Building Permits MoM

Building Permits MoM is the monthly change in new privately owned housing units approved in permit issuing places. It leads the Monthly New Residential Construction release from the Census Bureau and HUD, the housing department. The Building Permits Survey collects the approvals each period from about 8400 offices in a universe of about 19900 places. The universe covers over 99% of private residential building, so the count captures almost all planned homes.

It counts approvals only, not starts when work begins or completions when homes are ready. It excludes public housing, manufactured homes, group quarters, transient lodging, moved buildings and units inside existing structures. Permit valuation is the value on the application and often differs from true cost, so dollar volume needs caution.

There is no boom or bust cutoff and no target band. Agencies use it as a forward gauge of planned building, with higher readings pointing to pipeline expansion and mortgage demand. Monthly moves are often irregular, so more than a single period is needed to confirm a trend.

The reference period is permits approved in the prior period, with offices contacted at the start of each period. Response is voluntary through mail, phone, fax and online channels. Preliminary national and regional estimates appear first in the monthly release in the morning Eastern Time. Revised local detail for divisions, states, metro areas, counties and places follows in a separate release.

The same release also shows starts, completions, units approved but not started and units under construction. National estimates are revised in the following release and aligned to final annual totals once a year. Seasonal factors update over several years, while local estimates stay unrevised after first release.

Moves come from structure types and Census regions, with single family homes and southern areas carrying large weight.

• Single family: It captures detached and attached houses with ground to roof walls, the largest and steadiest part.
• Two home buildings: It captures buildings with two homes, a small slice of approvals.
• Small multifamily: It captures buildings with three or four homes, also a small slice.

• Large multifamily: It captures buildings with five or more homes, the volatile swing part.
• Regions: They capture the Northeast, Midwest, South and West, with the South holding a large share.

When large multifamily approvals swing, the headline can move with them while single family moves the other way. In the same release starts can wobble on sampling doubt while permits carry no range, so both can diverge. A lone period never sets a trend.

Collect gathers reports of units, buildings and valuation from offices plus field and third party files, with checks by region. Item imputation fills a missing item from annual ratios by region and city area status when a reply gives the other items. Nonresponse estimates a missing office from its prior year figure times a factor from replying offices in the same cell. The factor divides current period units from replying places by their annual totals for the same places.

Aggregate sums place estimates to counties, states, regions and the nation by structure type, with remainders carried so totals match. Seasonal adjustment removes regular seasonal, trading day and holiday effects each period with official methods, and the national total equals its parts. Annualize multiplies adjusted monthly units by 12 to get the yearly pace implied by one period, which is neither a forecast nor a projection. Compare divides the change from the prior period by the prior period times 100 to get the monthly change.

Table tallies use a non probability count without random sampling, so no sampling error or confidence range applies. A permit is approval to build, not a start, and many approved units start periods later or never. The release tracks approved but not started units separately, and monthly noise never sets a trend.