Partly cloudy

Economists could see less clearly than weathermen

Why did God invent economists? To make weathermen feel good about themselves. The old joke is still good. Just look at the sharp debate about US GDP growth in the first quarter of 2015, which ended six weeks ago. At least meteorologists know what the weather was yesterday.

The Bureau of Economic Analysis (BEA) reported 0.2 percent annualised growth on April 29. The number was a big disappointment - the consensus forecast was 1.0 per cent. It was also far below the prior three quarters' average of nearly 4 per cent.

However, there have been two strangely bad first quarters in recent years. In 2014, the BEA reported January-March GDP shrinking at a 2.1 per cent annual pace, while the preceding and succeeding quarters hummed along at around a 4 per cent clip. The weak-surrounded-by-strong pattern was similar in 2011.

Some Federal Reserve analysts were puzzled. Their number-crunching did not reveal any definite problems with the seasonal adjustments which are used to take into account the winter quarter's inevitable slowdowns. However, they did suggest "something a bit unusual."

Barclays' economists rejected the Fed conclusion. They focused on estimates for investments, accounting for about 9 per cent of GDP. The bank reckons that a more accurate adjustment yields a first-quarter annual growth rate of 1.8 per cent - something the San Francisco Fed also concluded in a study published Monday.

To add to the confusion, the Atlanta Federal Reserve Bank's GDPNow model showed a measly 0.7 per cent pace for the current quarter last week. That suggests a slowdown lasting more than a quarter. But GDPNow is new and not all that accurate. The error range halfway through the quarter is about 2 percentage points on the annual growth rate.

The focus on the BEA's first advance GDP number is overdone in any case. The calculation relies on past seasonal trends, and just one quarter of the data is firm. The rest comes from data which is indicated, indirect or inferred.

An accurate count of quarterly GDP cannot be made quickly. Investors would do better looking at simpler statistics about current activity - and let academics worry later about how they all add up.