Wages, Compensation & Growth
I’ve already blogged on the Treasury analysis of pay, productivity and total employee cost. As I said earlier this week, I don’t find it especially convincing but the debate is rumbling on and is a very important one.
So there you have it: if the Treasury numbers stack up, the biggest reason for subdued pay in recent years – apart, of course, from a horrible drop in productivity – is higher taxes and pension contributions. It’s worrying – but certainly not as existentially so as most of us had feared.
The thing is – I’d actually like to believe this. If the reasons for pay stagnation really were that straight forward then the solution would be equally straight forward. But I think the most important word in the above paragraph is ‘if’ and I’m not convinced the Treasury numbers do stack up.
So, to recap and expand on the earlier blog, here are my reasons to doubt the Treasury analysis – or at least to say it is not the full story.
First – even on their preferred measure of comparing output per worker to real compensation per employee, there appears to be a gap if we look at the period 1992-2007. I think this is the period to look at – one of economic growth, so we can exclude the impact of the recession from the analysis. After all if we are examining the question of has the link between productivity and earnings been broken, then is makes more sense to look at a period of growth.
Over this period, output per worker grew by around 8% more than real compensation.
A big gap opened up in the early to mid 90s, closed in the late 90s and then started to open again.
In favct the current situation looks even worse than the early 1990s – in that recovery real earnings rose for a while and then stagnated – this time they are still falling.
Second – this is (possibly due to the data used) a mean rather than a median analysis. That may sound a bit techy but it matters. A mean measure is liable to distortion by soaring high pay at the top of the distribution.
We know from previous ONS analysis that median measures show a different pattern, As they noted on Monday:
The mean simply divides the total income of households by the number of households. A limitation of using the mean for this purpose is that it can be influenced by just a few households with very high incomes and therefore does not necessarily reflect the standard of living of the “typical” household…
Median equivalised disposable household income more than doubled between 1977 and 2011/12. In 1977, the median was the equivalent of £11,200 in 2011/12 prices. In 2011/12, median household income was £23,200, having grown at an average rate of 2.2% per year over the intervening period.
Over this same period, mean household income increased more quickly than the median measure, growing at an average annual rate of 2.4% from £12,600 to £28,200. The faster growth of the mean measure was primarily due to incomes of high-income households growing at a faster rate than incomes in the middle and lower parts of the income distribution between 1977 and 1990.
So, if we really want to measure if the earnings of the typical household have become disconnected from growth – we need to look at medians rather than means.
Third – some of the interpretation around the Treasury numbers strikes me as misguided. Quite a few people seem to think this story is mainly about increasing employer National Insurance Contributions (NICS) payments. It isn’t. Between 1992 and 2012 employer NICS as a percentage of the wage bill rose from 6.22 to 7.22%, the big change came in ‘funded pension schemes’ which rose from 3.39% to 7.19%.
Now, for a start calling this ‘real compensation’ is questionable – it’s not as if we are experiencing some post-2000 ‘pensions bonanza’ that makes up for flat wages. Whilst there are legal & valuation changes that will have increased this amount, there is also the performance of the stock market post-2000 and underfunding of schemes in the 1990s.
But to test if this analysis is correct – i.e. that an increase in pension contributions is responsible for weak wage growth – then one would need to compare firms that run a scheme to those that don’t. Without such a test, I’m not sure we should read too much into this analysis.
Finally – this is a rather parochial analysis. It looks just at the British experience but we know that the median wage squeeze is afflicting many developed countries to different extents. Various international bodies from the IMF to the OECD have blamed different factors – declining labour bargaining power, globalisation and technological change being the most commonly identified. Te Treasury at least needs to consider these arguments before zooming in on the politically convenient explanation that it’s ‘all about NICS and pensions’.
I don’t rule out rising employment expensive as one factor behind some of the weakness in wage growth but it feels like quite a stretch to say it is the driving one.
This might seem like a fairly academic debate but it is one with very serious implications. We’re essentially debating the really big question in modern macroeconomics – have ordinary families become detached from the proceeds of economic growth?
I think they have and this means we need fundamental economic reform. The Treasury seems to think they have not and so the only any recovery will eventually raise median living standards.
Time will tell who is right.