Stephen King: Illusion of sustained US recovery is fading fast

If there's one bit of data guaranteed to get economists excited from one month to the next, it's the US employment report. Whatever its vagaries and inconsistencies, it's probably about the best and most timely barometer of what's going on in the world's largest economy. The numbers are often revised and so history is constantly changing, but the payrolls report has taken on an almost mythical quality.

This year, the numbers are particularly important: the US Federal Reserve has turned the corner on interest rates and, whatever happens in Iraq, you can more or less guarantee that jobs will be a central issue in the run-up to the presidential election. So if you want to know what the Federal Reserve is up to - and perhaps have a wager on whether George W Bush will still be President in 2005 - it's wise to have a quick look at the latest US labour market trends.

Friday's release was extraordinary. Even if you didn't know the precise details, one look at the markets was enough to tell you that the July numbers were genuinely shocking. In intraday trading, the Dow Jones Industrial Average fell 1.5 per cent, the FTSE 100 index was down 1.8 per cent and the Nasdaq composite was down 2.5 per cent. Meanwhile, the yield on US 10-year Treasury bonds dropped to just 4.2 per cent, a four-month low.

Why were the figures such a surprise? Payrolls rose 32,000 on the month, so at least there were no net job losses. The market, though, had anticipated an increase well in excess of 200,000. There were also downward revisions to the gains recorded in the previous two months. All in all, the strong recovery in payrolls that was seemingly under way earlier in the year suddenly appeared to be coming to a juddering halt.

It's worth putting these numbers in context. The chart shows the performance of payrolls in this economic cycle compared with all other economic cycles in the past 50 years. I have indexed employment in each cycle to equal 100 at the peak in economic activity as defined by the National Bureau of Economic Research. By doing this, it's possible to plot America's employment experience through both downswings and subsequent recoveries.

The initial job losses vary quite a lot from one cycle to the next. The chart shows that, in the latest cycle, the initial job losses were nothing extraordinary: a decidedly average recession, one could say. What has set this cycle apart has been the lack of decent jobs growth in the subsequent recovery. Other than the double-dip recession of the early 1980s, all other recoveries have been associated with a level of employment higher than at the previous peak.

Up until Friday, the one redeeming feature of the latest employment cycle was the better overall performance relative to the earlier double-dip period. Sadly, this claim can no longer be made. Although there are bound to be subsequent revisions, it's now possible to say that, from the last peak in economic activity, this economic cycle represents the worst employment performance from the US economy in the past 50 years.

Of course, optimists will turn this observation on its head and argue that, given the relatively shallow recession and reasonable subsequent recovery in output, this economic cycle represents one of the best productivity performances from the US economy in the past five decades.

Both observations are, of course, true. But what do they imply? For the election, it's surely the first claim that matters more. After all, a lot of voters are also workers and, productivity or no productivity, workers need jobs. For monetary policy, though, the story is more confusing.

In the past, the Federal Reserve has tended to be steered by the employment report. In a standard US economic recovery, job gains average somewhere between 200,000 and 400,000 a month, enough to persuade the Fed to go on a sustained monetary-tightening spree.

The Fed's problem is that jobs growth has fallen to such a pitiful pace that it becomes more difficult to justify rate increases. Yet, having already raised interest rates on one occasion, and having given very clear hints about further rate increases later in the year, a failure to do anything more now would reveal a sudden loss of faith in the US recovery that could do lasting damage to the Fed's reputation. It would look as though the Fed had lost its way, had perhaps jumped to the wrong conclusion about the nature of the economic recovery.

The Fed can still argue that, with or without jobs growth, the economy is growing and that, as a result, some of the excessive monetary accommodation of earlier years should be removed. This argument, though, is a lot more difficult to justify if the employment numbers are heading in the wrong direction: after all, the Fed only felt comfortable in raising interest rates after the payrolls numbers had shown apparently clear signs of recovery so it's difficult to see how the Fed can continue to raise interest rates in the light of renewed labour market weakness.

So what's gone wrong? Maybe nothing at all: the vagaries of the employment report are such that a major rebound in August is a distinct possibility, thereby alleviating some of the current unease. Let's say, though, that the rebound comes, but is not particularly large: sufficient, perhaps, to keep the average monthly increase in jobs at a rate of 100,000 or thereabouts, but below the threshold typically associated with rate increases.

What sort of explanations would then come into play? For me, the more obvious structural explanations are related to offshoring and outsourcing. US companies are hiring workers, but they're not US workers.

I'm beginning to wonder, though, whether the standard cyclical arguments are also not working terribly well. The Fed's view of the world is straightforward: get companies to invest, to take risk and to expand and, eventually, employment will recover.

For a time, this story looked good: the recovery in business surveys in recent months suggested that all was going to plan. But if employment is now stalling, the story looks a lot less convincing.

I suspect that the relationship between profits, investment and jobs has changed. Profits did rise strongly last year. Capital spending did pick up. But the recovery in capital spending was not necessarily a sign of sustained economic health. To pay off debt, companies had to extract large productivity gains from existing workers and were helped in doing so via an environment of remarkably low interest rates. Those low interest rates persuaded companies to alter the balance of their factors of production: capital spending rose because it was cheaper to employ capital than labour. At no point, though, did companies really choose to add a lot to existing capacity: this was cost-saving investment, not demand-expanding investment.

In other words, the environment of low interest rates perhaps fostered the illusion of a sustained recovery but, ultimately, did little more than that. If the payrolls gains do fade, the Fed will eventually be forced to change course. Then the future of its illusion will no longer look quite so bright.

Stephen King is managing director of economics at HSBC

stephen.king@hsbcib.com