Monday, June 11, 2012

Why we can't read the economy without help

The nation's economists, commentators and business people got caught with their pants down last week. They'd convinced themselves the economy was weak, but the Bureau of Statistics produced figures showing it was remarkably strong.

It's not the first time they've failed such a reality test. They prefer not to think about such embarrassing, humbling occurrences, but it's important to ask ourselves why we got it so wrong.

The bureau told us real gross domestic product grew by 1.3 per cent in the March quarter and by 4.3 per cent over the year to March. Then it produced labour force figures for May, showing employment has been growing at the rate of 25,000 a month this year, with much of that growth in NSW and Victoria.

So why is there such a yawning gap between what we thought was happening in the economy and what statistics say is happening?

Well, one possibility is the figures are wrong. That's likely to be true - to some extent. They're highly volatile from quarter to quarter and month to month, and much of that volatility is likely to be statistical "noise" rather than "signal".

But the financial markets, economists and media knowingly add to the noise by insisting on using the seasonally adjusted figures rather than the trend (smoothed seasonally adjusted) figures as the bureau urges them to. Truth is, both markets and media have a vested interest in volatility for its own sake - it makes for better bets and better stories.

However, even if the latest figures are likely to be revised down, their "back story" still contradicts the conventional wisdom. Cut March quarter growth back to the 0.6 per cent economists were forecasting and you're still left with above-trend annual growth of 3.6 per cent.

Consumer spending may not have grown by as much as 1.6 per cent in the March quarter, but - and notwithstanding all the retailers' complaints - it's been growing at above-trend rates for a year.

Another argument embarrassed economists are making is that the March quarter figures are "backward looking". All the news since March has been bad. They always use that excuse. But there's nothing out of date about job figures for May, and they, too, tell a story of strengthening growth.

If you accept, as you should, the figures are roughly right - especially viewed over a run of months or quarters - you have to ask how our perceptions of the economy have got so far astray from statistical reality.

It's less surprising business people's perceptions are off the mark. They're not students of economic theory or statistical indicators; their judgments are unashamedly subjective, based on direct experience and the anecdotes they hear from other business people, plus an overlay of what the media tell them.

More surprising is the evidence economists' judgments and forecasts aren't as rigorously objective and indicator-based as they like to imagine. They're affected by the mood of the business people they associate with and aren't immune to the distorted picture of reality spread by the media (because they highlight events that are interesting - and, hence, predominantly bad - rather than representative).

Like the punters, business people probably overestimate the macro-economic significance of falls in the sharemarket - particularly when our sharemarket is taking its lead from overseas markets reacting to economic news in the US and Europe that doesn't have much direct bearing on our economy.

Similarly, all the bad news from America and, particularly, Europe we're hearing from the media night after night can't help infecting our views about our economy. We're getting more economic news from China these days but we hear about the threats rather than the opportunities.

The familiar refrain about the alleged two-speed economy is tailor-made for the media but, as last week's figures make clear, an exaggeration of the truth. Consumer spending is reasonably strong in the non-mining states, as is employment growth this year.

In the absence of anything better, economists and the media persist in setting too much weight on the bureau's quarterly figures for state final demand, unaware they give an exaggerated picture of the differences in gross state product between the mining and non-mining states (because Western Australia and Queensland use much of their income to buy goods and services from NSW and Victoria).

The risk is the more we repeat the two-speed story to ourselves the more it becomes a self-fulfilling prophesy. This may be part of the explanation for the weakness in non-mining business investment spending, but as yet it doesn't seem to have affected consumer spending.

The media's highlighting of announced job lay-offs is a classic example of the way their inevitably selective reporting of job movements leaves the public, business people and maybe even economists with a falsely negative impression of the state of the labour market.

A recent list of 25 lay-off announcements showed total job losses of 17,000. When people wonder how the bureau's employment figures could be right when we know so many jobs are being lost, they're showing their ignorance of how selective media reporting is and how big the labour market is.

In a workforce of 11.5 million people, job losses of 17,000 are peanuts (though not, of course, to the individuals involved). Far more than 17,000 workers leave their jobs every month and far more take up jobs every month. The media tell us about just some of the job losses and about virtually none of the job gains.

The unvarnished truth - which none of us can admit, even to ourselves - is we think we know what's happening in the economy, but we don't. We're too fallible, and it's too big and complicated.