Questions are raised over accuracy of data

When the Office for National Statistics’ chief economist announced the first official estimate of economic output on Wednesday morning, he faced as many questions about the accuracy of the data as he did about the data themselves.

“We have no reason to believe these figures are any less reliable than would usually be the case,” Joe Grice of the ONS said several times in a press conference about the 0.2 per cent drop in output in the first quarter.

But not everyone was reassured. Kevin Daly, the UK economist at Goldman Sachs, called the ONS estimate “unbelievable”. Peter Dixon, the UK economist at Commerzbank, said: “Frankly, I don’t believe it.”

They were sceptical because unofficial indicators over the past three months had suggested the economy was growing again. The popular Markit purchasing managers’ index surveys of the construction, manufacturing and services sectors, for example, were consistent with output growth of about 0.5 per cent in the first quarter.

On Wednesday, the CBI employers’ group released a survey of the manufacturing sector that appeared to show improving orders and output volumes in the three months to April and the first improvement in sentiment in a year.

Disagreements over the reliability of official data are not uncommon, but they are important this time because the Bank of England’s Monetary Policy Committee appears to have sided with the sceptics, making it less likely the MPC will approve more quantitative easing next month.

In their meeting this month, MPC members said GDP might shrink in the first quarter but attributed it to “perplexing” official construction data. “Underlying aggregate activity growth was likely, if anything, to have picked up since the second half of 2011,” the minutes said.

The ONS’s “flash” estimate of GDP comes just 25 days after the quarter finishes, making it the fastest official estimate produced by any major developed country. By this time, statisticians have only gathered data covering 42 per cent of economic activity, so they use forecast models to fill in the blanks. Revisions are common in subsequent months and years, though they are usually small. The average GDP revision since 1997 has been to add 0.1 per cent, according to Capital Economics.

Some economists said the problem might lie with the unofficial surveys rather than with the official data. Allan Monks at JPMorgan said he was nervous about the reliability of the services PMI as a growth indicator. “For this reason, we are hesitant to fully downplay this GDP release.”

Richard Barwell, an economist at RBS, agreed. “A very bad indicator has lulled people into a false sense of the economy,” he said. “There is no evidence that they’ve [the ONS] got it wrong.”

But Chris Williamson, chief economist at Markit, which compiles the PMIs, said the surveys were sound. “The whole reason why we exist is because we didn’t believe the official data … and this just [reminds us] why we do what we do.”