The recession of last month’s ONS data disappears, not unexpectedly, in this month’s revisions. But some practices are still feeling a bit spooked
Good news from the latest release of construction output estimates: the industry does not now appear to be in recession.
Last month the figures suggested a technical recession in the first two quarters of this year. This month, after significant revisions to back data, the Office for National Statistics estimates that the industry grew in the first quarter of 2016 before declining slightly more recently.
The disappearance of this ‘phantom’ recession was not unexpected. Revisions are part and parcel of statistical series such as the one that defines construction activity. Technically, later figures might point again to a recession, though this seems unlikely.
It would be wrong to criticise the series for this seeming vacillation and argue that such turnarounds undermine its value. It simply illustrates the need for caution over the first estimates of data, particularly in construction where measurement is extremely tricky.
Chart 1 shows the picture of past construction activity the ONS is now presenting. It points to an industry that has pretty much flatlined for the past year or so, after surging out of a double-dip recession in early 2013.
Look in more depth at the figures and it’s clear that the driving force of that rise in activity was housebuilding and a surge in infrastructure work. It’s not surprising that the subsequent slowdown can also be traced to these sectors. Both have coming off the boil, since late 2014 in the case of housebuilding, and early 2015 for infrastructure.
There is obviously a link between construction output and architectural workload, but it is far from straightforward. A strong construction sector supports architectural activity, but the weight of work undertaken by architects tends to run ahead of work on site.
However, as we see from the latest RIBA Future Trends survey, the broad picture of workload easing in 2015 in the construction output data is mirrored in architects’ expectations of future work. Put these two data series together and the hint is that the appetite for construction work weakened to a more modest level.
But the consistency we might have found between these two data series ended abruptly with the latest RIBA survey. In June, a jolly 22% balance of architectural practices expected future work to rise, but in July the balance plunged to a negative 7%. Chart 2 shows confidence metaphorically falling off a cliff.
This is a Brexit effect. Unpicking the full meaning and gauging how seriously we should take the latest figures is far from simple. The demographic, cultural norms and London-centricity of architecture in the UK all heavily point to a high proportion of its practitioners voting to remain in the EU. Irrespective of the actual and likely effects of leaving the EU, it is reasonable to assume those who voted to remain would become more fearful for the future after the result, while there would be greater satisfaction among those voting to leave.
That doesn’t mean the result has no substance, and that it is simply a reflection of sentiment. There is plenty of anecdotal evidence to suggest that potential work has been shelved as a result of the vote to leave the EU and the uncertainty it has generated.
Sentiment will, however, have played a part in this sudden drop in confidence. We saw that within the Markit PMIs (purchasing managers’ indexes). Taking the Markit/CIPS construction index, following the referendum vote there were big drops in June and July to negative levels not seen for seven years. As we see in Chart 3, this was followed by a sharp bounceback in the index – albeit still to a negative position.
A wanton misunderstanding of the mechanics of this type of indicator led to some awful interpretations being peddled. This was not restricted to questionable politicians using them to puff up their positions, but included those whom others might look to as ‘experts’.
We should expect these types of indicators to be rather frenetic over the coming months as people try to assess the impact of Brexit on their businesses. Uncertainty does create an oversensitivity in what are at least to some extent sentiment-based measures.
Turning to what might be described as more tangible measures, the latest quarterly ONS new orders data were released this month. For those who want to see good news here, there is plenty to latch on to.
The dip in orders for new construction work seen in the first quarter of this year was reversed in the second quarter. This could be read as an upturn for the industry, but taking that view could be, perhaps should be, dismissed as premature.
There is huge volatility in the value of contracts awarded in the construction sector. Added to this, reading the data is made all the tougher by the need to gauge to what extent, if at all, the slowdown in the first quarter was linked to delays or uncertainty caused by the referendum.
Encouragingly, though, we can say that the figures do not look like bad news. We will, however, need a quarter or two more sets of new orders data before we can sensibly assess whether there was or is a Brexit effect.
If this guarded approach to analysing construction seems a bit lily-livered, it is worth noting that the Construction Products Association has stepped away from its traditional position of producing a summer forecast to present scenarios. This tells us pretty much everything we need to know about the level of uncertainty.
Well almost everything. The wideness of the spread of possible futures presented by the forecasters, as shown in Chart 5, tells us much about the scale of uncertainty as seen through the eyes of some of the smartest experts in the sector.