It apparently happens in every business cycle so it isn’t something to get overly excited about. Yet it is an important distortion to those to whom it matters, as do the policy consequences.

Building activity in the traditional housing market (over 80m2) is today at half its long term average (building plans approved running at about 1000 units monthly compared to over 2000 monthly on average longer term).

The falloff in townhouse building activity has been even steeper after its long and exceptional run-up during the 2004-2008 boom. Plans approved (square metres) in 2010 were down by 77% from their 2005 peak.

Additions and alterations activity has suffered similar slump conditions after reaching their cyclical peak.

Non-residential building activity has been rapidly cooling off this past year, with building plans passed over 45% down since the 2008 peak, with further declines still possible this year.

The construction industry reports a 35% decline in civil engineering turnover last year, with this sector laying off some one-third its labour force these past 18 months.
              
Contrast these industry impressions and StatsSA data reports with the way SARB reports the relevant data in its fixed investment component per the quarterly bulletin (the 4Q2010 issue released early last week).

Residential fixed investment (in real terms) for 4Q2010 is reported as still declining at a 4% annual rate, being 6% down on a year ago (and 7% down in 2010 overall as compared to 2009).

The cyclical peak in residential fixed investment was 1Q2007 and by 4Q2010 fixed investment in this area was down some 25% (peak-to-trough).

Non-residential fixed investment showed little change in 4Q2010, but was still 2.2% up on a year ago, showing over 2% growth during 2010 compared to 2009, with this sector yet having to peak out.

Construction fixed investment also showed little change in 4Q2010, but was still up nearly 1% on a year ago, with 2010 being 2.5% up on 2009 as a whole, with this sector also as yet having to peak.
               
Are these contrasts for real?

Yes, they are. Apparently they happen every cycle.

In some instances, such as non-residential building activity, SARB adds in data from other sources in the public sector not otherwise easily accessible, presenting an overall fixed investment picture.

But the more pertinent difference has to do with what is taken into account. SARB tends to count as fixed investment only activity that has been completed (or an assessment of work completed).

As work completed takes much longer to tail off at the end of the building pipeline as compared to plans passed at the front end (much of which in any case may never come to fruition), the SARB fixed investment data tends to be very backward looking, with little inkling of changing conditions that may in fact already have deeply changed activity levels in the respective sectors.
               
On this score, present SARB estimates of real residential fixed investment is probably closest to being an accurate snapshot of operating conditions in this important building sector segment, even though the SARB data may still be subject to downward revisions to come as the full story trickles through the data collection process.

The biggest contrasts can be found in non-residential building activity and construction work.

Here one senses that with non-residential building plans passed run down as heavily as they have been and many existing contracts in their final completion phase, the fixed investment data in coming quarters could see quite a falloff, with this sector still far from complete in making its cyclical downward adjustment (on average lagging residential trends by up to two years).

The same reasoning applies to construction.

With important World Cup projects completed (think stadia, airports, Gautrain) and new civil engineering contract work having fallen off, recorded fixed investment in construction work is also likely to tail off in coming quarters despite large workflows-in-progress (Eskom power stations, roads, water projects).

Some of such follow-up work (Eskom power stations) qualifies in any case as machinery and equipment (thinking generators and turbines) and not necessarily as construction work.

The sheer falloff in civil engineering and cement sales are such that this should eventually become reflected even in SARB fixed investment data, especially during 2011-2012 if traditional lags are taken into account.

With residential, non-residential and construction works amounting to over one-third of total fixed investment (and 7% of GDP), these represent significant drag anchors on an otherwise rather rosy overall growth story, today still mainly consumption-led.
                      
Only if the contract flow were to pick up significantly might the overall picture in these sectors eventually change for the better beyond 2012.

In residential building activity a cyclical bottom is being ploughed at present. The fact that these past two years the traditional inverse relationship with interest rates seems to have dissolved ‘completely’, probably due to structural changes in the credit story, suggests that the present is NOT a normal building cycle.

It may take an inordinate period of time to restore traditional relationships between population and urbanization growth, income changes and interest rate levels.

Given the lags involved, the non-residential building sector may face similar realities yet to be fully faced up to.

Construction faces its own specific constraints, to identify projects and allocate contract work, as well as the lack of the necessary technical manpower available in the public sector to manage such work flow.

With technical manpower shortages apparently the main reality throughout much of the public sector, addressing these shortages may take time, for the duration probably remaining an important bottleneck.