Be Careful Interpreting the Housing Data from the State Redistricting Files

Print This Post Print This Post

Written by: Director Robert Groves

I’ve read a couple of articles expressing shock at the level of vacant housing units in the redistricting files from the 2010 Census. I’ve also read a blog from one user who thought the totals for vacant units from our Housing Vacancy Survey compared to the 2010 Census were too high for the four states released thus far. To avoid misinterpretation of what these the counts mean, we should make clear a number of points.

We delivered counts of total housing units and counts of vacant housing units. However, the total vacant count includes units intended for rent or for sale, units intended for seasonal, recreational or occasional use, units for migratory workers (a very small number), and other vacant units that did not fit into these categories (usually held off the market for personal reasons of the owner). Seasonal, recreational, or occasional vacant houses were prevalent in April, 2010, especially in resort areas where there are cottages, condominia, and homes that are used only during the given tourist season.

Using the delivered total vacancy counts that include seasonal, recreational or occasional use vacant units can mislead a user who wants to measure the impact of the housing crisis in such an area. On the other hand, a large number of this type of vacant units in areas where you would expect higher occupancy rates at the time of the census or areas where there may have been high expectations of second home purchases may also indicate a housing problem. Later reports from the 2010 Census will show the vacancy status for all vacant housing units and will also show whether occupied units were owned or rented.

I would also note that we collect data on vacant units from several other surveys, and you will find that there will be differences, sometimes noticeable differences, between the results provided by the 2010 Census and those from these other surveys. There are number of reasons why these numbers may differ. For example, we attempt to measure the occupancy status of units on April 1, 2010 – a single day. Most surveys that supply vacancy rates measure the status of sample units at the time the field representative conducts the interview. Census enumerators returned to units thought to be vacant over several months to verify the status, but always attempted to measure status as of April 1, 2010. We will be actively investigating these differences as we evaluate the 2010 Census results.

This entry was posted in Redistricting. Bookmark the permalink.

6 Responses to Be Careful Interpreting the Housing Data from the State Redistricting Files

  1. Bruce Fogarty says:

    The 2010 Census Data Release schedule does not indicate when the mail return rates will be reported. It should. Our Complete Count Committee was surprised by the mail participation rates in areas in and adjacent to our military bases. We’re waiting to see if the return rates differ “substantially” from the participation rates and reviewing these data with our base commanders. It may suggest a problem in confusion over where to report, a surprising number of vacant housing units, or something else. In any case, our After Action Report is eagerly awaiting this information. We know the Bureau doesn’t expect much difference between these two rates, given what USPS was asked to do in 2010. We’ll see.

  2. Sataris says:

    I’m curious as to how long it takes to actually accumulate the data for such an overview?

  3. Patty Becker says:

    We need counts of “year-round” housing units which eliminate those intended for seasonal use only. If occupied and vacant HU counts are available for the year-round stock only, we can calculate vacancy rates which makes sense.

  4. RIchard Barnes, Ph.D. says:

    I would like to know if Vacant Housing appears in the TIGER file and whether this can be displayed using Google maps…
    Thank you.

  5. Our census block shapefiles with 2010 census population and housing unit counts ( do not include vacant housing numbers. Vacant housing data would need to be downloaded from American FactFinder2 and joined with the shapefile (
    Our TIGER/Line Shapefiles cannot be added to Google maps. You would need GIS software to view our TIGER/Line Shapefiles properly.

  6. CRM says:

    What are the implications for the vacant housing numbers?

Leave a Reply

Your email address will not be published. Required fields are marked *