The first casualty of elections, too

You don't expect much sense to be talked about anything during a general election (and politicos wonder why large groups of the electorate clock out of the whole process...). The big economic statistics are as much casualties as everything else.

So in the interest of clearing away some of the smoke that the spin merchants are trying to blow up your (insert body part of choice here), let's just revisit some of the basics of what's happening to pay. Because, while yesterday's labour market statistics were, in general, pretty good, especially the sharply lower unemployment rate, there was quite a bit of partisan argy bargy about what had happened to wages: were they or were they not going up as strongly as you might have expected in a strong labour market? With attendant scuffles over what role, if any, raising the minimum wage had played in it.

As it happens, there's ample scope for the political mischief makers to muddy the waters, because the two main gauges of what's happening to wages are easy to misinterpret, even when there aren't wilful manipulators shouting in your ear.

Here's the relevant bit of Stats' media release.

Annual wage inflation, as measured by the labour cost index (LCI) salary and wage rates (including overtime), increased 1.7 percent compared with annual consumer price inflation of 1.6 percent. Average ordinary time hourly earnings (QES) rose 2.5 percent over the year.

So which is it? Wages only marginally ahead of inflation, or modestly ahead?

The answer is, and I'm sorry if this sounds quintessentially like what an economist might say, it depends on what you're measuring. The two measures that Stats publish - the Labour Cost Index (LCI) and the Quarterly Employment Survey (QES) measure of earnings - do different things.

The LCI looks at the rate of pay for exactly the same job (strictly speaking, we're talking about the "adjusted" LCI here, which is the one used in the Stats press release). If you're a middle manager with exactly the same set of responsibilities you had a year ago, and you're doing it with the same qualifications, the LCI is the measure of the rate for your job. It's a pure "like for like" comparison. On average across the country, on this basis, the pay rate for the same job done by people with the same qualifications is up 1.7% on a year ago.

As a statistic, this has its uses, but it's quite a precise concept, and in particular it doesn't neatly line up with what actually turns up in people's bank accounts from their employer's payroll run, which is probably more what people have in mind when they talk about what's happened to their pay over any given period.

For a start, at the individual level, people get promoted. Or they get a bigger bonus. Or they move to better paying jobs at another place. Or they get more qualifications. Or (on the downside) their employer goes bung and they have to take a job that pays less. Or they jack in the high paying job they've always hated and gone and done what they always wanted to do. Whatever.

There are also all sorts of change happening at the business and industry level, too. "For example" - I'm quoting from p14 of Stats' explanation of the QES here - "average total hourly earnings in the retail trade industry are lower than the national average, and represent about 10 percent of the total paid hours of all industries combined. If the retail trade industry increased total paid hours relative to other industries, the average total hourly earnings for all industries would fall, everything else being held constant, because there is a relative increase in influence from a lower-paying industry". If all the new jobs are McJobs, that probably fits into the measure that most people would have in mind when they're thinking about what's happening to the wages being paid in New Zealand.

So for many purposes - and I'd say for the general citizen's interest in what has happened to pay over time - the QES measure is better going to reflect the messy reality of life. Businesses open and close, opportunities come and go, demand runs hot for some skills while nobody want a bar of others, individual careers thrive or falter. Net net net, after all of this, you get the QES measure of average hourly earnings, which is up 2.5% on a year ago. No, it's not "apples for apples", and if you were a nurse a year ago and a nurse today, the LCI's 1.7% may a closer stab at the true state of affairs. But if you moved to the DHB next door, or got a better paying job in A&E, the QES's 2.5% is likely to be more representative.

As for the impact of the minimum wage - it's there, but it was really small. As Stats calculated,

If the increases that were due to minimum wage had been processed in the LCI as no change, then in the year to the June 2014 quarter:

• all salary and wage rates (including overtime) would have increased 1.6 percent, instead of 1.7 percent