Housing starts plunged 9.0% but permits rose 6.3% in September according to a New Residential Construction report by the commerce department.

The Econoday Consensus estimate for starts was 1.18 million units at a seasonally adjusted annualized rate (SAAR). The consensus missed by a mile. Starts came in at 1.047 million units.

Permits rose to 1.225 million (SAAR), far ahead of the consensus estimate of 1.165 million units.

Econoday calls the report “Deceptively Solid”.


Starts are mixed but permits are up in what is a deceptively solid housing starts & permits report. Starts plunged what looks like a shocking 9.0 percent in September, to a 1.047 million annualized rate. But the drop is tied entirely to the volatile multi-family component where starts fell a massive 38 percent in the month to a 264,000 rate. The more important single-family component is up sharply in its own right, 8.1 percent higher to a 783,000 rate.

Permits for both components are up with single-family 0.4 percent higher to a 739,000 rate and with multi-family, in contrast to the big decline in starts, up 17 percent to 486,000. Together, permits are up 6.3 percent to a 1.225 million rate that far exceeds Econoday’s top estimate of 1.182.

By region, year-on-year starts are down the most in the Northeast (minus 32 percent) and the South (minus 16 percent) while permits are up the most in the Northeast (plus 13.9 percent) and West which is a focused region for the nation’s builders (plus 13.3 percent). The negative headline aside, there are more positives in this report than negatives, positives that include gains for single-family starts and permits in what are pluses for new home sales.

Housing Starts 1960 – 2016


Private Single-Family Housing Starts 1960 – 2016


Private Single-Family Housing Starts 1960 – 2016 Percent Change


Those charts add a much needed perspective to the housing report. But questions remain. The key question is whether or not the report adds or takes away from GDP, and when.

Seasonal Adjustments Suspect

Before anyone gets too excited by permits or single family starts jumping 8.1%, let’s look at quarterly data, seasonally adjusted.


That has to be one of the worst looking seasonally-adjusted charts in history.

Note the unit size. This practice of taking one month of data, then multiplying it by 12, then seasonally adjusting it, serves one purpose: to make the data look better than it really is. The practice backfires in downturns.

Persistent Spin

Reuters reports U.S. single-family housing starts surge; multi-family segment falters

  • Overall home building activity is likely to rebound in the coming months, as permits for future construction surged 6.3 percent in September to their highest level since last November. Single-family permits rose 0.4 percent last month.
  • “Permits are outpacing starts, which indicates that developers and builders are finally planning for more growth ahead. That’s great news for both the economy and the consumer,” said Jonathan Smoke, chief economist at realtor.com in Washington.

Notice the persistent spin. It’s similar but even more hyped-up than Econoday spin.

This note about permits is mostly hooey. Builders start homes when they get a contract. Yes, they need a permit to start, but they won’t start without a contract. This is not a 2006 build on spec environment.

Half the equation is missing. With traffic low, are buyers going to suddenly show up? If so, why?

And for all the spin on permits, single-family permits only rose 0.4 percent. It seems there is a bias to play permits and starts both ways, ignoring anything bad.

Mike “Mish” Shedlock