CMD announced today that November’s level of U.S. construction starts, excluding residential work, was $24.0 billion, a decline of 2.5% versus October. The small drop was in line with the usual or long-term November-versus-October percentage change, due to seasonality, of -2.0%.

Versus the same month of last year, November 2015’s starts level was +1.6%. Com pared with average November starts during the preceding five years, the latest month’s
level was +9.4%.

Year-to-date starts in 2015 are ahead of the same January-to-November period in 2014 by 4.6%.

The starts figures throughout this report are not seasonally adjusted (NSA). Nor are they altered for inflation. They are expressed in what are termed ‘current’ as opposed
to ‘constant’ dollars. ‘Non-residential building’ plus ‘engineering/civil’ work accounts for a considerably larger share of total construction than residential activity. The former’s combined pro portion of total put-in-place construction in the Census Bureau’s October report was 63%; the latter’s was 37%.

CMD’s construction starts are leading indicators for the Census Bureau’s capital invest ment or put-in-place series. Also, the reporting period for starts (i.e., November 2015) is
one month ahead of the reporting period for the investment series (i.e., October 2015.)

According to the Bureau of Labor Statistics (BLS), total employment in the con struction sector increased by 46,000 jobs in November. That was the greatest month-to- month gain since January 2014’s +69,000 figure. The year-over-year percentage climb in construction jobs has been +4.2%, the fastest pace of increase among all major indus - trial job categories. Construction’s NSA unemployment rate in November was 6.2%. Twelve months earlier, it had been 7.5%.

Before there can be on-site construction activity, projects must be contemplated and planned by owners and rendered into working drawings by design professionals. On a month-to-month basis, total jobs in architectural and engineering services in November were -0.1%. On a year-over-year basis, the percentage change remained posi tive (+2.7%), although less so than at mid-year.

Almost equal month-to-month mild increases in commercial (+3.7%) and institutional (+3.6%) starts in November could not make up for heavy engineering’s (-10.4%) larger pull-back. Industrial work (+29.6%) had a nice gain, but it usually has a much lower dollar volume. Also, it can display severe variation based on whether or not there is a go-ahead for a mega project.

As for November of this year versus the same standalone month of 2014, all three major categories recorded pluses: commercial, +15.4%; engineering/civil, +5.5%; and institutional, +3.9%. Industrial (-90.2%), which often has wide swings, recorded an outsized drop.

Year-to-date starts continued to be led by the heavy engineering category, +14.3%. Commercial starts, at -0.2%, were almost exactly even with the same January-to- November time frame of last year. Institutional starts were down, -5.1%. Smaller-category industrial starts were +26.8%.

‘Road/highway’ work, with a 42% share currently, is the largest sub-component of the heavy engineering category. Such ‘street starts’ in November were +5.5% month-to- month (m/m); +17.5% year over year (y/y); and +11.8% year to date (ytd).

As for ‘water/sewage’ starts, which account for almost one-quarter of the civil total, they suffered a setback m/m (-16.6%), but were on the upside both y/y (+9.8%) and ytd (+15.0%).

Before moving on from engineering, it should be noted that another of its important sub-categories, bridges, had a good November: +4.0% m/m; +10.5% y/y; and +15.6% ytd. With nearly two-thirds of such work, the ‘school/college’ sub-category is dominant within institutional construction starts. In November, educational facility starts were: -5.4% m/m; but +11.3% y/y; while a barely-different +0.2% ytd.

Also significant under the institutional umbrella, ‘hospital/clinic’ starts in November surged +48.6% m/m; improved +5.0% y/y; but were a disappointing -13.1% ytd.

The ‘police/courthouse/prison’ category within institutional had a strong November: +60.3% m/m; +132.9% y/y; and +61.2% ytd. Prison starts (+183.8% ytd) have been particularly bullish.

Commercial starts are being bolstered by strong retail representation: +19.2% m/m; +18.9% y/y; and +7.8% ytd. Hotel/motel work has been mainly upbeat: +183.2% m/m; +26.7% y/y; but a barely better +1.6% ytd.

While it may appear that ‘private office building’ starts have been seriously lagging (-42.7% m/m; -25.3% y/y; and -24.5% ytd), the numbers fail to take into account the exceptional strength in a ‘warehouse’ category (+194.0% m/m; +266.6% y/y; and +42.4% ytd) that includes a substantial office, or at least office-similar, component. CMD categorizes large call center and data center projects to its ‘warehouse/office’ type-of-structure category.

Among the 12-month moving average trend-line graphs on page four of this report, the non-residential building categories are staying flat or declining, while the civil/engineering subsets are moving higher. In the ‘Top 10’ project starts list on page three, Texas stands out among the six states, contributing four projects. Next in line is Pennsylvania with two.The rest have one project each, although Ohio also deserves mention for its giant Amazon data center.

Growing tightness in labor markets has yet to affect compensation levels. According to the BLS, average hourly earnings for all jobs (including supervisory) in the U.S. economy fell from +2.5% year over year in October to +2.3% in November. Construction, though, did move in the opposite direction, from +2.6% to +2.7%. As for average weekly earnings, they slipped from +2.5% to +2.0% economy-wide and also retreated in the construction sector, from +4.1% to +2.9%.

The value of construction starts each month is summarized from CMD’s database of all active construction projects in the U.S. Missing project values are estimated with the help of RSMeans’ building cost models.CMD’s non-residential construction starts series, because it is comprised of total-value estimates for individual projects, some of which are ultra-sized, has a history of being more volatile than many other leading indicators for the economy.

Contact: