Economist writer H.C. has an interesting Free Exchange post discussing Scott’s “musical chairs” model and British unemployment. Scott also comments. Here is my curve-fitting exercise:
One of the difficulties in working with the UK macro data is that the ONS don’t produce a good quality high frequency time series for hourly wages. Of the available data we have low frequency, high accuracy data in the Annual Survey of Hours and Earnings (ASHE) based on very large sample of employers’ payroll data held by HMRC. At the other end we have high frequency, low accuracy data in the Average Weekly Earnings (AWE) data, produced from a small survey which deliberately excludes a number of workers (notably, I believe, high earners and small employers). I compared the range of various sources at the beginning of last year.
In this post I’ll use the lower quality AWE data, from which we have time series for both regular (ex bonuses) and total weekly pay, and also for average actual hours worked. I’ll also assume that bonus pay is not very sticky. From either pay series we can then calculate average hourly wages by simple division.
The musical chairs model says there is a strong positive correlation between the unemployment rate and the ratio of hourly wages to nominal GDP. I’ll cheat slightly (I did warn you this was a curve-fitting exercise) and use nominal GVA at basic prices instead of NGDP. An increase in nominal GDP which is accounted for entirely by an increase in indirect taxes would not be expected to increase the resources available to hire workers; since we’ve had big VAT shocks in the UK I think this is a reasonable “cheat”.
Here’s the first version of the musical chairs graph, using regular wages to derive hourly wages:
I would say this has done quite well since 2008: though the recovery is predicted slightly too fast, and the correlation is absent for the period prior to 2008.
A second interpretation is to use aggregate employee income instead of nominal GDP/GVA, so changes in the labour share of GDP don’t affect the results:
This is a better fit.
One thing that either of these models will struggle with, if only because of data limitations, is the shift between employment and self-employment. There have been two very large swings in UK self-employment in the period covered above: a 9% rise in 2003 and a 10% rise in 2013/4. We cannot really measure the wages of the self-employed, and as far as I know those workers are entirely excluded from the AWE data – I would not find it surprising that the musical chairs model appears to breaks down for that kind of shock. If there are studies on self-employment and wage stickiness I’d be interested to hear in the comments.
I see some fuss over the wage data (again)… but I’m not convinced, especially since this happened earlier in the year and it was a false “dawn”. Declaring my bias: I want to believe there is still a massive hole in labour supply, either in the form of unemployed workers, or workers not getting enough hours. Hence, we still have a significant output gap, and we can expect to see unemployment fall to somewhere near 5%. Fast wage growth now would be a disconfirmation of labour market slack, so in a sense it is not what I “want” to see. (I also prefer that we’d had a macro policy since 2008 which had aimed for 4% wage growth and avoided large shocks to unemployment.)
Martin Weale and others are citing survey measures of pay settlements. I don’t see any reason to trust that over the ONS data. But the ONS labour market update for 2014 Q3 gave us a spike in the 6m growth rate:
That measure is clearly quite volatile.
The annoying thing here really is the “policy-based evidence-making” by Weale (et al), who has spent the last four years cherry-picking whatever data best supports his preferred policy of higher interest rates. In 2011 Weale told us to look at the GVA deflator, in 2013 the excuse was unit labour costs, and in 2014 the excuse is that he spoke to some business owners who said wages were rising. And by the way in 2014 the GVA deflator is running below 2% y/y and unit labour cost growth is around 0%.
Anyway, here are trends and levels for private sector regular weekly wages:
That tiny spike is enough to warrant rate rises? Really, that’s the best argument there is? We also have the quarterly estimates of hourly earnings, with the update to the “EARN08″ table, although this survey measure excludes very high earners:
Again… there is no “inflation”.
There is an interesting asymmetry in how people read the macro data.
For a given increase in aggregate nominal spending (income) I think it would be generally agreed that “what we want to see” is a higher volume of output and not much inflation. Does anybody disagree? Anybody out there who would prefer the trade-off shifts towards higher inflation and lower output growth? No?
OK. For a given increase in aggregate nominal income (spending) we can consider the same trade-off between employment and wages. I had taken it as given that we had a depressed labour market and so “what we want to see” is that increases in aggregate income will translate primarily into higher employment.
What we have seen over the last year looks quite amazing. Over the year to the March-May 2014 period, hours worked has risen 3.7%. We only have nominal data for Q1, but that showed a 4.1% rise in nominal aggregate labour income. In other words, the increase in aggregate income has translated almost entirely into a higher volume of labour employed and there is no inflation – nominal wage growth is maybe just positive.
Yet this is seen somehow as a bad thing, see, for example the Guardian here, which puzzles me. Do you have a sticky wage model of the labour market, in which AD shocks can raise/lower employment, or not? Is higher employment in 2014 a good thing, or not? These questions have simple answers for this simpleton blogger.
6.5% is a good news story, and let’s hope they keep coming.
For me one of the most important lessons for British economic policymakers over the last six years should be to fear the interaction of micro with macro, supply-side policies with demand-side policies.
I do not think it is a merely a co-incidence that the worst fall in nominal demand since the 1920s occurred at the same time as a supply-side shock (collapse) in 2008. All recessions in British history have been driven by tight money aimed at lowering inflation. Was this time different? It’s not obvious why… CPI rate, September 2008? 5.2%.
And I do not think it is merely a co-incidence that the worst recovery in demand on record has occurred at the same time as inflation has sometimes hovered, sometimes soared above the inflation target. CPI rate, September 2011? This is during the time when the Darling/Osborne austerity drive “sucked demand out of the economy.” That CPI rate in September 2011 was, again, 5.2%.
Here is MPC hawk Martin Weale writing this week:
If wage growth picks up more rapidly than I expect, it will be an indication of inflationary pressure in the economy and Bank rate will need to rise sooner. If wage growth remains subdued, Bank rate should rise more slowly. Because the future is uncertain, we cannot make any promises about where Bank rate will be in a year or two years.
Raising the minimum wage by 3% at a time when hourly wage inflation is 1-2% at best, and is one of the indicators preventing the Bank from screwing up the demand side again… that would surely be an incredibly foolish gamble.
What, exactly, have policymakers learnt from six years of negative supply shocks and disastrous demand-side outcomes? Have we even learnt anything about wages, nominal shocks and employment? It does not appear so.
So sure, let’s try another supply shock. Maybe we’ll get “lucky” and the labour market tightens enough this year that hourly wages pick up, so that a 3% NMW raise doesn’t hurt many more people. Maybe.
Here is the Low Employment Commission report for 2008 (before the recession):
3.18 The decline in the labour market position of young people has been general across the UK. The proportion of young people not in FTE aged 16–21 who were in employment fell in almost all regions between 1998 and 2007, unlike those aged 22 and over who saw their employment share increase in all areas of the UK except London. However, by European standards, young people’s labour market position in the UK is relatively strong.
Good one! Our labour market may be doing badly, but just look at Spain! Those guys are really screwed. They continue:
Given that employment in the UK has been at record levels, it is difficult to explain why young people have not done better in the labour market. Two significant developments in the labour market in recent years have been the increase in the number of people of pension age becoming economically active and the arrival of predominantly young migrant workers from the European Union accession countries.
It’s “difficult to explain”… right. A total mystery. I can’t think what might have caused it, so let’s blame immigrants and old people, those are surely the most “significant developments” in the British labour market in the years to 2008. If anybody does have any better ideas about what happened, be sure to write to Card, Krueger, Dube, etc.
In the UK the spotlight is usually on the weekly wage measure produced by the ONS, Average Weekly Earnings. Scott Sumner tells us to focus on hourly wages. Why? I think the heart of the theory is that the hourly wage is a stickier price. Weekly wages can adjust via a change in hours, or the hourly wage rate.
In this post I’ll take a look at the hourly wage data. The ONS do not produce an “official” time series for mean hourly wages; but we can get to the data by a number of different routes:
- The Annual Survey of Hours and Earnings. This survey should provide the most reliable data for employed workers, and provides the gross mean hourly earnings (amongst others) from a survey of employers.
- The Labour Force Survey. Every quarter the Labour Market Statistics provide an update to a measure of mean gross hourly earnings (table “EARN08″) from the LFS. This data is known to underestimate the mean, it excludes workers earnings more than £100/hour.
- The monthly labour market update provides both an estimate of Average Weekly Earnings plus average weekly hours, again from the LFS. A simple matter of division should give us the mean hourly wage.
- The national accounts, combined with the Labour Force Survey. The national accounts tell us aggregate national wage income. The LFS data tells us total hours worked. From these two we can calculate mean hourly wages. I don’t know of any reason to doubt the LFS hours data. The national accounts do of course get revised. I am not sure how reliable this measure should be.
The ONS does also have an experimental Index of Labour Cost per Hour series. This data is also available from Eurostat as the Labour Cost Index. Annoyingly we are not given the underlying nominal data in either case, only the index level; I will ignore those series for this post.
This is what the four different sources of hourly wages look like:
I was pleasantly surprised that these estimates came out relatively close together; the data from the “EARN08″ table (green line) is as expected an outlier.
Comparing weekly with hourly wages it does appear that a reduction in weekly hours worked in 2009 contributed to the weakness of weekly wages. Here I’ll stick with the LFS data using average weekly wages/hours as the hourly wage:
Similarly the recovery in average weekly hours since 2011 explains why weekly earnings have grown faster than hourly wages.
On the ASHE measure the average annual growth rate of gross hourly wages was 4.1% between 1997 and 2007, falling to 1.5% between 2008 and 2013. Remember that 4.1% figure when you are told silly stories about how globalisation reduced wages in Britain, and remember the 1.5% figure when you are told that inflation is the “real threat”.
Three graphs to try to convince you why slow wage growth is a “good thing” for the UK, given the path of aggregate demand.
Graph one: the change in nominal gross value added since 2008. This is nominal GDP at basic prices not market prices, so factors out changes in indirect taxes such as VAT.
This is very similar for France and the UK.
Graph two: the change in nominal hourly wages since 2008.
The fact that nominal wages increased in both countries between 2008 and 2009 despite sharp falls in nominal demand is a good illustration that nominal wages are indeed sticky. But there is “sticky” and there is “gravity-defying”. French wages are up nearly 12% since 2008 despite a 7% rise in national income over this period.
What’s the result? The “hard-working families” of the UK are actually working hard (relatively speaking; not to deny there is slack in the labour market). In France, not so much.
A little poetic license there… but I think there is an important point here.
In the long run all that matters is productivity. Yes, it would be fantastic if Britain can turn into Switzerland. But Keynesians have a nice turn of phrase about the long run… let’s forget about the long run for a minute.
In the short run, coming out of recession, the biggest difference the government (and I mean you, Carney et al) can make to British “living standards” is to deal with the problem of unemployment and under-employment. In that short run, the growth of real hourly wages is inversely correlated with rising employment and hours worked. Hence I think it is reasonable to argue that falling real wages are a good sign that British living standards are rising because it will be a good sign that employment and hours are rising.
And note again that real wages are not real incomes. Consider a part-timer who goes from 10hrs/week to 20hrs/week at the same time their real hourly wage drops 2%; their real income rises 96%. And then consider the unemployed person who goes from the dole queue to a job. The “real hourly wages” data does not capture a change in either person’s “standard of living”.
This inverse correlation is obviously not generally true. Outside of recessions any supply-side issue which lowers real wages (because productivity falls) is unambiguously bad. But recessions are special.
I couldn’t find a source for historic UK hourly wages, but I did find a couple of studies which imputed hourly wages from the compensation of employees data in the national accounts divided by total hours worked in the labour market stats. Here is the 1990s recession; hourly wages are deflated by the GVA deflator, the best measure of whole-economy inflation:
Now where do you want to argue that British “living standards” were rising? When real hourly wages were soaring in 1991, or when real hourly wages were stagnating in 1994/5?
Maybe this argument is uncontroversial. But if so then why the focus on falling real hourly wages from opponents of current UK macro policy? That is exactly the wrong focus from the perspective of macro policy. If the (political) focus is now purely on the supply-side, on productivity, then Osborne is surely right to claim he has won the argument on macro policy, merely because his opponents have abandoned the debate at the first whiff of positive real GDP growth. Oh, and forget about the unemployed, forget about the labour market… that is the ultimate victory of the inflation hawks.
[Inspired by, and see also: old Sumner posts on real wages in the Great Depression.]