Target population: All manufactured home shipments in the United States

Sampling frame: A listing of new HUD inspected manufactured home sections shipped each month provided by the Institute for Building Technology and Safety (IBTS). Sections shipped to Canadian dealers or to FEMA only dealers are removed from the frame prior to sampling. 

Sampling unit: Individual section of a manufactured home.

Sample design: A systematic sample of fixed size is taken from each of the four Census regions without replacement for a total of 405 homes per month. Occasionally the sample size will be slightly less if duplicates (sections from the same home) are selected. FEMA units are included in the sample with certainty for estimation purposes but no data is collected for these units through dealer contact.

Frequency of sample redesign: The sample was last redesigned from a 1 in 40 systematic sample to its current design in Jan 2010.  There is no set schedule for sample redesign for this survey.

Sample maintenance: There are no sample maintenance procedures since the sample is selected from a new frame each month.

Data items requested and reference period covered: The survey form can be found here. The reference period is for a home shipped four months prior.

Key data items: Sales price (or intended sales price) and status (sold/placed for residential use, intended for sale for residential use, placed for non-residential use, other (lost, destroyed, FEMA, etc)).

Type of request: Voluntary

Frequency and mode of contact:  Dealers receive a mailed form four months after a sampled unit is shipped to them.  Dealers can respond by mail, fax, or telephone.  Calls are made to dealers who fail to respond.

Data collection unit: Data is collected each month from a sample of about 405 manufactured home shipments.

Special procedures: There are no special data collection procedures used for this survey.

Editing: Respondent data is reviewed for consistency across related items.

Nonresponse: Nonresponse occurs when survey participants choose not to take part or response data are unobtainable for other reasons. To account for nonresponse, a respondent may have their initial sampling weights adjusted based on characteristics such as activity status and/or response status. Economic surveys also perform imputation, which is the procedure for determining a value for a specific data item where the response is missing or unusable.

Nonresponse adjustment and imputation: Missing data items such as status code, final destination, leased status, length, width, and bedrooms are imputed using hot-deck imputation monthly.  In hot-deck imputation records with reported data for a particular item are used as donors for the record with the missing item. Donors are identified within pre-defined imputation cells based on known characteristics of the home such as dealer state/region and size. Each donor is only used once and if a donor is unavailable the mode or average value of the item is used.  Sales price is imputed by fitting a regression model on square feet and size (single/multi section) of home. 

Other macro-level adjustments:  Weights are adjusted so that estimates of total shipments equal the published totals of single and multi units shipped each month.

Tabulation unit: Manufactured Housing Unit (containing 1 or more sections)

Estimation: Estimates of manufactured housing shipments by status are calculated using the sample weights and an adjustment to the population control totals of homes shipped by single or multi section. Estimates of average sales price include both actual sales prices and intended sales prices.

Seasonal adjustment:  Seasonal adjustment is the estimation of the seasonal component and, when applicable, also trading day and moving holiday effects, followed by their removal from the time series. The goal is usually to produce series whose movements are easier to analyze over consecutive time intervals and to compare to the movements of other series in order to detect co-movements.

The seasonally adjusted series are shown as seasonally adjusted annual rates (SAAR). A SAAR is the seasonally adjusted monthly estimate multiplied by 12.

The seasonal adjustment indexes were developed using X-13ARIMA-SEATS software.

The X-13ARIMA-SEATS program provides summary statistics to indicate the overall effect of the seasonal adjustment. This table of summary measures shows some of these statistics. For more information on X-13ARIMA-SEATS see the reference manuals posted on the Census Bureau's website.

An assumption underlying the seasonal adjustment process is that the original series can be separated into a seasonal component, a trading-day component, a trend-cycle component, and an irregular component. The seasonally adjusted series consists of the trend-cycle and irregular components taken together. The trend-cycle component includes the long-term trend and the business cycle. The irregular component is made up of residual variations, such as the sudden impact of political events and the effects of strikes, unusual weather conditions, reporting and sampling errors, etc.  For more information on Seasonal Adjustment, view our Seasonal Adjustment Questions and Answers here.

Seasonal indexes are developed concurrently for each month for U.S. total shipments.

Disclosure avoidance:  Disclosure avoidance is the process for protecting the confidentiality of data. The Census Bureau uses two methods of preventing disclosure of business data, cell suppression and noise infusion.

Cell suppression protects the confidentiality of individual businesses by replacing cell values with symbols in tables, where the amount of the cell if it were known, would allow one to estimate a single contributor’s value too closely. This occurs when there are very few contributors, or when there are one or two large contributors that dominate the aggregate statistic.

The cells that must be protected are called primary suppressions.

To make sure the primary suppressions cannot be closely estimated by subtracting the other cells in the table from the higher-level totals, additional cells may also be suppressed. These additional suppressed cells are called complementary suppressions.

The process of suppression does not usually change the higher-level totals. Values for cells that are not suppressed remain unchanged. Before the Census Bureau releases data, computer programs and analysts check published tables for both primary and complementary disclosures.

Only cell suppression (primary and complementary) is applied to estimates for the Manufactured Housing Survey.

The Census Bureau has reviewed this data product for unauthorized disclosure of confidential information and has approved the disclosure avoidance practices applied. (Approval ID: CBDRB-FY22-278)

The sampling error of an estimate based on a sample survey is the difference between the estimate and the result that would be obtained from a complete census conducted under the same survey conditions. This error occurs because characteristics differ among sampling units in the population and only a subset of the population is measured in a sample survey. The particular sample used in this survey is one of a large number of samples of the same size that could have been selected using the same design. Because each unit in the sampling frame had a known probability of being selected into the sample, it was possible to estimate the sampling variability of the survey estimates. The Census Bureau recommends that individuals using these estimates incorporate this information into their analyses, as sampling error could affect the conclusions drawn from the estimates.

The Standard Error measure indicates the extent to which a survey estimate is likely to deviate from the true population and is expressed as a number. The Relative Standard Error (RSE) is the standard error expressed as a fraction of the estimate and is usually displayed as a percentage: RSE = 100 x SE/estimate.

A relative standard error (RSE) is provided for each survey estimate.  The 12-month average reliability of the key estimates for 2021 are displayed in the table below:

The nonsampling error of an estimate based on a sample survey encompasses all factors other than sampling error that contribute to the total error of the estimate. This error may also be present in censuses and may be attributed to many sources: inability to obtain information on all units in the sample; response errors; differences in the interpretation of the questions; mistakes in coding or keying the data obtained; and other errors of collection, response, coverage, and processing. Although no direct measurement of the potential biases due to nonsampling error was obtained, precautionary steps were taken in all phases of the collection, processing, and tabulation of the data in an effort to minimize their influence. The Census Bureau recommends that individuals using these estimates incorporate this information into their analyses, as nonsampling error could affect the conclusions drawn from the estimates.

A potential source of nonsampling error in the estimates is nonresponse. Nonresponse is the inability to obtain all the intended measurements or responses about all selected units. Unit nonresponse is used to describe the inability to obtain any of the substantive measurements about a sampled unit. For the 2021 survey, the average unit response rate was 73.0% and the total quantity response rate for price was 59.8%.

The methodology for collecting information on new manufactured homes for 1974 through 1979 involved contacting a sample of manufactured home dealers each month within 137 geographic areas or primary sampling units. The dealers were asked to provide data on the number of manufactured homes received from manufacturers, the number placed on a site for residential use, and the number held in inventory.

The methodology used after 1979 involved a monthly sample of new manufactured homes shipped by manufacturers. The dealer to whom the sampled unit was shipped was contacted by telephone and asked about the status of the unit. This was done each month until that unit is reported as placed.  Data prior to 2014 can be found on our previous data page here.  This data is not comparable to the current data series noted below.

The methodology used beginning in August 2014 involves contacting the dealer four months after the unit was shipped to ask about the status of the unit. The dealer is asked to report a sales price if the unit is already sold and placed for residential use or to report an intended sales price if the unit is intended for sale and for residential use. The dealer is no longer contacted each month until the unit is placed. Estimates of average sales price include both actual sales prices and intended sales prices.  In January 2021, the survey was modified to identify leased units and to better distinguish units that were placed or not yet placed at time of survey.  In addition, changes to select characteristics questions were modified to align with current industry standards.   These modifications allow for more detailed characteristics and did not alter the survey methodology. Annual characteristics data prior to 2021 are based on the old survey form found here.