check that the definitions and sources match. When they do not, differences may reflect method more than reality.

A well-documented count is a starting point. It tells a city where to look and what to ask, and it is most helpful when paired with a field check of individual properties and with the knowledge of people who live and work nearby.