What we can learn about the economy from construction data

Years from now, when the bards of economic lore regale audiences with tales of the Great Pandemic of 2019-22, few stories will be more captivating than the ones based on the U.S. construction sector.

The data has all of the dramatic elements of the day: rampant inflation, acute labor shortages, large-scale supply chain disruptions, seismic cultural shifts, lots of exposure to interest rate risk, an unprecedented level of pent-up demand.

And to think we get to live through all of this in real time.

For plastics materials suppliers and processors in this sector, I expect strong demand to continue for the remainder of 2022. However, the pace of growth will decelerate as the year progresses.

Jerome Powell has signaled that the Fed will be increasingly aggressive in its efforts to quell the rate of inflation. There is no way of predicting what the Fed will do, but no sector provides a better illustration of the complicated cause-and-effect relationship of inflation than the construction sector. Here is what I see in the current data as we embark on a new phase of Fed tightening.

Despite all of the turbulence reflected in some other sectors, total U.S. construction spending increased by a robust 12 percent in the first quarter when compared with the same period in 2021. Total spending for residential projects jumped almost 19 percent, while spending in the nonresidential segment increased a very respectable 6 percent.

I have graphed the historical 12-month rates of change for the data. The chart shows there was a brief dip in residential spending in 2019 with the onset of COVID-19, but spending levels started to recover in the second half of 2020. By 2021, the pace of growth hit another gear. Overall, the rate of change coming out of the moderate cyclical dip in 2019 has been much stronger than I expected, given the historical patterns.

Read more of Bill Wood's Numbers That Matter columns.

If you look at the longer-term demographic data, it was apparent that the residential construction sector was underperforming since the housing bubble burst in 2008. In an attempt to explain this subpar performance, there were stories of millennials living in their parents' basements and waiting to start their own households until they were older.

This did not square entirely with the data on household formation, which was showing steady expansion during most of this period. And during all of this time, interest rates were at historically low levels. So the reason for underinvestment for such a long period in new housing is still not entirely clear. But none of that really matters now because the pandemic changed everything.

Higher interest rates are needed to put the brakes on the demand pressure in this market segment and overall inflation in the economy. But how much higher and how fast they must be raised are definitely open questions. In my opinion, activity levels in the residential housing sector for the next year or two will be the most vital indicators for understanding the U.S. economy.

For the processors and suppliers that are affected by the trends in the nonresidential sector, the historical pattern is a significantly different one. But there are still some trends worthy of mention.

The rate of growth in the overall nonresidential data did not fall into negative territory in 2019. These types of projects are longer term in scope, and the short duration of the downturn in 2019 combined with massive amounts of federal relief kept overall spending levels positive that year.

By 2020, the combination of supply chain disruptions, labor shortages and shifting cultural behavior patterns took its toll. Total spending for nonresidential projects actually declined by 3 percent — a sharp contrast from the unprecedented 23 percent spike in residential data.

So what does all of this mean for the nonresidential segment going forward? The trends of the past couple of quarters offer some clues. By the way, the monthly data from the Census Bureau is broken down into a number of useful nonresidential categories, and I encourage you to log on and read the monthly news release periodically.

The fastest growth in the first quarter, over 33 percent when compared with last year, was for projects related to manufacturing. I find this encouraging because it just might indicate a trend toward reshoring and rebuilding our domestic-based supply chains.

It will take more than one strong quarter to prove this hypothesis, so I will monitor this category closely. Another strong gainer last quarter was the commercial category, up almost 18 percent over a year ago. This is the category that includes new warehouse projects.

The most surprising increase for me was the moderate gain of almost 5 percent in new office projects. I have traveled extensively by car in recent months, and one theme is a widespread abundance of vacant office and retail spaces. My working theory is that the workers who traditionally occupied these work areas are either working from home, or they were lost when the economy shut down.

There were also some large decreases posted in some of the nonresidential categories in Q1. The largest drop — down 31 percent — was in the relatively small category of public safety projects. The lodging category, another small category, registered the second-biggest decrease in spending — down 27 percent.

The market will continue to find a point of equilibrium in all of these trends, and the construction spending data will eventually have to reconcile with the new equilibrium. The ongoing challenge for processors will be to anticipate which of these changes are cyclical and which are permanent. That has always been the case, but we are presently living through once-in-a-lifetime conditions. And I do not believe that the high volatility and the large quantity of changes in such a short period of time will settle down this year.

At this moment, the interest rate story predominates. Changes in the trajectory of the data for construction spending are still dependent on the trend in interest rates, but with a lag. And to be sure, the current level of interest rates, when adjusted for inflation, are still quite accommodative. So the deceleration I expect for the rest of 2022 will be caused by supply constraints rather than weakening demand.

But if the Fed accelerates the pace of its rate increases, and if this faster pace surprises the financial markets, then the economy could fall into a widespread recession. In such an environment, it would not surprise me to see a sharp pullback in construction spending.