Housing Starts Unexpected Dive 8.2%: Economists Blame the Weather

Housing starts dove 8.2% in Dec., well under the consensus estimate and also lower than the lowest economist's estimate.

The Census Department released its monthly New Residential Construction report for December this morning. Here are the pertinent details.

Housing Starts

  • Privately-owned housing starts in December were at a seasonally adjusted annual rate of 1,192,000. This is 8.2 percent percent below the revised November estimate of 1,299,000 and is 6.0 percent below the December 2016 rate of 1,268,000.
  • Single-family housing starts in December were at a rate of 836,000; this is 11.8 percent below the revised November figure of 948,000.
  • The December rate for units in buildings with five units or more was 352,000. An estimated 1,202,100 housing units were started in 2017. This is 2.4 percent above the 2016 figure of 1,173,800.

Building Permits

  • Privately-owned housing units authorized by building permits in December were at a seasonally adjusted annual rate of 1,302,000. This is 0.1 percent below the revised November rate of 1,303,000, but is 2.8 percent above the December 2016 rate of 1,266,000.
  • Single-family authorizations in December were at a rate of 881,000; this is 1.8 percent above the revised November figure of 865,000. Authorizations of units in buildings with five units or more were at a rate of 382,000 in December.
  • An estimated 1,263,400 housing units were authorized by building permits in 2017. This is 4.7 percent above the 2016 figure of 1,206,600.

Economists' Estimates

The Econoday consensus estimate for starts for December was 1.280 million units at a seasonally-adjusted annualized rate (SAAR). Instead starts fell to 1.192 million SAAR, well below the lowest estimate of 1.230 million.

Quick Retreat

  • The housing starts number was substantially lower than analysts had anticipated. Those polled by Econoday had made predictions ranging from 1,230,000 to 1,320,000 units, with a consensus of 1,280,000.
  • The performance in the single-family sector was significantly worse. Those starts dropped 11.8 percent compared to November, to a rate of 836,000 units although they still retained a 3.5 percent edge over the previous December. The estimate for November starts was revised from 930,000 to 948,000. Starts in buildings with five or more units increased from 343,000 to 352,000, a 2.6 percent gain, and were 21.6 percent lower than the same month in 2016
  • On a non-adjusted basis, there were 80,300 housing units started in December compared to 98,100 in November. Single-family starts dropped from 69,300 to 54,300.
  • The year-end estimate for all of 2017 is a total of 1,202,100 housing starts. This is an increase of 2.4 percent compared to the prior year.

By Region

Single-Family Starts

  • Total: -1.8%
  • Northeast: - 24.2%
  • Midwest: -8.5%
  • South: -16.6%
  • West: +0.0%

"Very Solid Report"

Let's return to Econoday for the laugh of the day.

Here is Econoday's opening gambit.

"A surprising but perhaps one-time drop in single-family starts masks what is otherwise a very solid housing starts and permits report for December. ... But the backlog behind future starts continues to build as permits came in very strong, virtually steady at a 1.302 million rate and showing a noticeable 1.8 percent gain for single-family permits to 881,000."

That is easily disproved nonsense.

Starts vs Permits

Starts vs. Permits Synopsis

  • Permits are not a leading indicator of anything.
  • If there are sales, homebuilders will start homes.
  • If there are not sales, homebuilders will not start construction no matter how many permits they have.
  • Although permits are required to do a start, permits do not represent a "backlog behind future starts".

The Weather

Econoday made this claim: "Starts can be affected by weather which along with related adjustments are always factors for this reading in the winter months."

It's amusing how economists never seem to know what the weather "was" until economic reports come out a month later.

Mike "Mish" Shedlock

Comments (4)
No. 1-4

Record high prices, rising mortgage rates and now caps on mortgage interest and property tax deductions obviously had nothing to do with it.


The North East - which gets hit the hardest the tax reform caps on mortgage interest and SALT deductions - took by far the largest hit of any region.


city and county government is a bit overwhelmed and eager to get everything out of their permit fee structure that is reasonably possible. where the weather is nice the slowdown has more to do with local government. and the economy is not good enough to support these prices


Weather here in the pacific Northwest is a major factor. Nothing happens for like 3 months of cold winter. So yes in the real world it does matter. However in fantasy I am sure it is meaningless.