It’s a Floor Wax and a Dessert Topping

In considering the persistence of unemployment after the late recession, Cardiff Garcia at FT Alphaville argues that the decline in construction employment is solely cyclical:

There’s a decline that immediately precedes or coincides with each recession, and an increase (at what appears to be a very similar slope) that begins shortly after each recession ends. And each time, construction employment climbed to levels well above the previous peak.

Far from being bubblicious, this is simply what you would expect in a developed country with a growing population and generally rising living standards. And although the US faces some difficult demographic issues, they’re not so dramatic just yet that you should anticipate a radically different trend in this recovery

Take a look at the post which has plenty of charts, graphs, and links to support the claim.

I’ve taken the liberty of producing a graph of my own:

This represents total construction workers related to population over the period 1952 to the present. I’ve also introduced two trend lines, one linking peak to peak, the other trough to trough. They’re obvious enough in the graph over at FT but there are three anomalies in the cyclical construction employment story: the sharp deviation from trend on the upside in the 1990s and continuing until the Great Recession, the deviation from trend on the downside in the early Aughts, and the very sharp downwards decline following the last recession. I don’t see a cyclical explanation for any of them and, particularly, I think the “growing population” is clearly irrelevant. If anyone can produce a cyclical explanation for those anomalies, I’d welcome it.

My explanation is that there are both cyclical and structural forces at work in the present situation in construction employment and the present decline is approximately a reversion to the old trend. If that’s the case, it suggests that construction employment will, indeed, grow in the coming years but is unlikely to return to its heights of before the late recession.

3 comments… add one
  • Drew

    I think you’ve clearly got the better read on it, Dave.

    Our firm has three general areas of focus, and one is infrastructure and “building products.” We have done very well. It started in the early 90s, and our last investment in the space was realized in 2006. We did not foresee the debacle. We aren’t that smart. but we saw things were very long in the tooth and overheated. We ran away.

    So I’ve seen and read more graphs and studies on commercial and individual real estate construction than I care to admit. The two are traditionally out of phase, so we will have to see how commercial rebounds even if r evidential does not, but I’m not optimistic.

    From where I sit, the “Dave-o-Graph” even though it mixes the two sectors is spot on. Like Mr Biden, we got out over our skis. Residential will find a so called new normal at lower levels. But in addition, commercial, the usual cyclical, is in a terrible state of oversupply at both retail and office. Without a significant uptick in GDP and employment (you gotta employ people to go in those empty offices and retail establishments) it’s going to be a long time.

    Shorter: structural will dominate cyclical for quite some time.

  • The CBO’s analysis finds about a third of the increase in total unemployment relative to before the recession structural. That may be about right. However, in construction, which is what this post is about, the ratio is a bit different. I have no doubt that some of the unemployment in construction is cyclical. But I think a lot of it is structural which in turn implies that a lot of the structural unemployment in the economy is in the construction sector which seems to me to stand to reason.

    If that’s the case there’s actually a larger issue which I thought about but didn’t work into the post. The country with the highest proportion of people having college educations is Canada at about 50%. That’s probably about the limit. That means that there will need to be jobs for about half the people that don’t require college educations.

    Where were the jobs that didn’t require college educations? Other than the low-compensated jobs in healthcare, I mean. Manufacturing, construction, retail, hospitality, farming, oil and gas production, mining, lumber.

    Construction is likely to be in the doldrums for some time. Increasingly, manufacturing may not require college educations but it will require college education equivalents—various different skills in high tech equipment and processes. Retail and hospitality are dead ends—they’re likely to continue to be poorly compensated. That leaves farming, oil and gas production, mining, and lumber.

    Those fields really need to be seen as more acceptable, both to people seeking jobs and to policy makers.

  • steve

    Farming can take some real education also (beyond the high school level). I would add child care to the list of jobs not requiring college. Mostly a dead end also.


Leave a Comment