Recession may be revised, but too late?

The Office for National Statistics said its preliminary estimate of first-quarter gross domestic product shows the economy shrank 0.2% in the three months to March, compared with the previous quarter. It shrank in the final three months of 2011, and economists typically define a recession as two consecutive quarters of contraction.

Data showed that construction sector output slumped, industrial production was down and the U.K.’s powerhouse services sector eked out a miserable 0.1% expansion.

But economists were left scratching their heads. They expected the economy to have grown in the first quarter, albeit only by a measly 0.1%, according to polls.

That’s largely because the multitude of surveys covering U.K. businesses haven’t been consistent with falling output. Economists were particuarly puzzled by the weakness of the services sector and the sharp drop in construction activity.

Surveys of purchasing managers in services suggested activity in that sector has been considerably stronger than the headline numbers suggest. The apparent collapse in construction activity is also raising eyebrows, since the U.K. had a much milder winter this year than it did last year and house prices appear to be holding up better. Both ususally aid building activity.

The ONS takes pains to remind people that its first estimate of GDP if based on 42% of the solid data that eventually goes into its final estimate and revisions aren’t unusual. It’s therefore possible this recession will be revised away at a later date.

But that will be too late late for the households and businesses tuning into the evening news or picking up newspapers tomorrow morning. All they will hear is that the U.K. has double-dipped and is back in recession less than two-and-a-half years after it crawled out of the previous one.

The risk, as the Bank of England has acknowledged, is that the gloomy data become self-fulfilling. Firms’ and consumers’ already-fragile confidence may take a severe knock if all they hear is ‘recession’. Then arguments about the relative merits of different data series will be all for nought.

(The article was first published in The Wall Street Journal)