The AEC Lens
Alex Carrick, Chief Economist at ConstructConnect
Alex Carrick is Chief Economist for ConstructConnect. He is a frequent contributor to the Daily Commercial News and the Journal of Commerce. He has delivered presentations throughout North America on the Canadian, United States and world construction outlooks. A trusted and often-quoted source for … More »
U.S. Put-in-place Construction Growth to be near 9% in 2016 and 2017 (Parts 1)
December 23rd, 2015 by Alex Carrick, Chief Economist at ConstructConnect
Article source: CMDGroup
The focus for CMD’s construction statistics, both in the U.S. and Canada, is on actual and forecast levels of starts.
There is another data set supplied by government agencies – i.e., the Census Bureau and Statistics Canada − known as the put-in-place (PIP) investment spending series.
For ‘starts’, the total value of a project is entered in the month when, according to a best estimate, ground is broken. The starts are often referred to as ‘lumpy’, since the largest projects play outsized roles.
Starts totals are built-up from the summation of all individual projects that are in the data base.
Conceptually, the PIP data set differs in that it simulates progress payments as projects proceed.
For example, while PIP numbers are actually based on owners’ and others’ estimates of capital spending across a total universe of construction activity, a $60 billion office building beginning in July of this year will be theoretically captured (approximately) as $20 billion appearing in 2015; $30 billion in 2016; and the final $10 billion in 2017.
When graphed, both sets of numbers display cyclical patterns (i.e., waves over time) for quite a number of different type-of-structure categories.
The curves for the starts series, though, have wider amplitudes and their turning points occur out front of (i.e., ahead of) the PIP numbers. In other words, the starts are leading indicators for the investment spending series
Since many executives and analysts working in, or fascinated by, the construction sector follow both series, CMD calculates forecasts for the PIP numbers as well as for its own starts information.
CMD’s latest updated PIP projections for the U.S., with accompanying graphs, are presented in the next Economy at a Glance.
Before proceeding to that stage, however, let’s exam some history, as set out in the accompanying table.
Only those type-of-structure categories displaying clear cyclical patterns, when graphed, are shown with detailed information in Table 1.
For example, ‘n/a’ or ‘not applicable’ appears in most of the boxes for ‘Total Engineering’ work.
Neither total engineering, nor any of its sub-components − with the possible exception of ‘sewage and water disposal’ work − has displayed a cyclical pattern in seasonally adjusted and annualized (SAAR) monthly data from January 2002 to the present.
Cyclical waves have been almost exclusively confined to the residential and non-residential building categories.
For each category of construction, its pre-Great Recession peak value (and date when achieved) is shown in column ‘A’. Subsequent troughs are set out in column ‘B’. The percentage drop, peak-to-trough, appears in ‘C’.
From column ‘C’, it is apparent that residential construction (-65.4%) suffered a sharper descent on account of the recession than did non-residential building work (-41.0%), although neither remotely could be said to have escaped lightly.
Within non-residential building, the ‘lodging’ (-77.7%), ‘commercial/retail’ (-60.0%), ‘manufacturing’ (-54.6%) and ‘office’ (-50.9%) sub-categories all recorded declines of more than half. (This yields the logical inference that those are the type-of-structure categories most influenced by the overall economic cycle.)
I’ve written before that both total output (i.e., gross domestic product) and employment in the U.S. have recovered what they lost in the recession and have moved on to new heights, but that construction activity continues to lag.
The veracity of that statement is most clearly proven by the results shown in column ‘E’.
The figures in ‘E’ result from calculating ‘D’-minus-‘B’ as a percentage of ‘A’-minus-‘B’. They show the degree to which current levels in each type-of-structure category have reclaimed ground lost in the recession.
Only one category, ‘manufacturing’, has met and matched its former pinnacle and realized a substantial improvement as well (+35.5%).
Otherwise, ‘amusement and recreation’ (+87.5%) is doing best, followed by ‘offices’ (+65.7%).
‘Health care’ (+31.7%) and ‘educational’ (+34.1%) are only about one-third of the way back to their former glory, but it should also be noted that their peak-to-trough declines (-21.3% and -31.6% respectively) were not nearly as extreme as for the other sub-categories.
As for total residential versus total non-res building, the former has only made up 37.9% of its recessionary shortfall compared with 77.5% for the latter.
U.S. PIP series: http://www.census.gov/construction/c30/c30index.html
Category: CMD Group