Home construction estimates are notoriously variable. They can change oftentimes by a substantial amount on a month-by-month basis, and the numbers are just as frequently subject to serious revisions in the short run. Still, we have to note when there are big moves even in these high frequency series, though keeping in mind no definitive conclusions are available from them alone.

Housing starts plummeted in June 2018, at least according to the initial Census Bureau figures. From a high of 1.337 million (SAAR), starts are thought to have been just 1.173 million in the latest data. Total permits had fallen sharply in May and were down again last month.

These declines are the responsibility of both sides of the ledger, single family as well as multi-family construction. As to the former, it isn’t yet all that much but it may suggest some cooling as mortgage rates are being digested and/or the economy not living up to its mainstream billing.

More than that, however, apartment construction was where most of the recent declines were found. This reverses a partial trend reversal from earlier this year. Multi-family construction had been contracting since the middle of 2015 right alongside weakening labor incomes (opposing the unemployment rate). Beginning in October and November 2017, permit filings rose slightly while starts jumped to meet them.

It’s difficult to assess how much of a delayed effect there might have been due to the hurricanes Harvey and Irma. The increase in activity could have reflected expectations for mortgage rates, too. If borrowing costs rise making home buying slightly less affordable, it would stand to reason that rental units would be slightly more in demand.

With these latest and sharp declines, however, the trend in multi-family looks more like it did in the middle of last year. Combined with at least short-term weakness in single family, the net result was not a good one in June.

The big problem simply remains above all of it. Americans are still stuck with cohabitation. Household formation was tripped up again in the first quarter of 2018, as reported in separate estimates also provided by the Census Bureau. The number of American households was slightly less in March 2018 than in December 2017 despite population growth.

That puts potential home unit demand 3.7 million behind the prior precrisis baseline, a significant gap that no matter how much time passes and how low the unemployment drops it just does not get better. Akin to the labor market’s participation problem, almost certainly related to it, it may be that home builders on both sides have a household problem that they occasionally have to account for.

In other words, perhaps last year’s small bump in home construction was something like a forecast error in this market. We can’t discount how builders may have been a little more enthusiastic about their business prospects given how sure it had appeared Economists and central bankers (redundant) were about the coming economic acceleration. Sprinkle in some inflation hysteria and perhaps builders were a little too sensitive toward moving forward activity.

Maybe builders started a few more apartment projects than they otherwise might have figuring it would be good practice to get ahead of a real boom this time; only to find, yet again, there was as little behind the talk as there always is.

However it may turn out in the short run, the home construction business continues to be stagnant. It shouldn’t be this way not with home prices rising and the labor market sizzling. Because it is this way, maybe the economy isn’t really booming, there are just these times when people are receptive to the idea it might start to. Depending on how the next few months shake out, this part of the economy may even be rolling over instead – starts first.