Housing Starts Fall More Than Expected In February

New construction of US residential housing in February was considerably weaker than expected, raising more doubts about the resilience of the economy’s strength. Analysts were looking for a mild pullback to a seasonally adjusted annual rate of 1.048 million units for last month. As it turns out, that was far too optimistic. The Census Bureau reports that housing starts slumped 17% last month to an annual pace of 897,000, the lowest in more than a year. Is the weakness a danger sign for the macro trend? Or is it just a blip due to a rough winter that’s temporarily weighing on economic activity? No one really knows at this stage, although that doesn’t stop anyone from choosing a narrative that fits their economic worldview.

As for the data in hand, today’s bad news was tempered somewhat by the strength in newly issued housing permits, which perked up to a seasonally adjusted annual rate of 1.092 million—close to a post-recession high. As a result, we have an odd combination of rising permits and falling starts. The two indicators tend to track one another through time, but short-term divergences pop up at times… like now. Which measure should we focus on?

Some economists favor the idea that permits are effectively a leading indicator for starts, which implies that residential construction activity will improve in the months ahead. Given the recent strength in payrolls lately, that’s a reasonable forecast. But one reason for keeping a lid on this prediction arises from the fact that today’s soft data on housing starts follows disappointing February updates on industrial production and retail sales. It seems that there are more cockroaches than expected as we continue pulling up the rug.

The one corner of the economy that’s still holding up is the labor market, which is no trivial point. Private-sector payrolls rose 288,000 last month. That’s moderately above the rolling 12-month average (+267,000) through last month, which suggests that the labor market isn’t sharing in the stress that’s popped up in other indicators. But let’s be clear: the relatively sunny view via payrolls is critical at this point. If this gives way, the outlook will take on a distinctively darker hue.

The next clue for judging the trend in payrolls arrives on Thursday, with the weekly update on jobless claims. Recent data for this leading indicator has been volatile lately, although last week’s hefty decline suggests that there’s still a healthy degree of forward momentum for job growth. Not much is expected to change on Thursday—Econoday.com’s consensus forecast calls for marginal rise, which effectively is no change for a data set that’s subject to wild swings in the short run.

As for housing starts, the trend doesn’t look encouraging these days. Based on the year-over-year percent change, starts fell more than 3% last month vs. the year-earlier period. But permits are still rising — by nearly 8% a year through February. That’s a relatively stable pace compared with recent history, and so once again we have a reason to wonder if starts will show some strength in annual terms in the spring.


In any case, deciding what comes next is getting tougher. Depending on the indicators you favor, you can rationalize a range of possibilities at the moment. The Fed certainly will have a tough time with managing expectations in tomorrow’s policy statement and press conference.

Meantime, the optimistic case rests firmly on the idea that a harsh winter has temporarily derailed economic momentum. It wouldn’t be the first time, but for now that’s only a theory. Deciding if it’s true, or not, will have to wait until we have the first round of estimates for March numbers, which won’t start rolling in for several weeks.