How low can you go?

By Donal Curtin

Saturday 1st May 2004

Here's a question that might seem daft at first sight: how low should unemployment go? As low as possible, of course (ideally to zero), you'd be tempted to answer - what other answer could there be? Isn't full employment what every government is looking to achieve?

That's true - but there's a risk that if the labour market gets very, very tight, there's the potential for a nasty outbreak of inflation. If employers get desperate to hire the last person left, or to bid away staff from competitors, or if employees feel they can thump the table for a big pay raise, then there's a real chance of an ugly episode of wage and price inflation.

That, at least, used to be the story: economists reckoned there was some (lowish) level of unemployment beyond which it wasn't prudent to venture. They had various names for it - the "non accelerating inflation rate of unemployment" (often shortened to "NAIRU") was one ungainly description, while the "natural" rate of unemployment was another, suggesting - wrongly, as we will see - that there was some God-given unemployment rate we were stuck with.

And they even had a decent go at measuring what that "safety limit" unemployment rate might be. The OECD's website is such an unnavigable mess you can never be sure you've got the latest version of anything, but it's probable that their latest estimates (prepared in 2001 and relating to 1999) show that New Zealand's NAIRU was 6.1%, Australia's 6.8% and America's 5.2%. A few lucky countries had low rates - Switzerland's was only 2.4%, and Norway's 3.7% - while some had unusually awful rates, notably France with 9.5% and Italy with 10.4%, not to mention the scarcely credible 15.1% in Spain. (In other words, if Spanish unemployment fell below 15%, their inflation rate would start to rise).

At this point any semi-comatose readers will leap back into their usual state of alert interest, having spotted what looks like a clanger. If New Zealand's safety limit rate of unemployment is supposedly 6.1%, like these OECD wallahs reckon, how come we haven't had a massive outbreak of inflation, given that our actual rate of unemployment is well below the safety limit? Last September, for example, our unemployment rate hit a new low of 4.4%. Even though it bobbed up a bit to 4.6% by December, it's well into territory that, if this NAIRU stuff is at all on the mark, ought to have sent us into a wage-price spiral.

In fact, the real world unemployment rates in a number of countries, including New Zealand, have in recent years fallen below the alleged roadblock rates, with no ill effects by way of higher inflation. In fact, exactly the opposite has happened - unemployment and inflation have both fallen to low levels.

There's been, as you'd expect, quite a bit of debate on why this has happened, since it's a huge bonus if your economy can run at a 4.4% unemployment rate without blowing any gaskets, instead of being hobbled to 6% (using New Zealand's figures). It was especially debated in America: there, it had long been almost a given that unemployment couldn't fall much below 5.5% without bad things happening, such as the Fed having to raise interest rates. In reality, unemployment dropped to 4%, and the inflation bogey never turned up to spoil things.

The answer, as best economists understand it, is that these NAIRUs aren't a natural and immutable feature of the landscape, but can and do change. In many countries, they have been falling, for two reasons. One is that labour markets are working better in a number of countries: you can look at the NAIRU as a way of measuring how well your labour markets match jobs with people. The higher it is, the worse they are: if you get wage pressures developing when unemployment is 10%, clearly there's something amiss preventing the 10% from getting at the available jobs (if they could, the incumbent 90% wouldn't be able to get away with jacking up the rate for the job).

Let's sidestep the partisan positions taken on the success or otherwise of the Employment Contracts Act, and just note that, in principle, you can choose to devise smart industrial relations regimes - with correspondingly lower NAIRUs - or you can devise clunky, inflexible ones, with higher NAIRUs. Increasingly, the evidence is that a number of countries devised better arrangements in the 1990s, and saw the benefit of it in lower inflation for any given unemployment rate.

What's more, the size of the payoff may well have been substantial: one piece of research, by economists at the Federal Reserve Board in the US, included New Zealand in its reach and found that our falling NAIRU was good enough to take 3% off what our inflation rate would otherwise have been.

Better functioning labour markets are the main reason the unemployment speed limit has been easing, but there is another one as well. Unfortunately, it applies only to the US at the moment. That other reason is a sharply higher rate of productivity growth in the late 1990s. That also tends to ease the speed limit: when staff are scarce, the pay for the job rises, but employers can afford to pay more without raising prices, thanks to the productivity gains.

In sum, we've done some good things in New Zealand. Unemployment doesn't have to be pegged to 6.1%, and the roadblock is coming down all the time. All that said, 4.6% of the workforce is still unemployed: next month, we look at why.

Comments from our readers

No comments yet