Key housing challenges – including occupancy in the rural Southwest, manufactured housing’s potential to help solve the affordability crisis and the accurate provision of critical city resources – are all being addressed in some form with help from American Community Survey (ACS) data. But while the data is invaluable to researchers and policymakers, it requires very careful analysis.

L to R: Under Secretary of Commerce Mark Doms, Director of Minnesotans for the American Community Survey Joan Naymark and Zillow Senior Economist Aaron Terrazas speak on a panel at the recent American Community Survey Data Users Conference. (photo credit: U.S. Census Bureau)

Zillow recently participated in the Second Annual American Community Survey Data Users Conference organized by the Census Bureau, the Population Reference Bureau and Sabre Systems. The ACS is an annual survey of Americans conducted by the Census Bureau. It provides essential information about the American population, and Zillow often combines ACS data with our own data to analyze trends and compare housing markets nationwide.

A number of ACS users – from academic researchers, to nonprofits and state and local governments – spoke about how they use ACS data, and a number of presentations touched on housing issues.

Here are some of the highlights:

Hidden Communities: The Geography of Manufactured Home Parks in the United States

Lance George of the Housing Assistance Council, a nonprofit group, spoke about his research on manufactured homes. Manufactured homes – including mobile homes or trailers – represent a small part of the U.S. housing stock, but can be an important source of affordable housing in rural communities.

Manufactured homes lie somewhere in between traditional perceptions of owner- and renter-occupied homes, because residents typically own the structure but rent the land on which it sits. The cost of moving these homes from site to site typically ranges from $5,000 to $10,000, making mobile home owners less mobile than traditional renters. Manufactured homes are also often financed through high-cost or high-interest-rate loans, leading to greater incidence of financial distress among owners. However, foreclosures happen much quicker than traditional homes since manufactured homes can be repossessed on relatively short notice – similar to other vehicles.

As the nation confronts the growing challenges of rental affordability, most of the conversation has focused on urban markets and multifamily properties. But as George highlighted, affordability is also a concern in many smaller communities, where large multifamily developments may not make sense, but manufactured homes could offer a solution.

ACS: A Look Forward, Data Content, Dissemination and Controls

Joseph Salvo of the New York City Planning Department spoke about the challenges of using ACS data for very small geographies. He analyzed the total number of housing units, occupied housing units, the vacancy rate and total population of the Rockaways neighborhood in Queens around the time of Hurricane Sandy in October 2012.

The analysis showed how each year of ACS data for these smaller geographies can have substantial statistical noise from year to year. It is difficult to see a clear change in the housing stock in the Rockaways before and after Hurricane Sandy, despite an obvious change according to anecdotal reports.

The way the data are processed and weighted by the Census Bureau might explain the lack of conclusive evidence. For city governments, this is a challenge. For instance, city officials trying to plan the number of childcare facilities needed in different communities, and different weighting of the ACS data, can lead to a substantial under- or over-supply of services in different communities.

How Does the ACS Compare to Local Utility Data for Understanding Local Housing Occupancy

Nevada State Demographer Jeff Hardcastle spoke about his research comparing ACS data with local energy utility data to estimate housing occupancy rates.

Housing occupancy data are particularly important for Nevada lawmakers, since they are used to allocate state funds to local governments. Hardcastle combines public record data, permit data, Census data and, in this analysis, energy utility data to estimate occupancy rates – which is particularly challenging in Nevada’s small communities outside the major cities of Las Vegas and Reno. In rural parts of the state, the boom-and-bust mining economy has led to dramatic population swings in small housing markets.

Each data source has strengths and weaknesses. Census data includes housing units that are likely overlooked in permit data – for instance, in-law suites or temporary rentals – but are unavailable for small geographies on a high frequency. When available, utility data can help address some of these concerns.

Slides from the presentations can be accessed by registering with the ACS Data Users Group here.