Winner of the New Statesman SPERI Prize in Political Economy 2016

Wednesday 9 August 2017

Real wages are mainly a macro issue

What do I mean by this? Macroeconomists have many faults, but one clear positive is that we think about systems as a whole rather than just one particular component. One area where it is important to do this is in thinking about what determines economy wide real wages. Take, for example, this recent post by the Flip Chart Rick. (His posts are brilliant and I try and read every one, but unfortunately this one is too good an illustration of the problem I have in mind.) He starts with this chart from the FT that I reproduce below.

Why is the UK unique in having a combination of negative real wage growth but positive GDP growth? Now it just so happened that I had written a post about this, explaining I thought pretty well the key reasons. But Rick mentions none of these, but writes about a whole bunch of stuff related to labour market structure and trade union power that I think are largely irrelevant. I think he is making the same mistake that people make when they say immigration reduces real wages, or that we would all be better off if only unions were more powerful.

All these things are important in influencing nominal wages, and perhaps the distribution of wages between workers. But real wages also depend on prices, which are set by domestic or overseas firms depending on where goods are made. If nominal wages go up, prices are likely to go up.

So what do I think accounts for the fall in real wages in the UK over the last decade? We need to start with GDP per head rather than GDP: growth in the latter has been boosted by immigration. Here is what has happened to GDP per head over the last ten years.

GDP per head fell in the recession, and then steadily but slowly recovered: the slowest recovery in at least a century. To see how that is related to real wages (using ONS average earnings divided by the CPI), which I call real consumer wages, we first need to look at an intermediary measure: real product wages. These are real wages divided by the price of UK output: the GDP deflator.

This is an interesting measure because its closely related to a simple identity relating GDP to labour income and profits. We can see that real product wages have not changed very much over this period: the recession mainly hit profits, or it created unemployment. (Real wages are wages divided the number of workers, GDP per head is GDP divided by the total population, which includes the unemployed.) But if we are comparing 2007 with 2015, real product wages were as stagnant as GDP per head.

So why did real consumer wages fall? That must be because consumer prices rose more than output prices. There are two reasons why this happened in this case: indirect taxes increased (remember the 2011 VAT hike), and a large sterling depreciation during the GFC worked its way into higher prices for imported goods. It is of course another depreciation after the Brexit vote that is cutting real wages once again right now. As I always try and stress, real GDP growth per head is not a good guide to real income growth if the price of imported goods rise or the price of UK goods sold overseas falls (what economists call a decline in the UK’s terms of trade).

Real wage growth in the UK has not been lousy because of lack of union power, immigrants or higher profits, but because economic growth (properly measured) has been stagnant, austerity included raising indirect taxes and we have now had two large depreciations in sterling. [1] That is not to say that these labour market factors are not important. At a macro level they are important in keeping inflation low, which should have allowed a more rapid expansion of GDP growth than we have actually had. That is where fiscal austerity and Bank of England conservatism come in. At a micro level labour market structure helps influence the distribution of earnings between different labour groups. [2]

What I say about the unimportance of profits is factually true for the UK over this period, but it is not always the case. In the US and elsewhere we have seen a gradual shift from wages to profits over the last few decades. But even here it is not obvious that weak nominal wage growth is the main cause, because in a competitive goods market lower nominal wages should get passed on as lower prices. One explanation that is attracting a lot of interest is the rise of superstar firms. These firms make unusually high profits, or equivalently have low labour costs, and if output is shifting towards these firms labour’s share will fall. What these firms do with their profits then becomes an important issue. More generally, it may be the case that governments have become too lax at breaking up monopolies, allowing a rise in the overall degree of monopoly.

The consequence of growing concentration, superstar firms and a rising share of profits is that income derived from profit grows faster than income from labour. I say derived from profit because I would include in this CEO and financial sector pay, which in effect extracts a proportion of profits from large firms. The net result is that most of the proceeds of economic growth are going to those at the top of the income distribution. But it would be good if we could change that by making the goods market more competitive and removing the incentive for CEOs to extract surplus from firms [3], rather than by making the labour market less competitive.

Technical appendix

For those who are lucky enough to have learnt economics using the Carlin and Soskice text, this is a classic application of wage and price setting curves. If workers become weaker, this shifts the wage setting curve towards the (perfect competition) labour supply curve, reducing the equilibrium real wage (unless the price setting curve is flat) but increasing the equilibrium level of employment. An increase in the degree of monopoly (the mark-up) shifts the price setting curve further away from the perfect competition labour demand curve, which reduces equilibrium employment as well as the real wage.

[1] One possible caveat here is that low wage growth may have encouraged firms to use more labour intensive production techniques, which has depressed investment and productivity. But if we want to incentivise firms to invest in more productive technology, increasing demand is a much better method than increasing nominal wages.

[2] Another caveat. I'm not sure where the real wage data in the FT chart comes from, but the fall in UK real wages there is greater than you get by using the ONS average earnings data (which I have used), so it may be a different and more specific measure of real wages. In which cases labour market structure might be relevant in explaining that number, and I apologise to Rick in advance if that is what he had in mind. 

[3] By, for example, applying much higher tax rates on high incomes, or imposing a maximum wage.  


  1. Man, I always enjoy your posts and find them thought-provoking and informative, but this one in particular has turned my crank. I'm not trying to fluff you, but there is so much negative crap being slung around the internet these days, I simply wish to express positive energy when I feel it. Thanks for your work!

  2. I thought a supply of cheap labour (from weak unions and plenty of migrants) meant that firms don't need to innovate with new automation and more efficient use of labour. That lack of innovation and progress lies behind the phenomenon you describe:- "economic growth (properly measured) has been stagnant". If a firm has no choice but to use scarce and expensive workers, then they will be forced into investing in ever more efficient machines and working practices and economic growth follows on from that.

  3. I would echo the comment above: I found this a very interesting and thought provoking post.

    There is one or two other factor that may be relevant: the relative decline of the manufacturing sector versus the service sector and whether service pay generally is less than that in manufacturing and also the shift into self employment, increasingly important economically but very under reported statistically.

  4. One thing I can't fathom about macro economics is when GDP or average wages are used to explain the state of a nation where inequalities are a given.
    What would happen to these figures of the top 5% of GDP or wages (or anything) were removed?
    I understand you're introducing bias, but is there another way to allow for mean incomes being distorted by London pay, or for financial conglomerates (who contribute less to the national economy, due to their location and product) from distorting GDP.

    Thanks though.

  5. One thing I can't fathom about macro economics is when GDP or average wages are used to explain the state of a nation where inequalities are a given.
    What would happen to these figures of the top 5% of GDP or wages (or anything) were removed?
    I understand you're introducing bias, but is there another way to allow for mean incomes being distorted by London pay, or for financial conglomerates (who contribute less to the national economy, due to their location and product) from distorting GDP.

    Thanks though.

  6. Real wages aren't as much a macro issue as they are a government policy issue. I.e., real wages go nowhere because government changes the relative power of the buyers and sellers of labour.

  7. JW-Mason has a post (with lots of links) that goes into how and why a lack of union power and a slack labour market could be behind that lack of economic growth "Abundant labor and low wages discourage investment in productivity-raising technologies. As Avent notes, early British and even more American industrialization owe a lot to scarce labor and high wages." - I'm also thinking that a knock-on effect could be that less capital intensive work could further erode worker's bargaining power. If say a warehouse is largely automated with very expensive machines sorting the goods and a few highly trained workers installing and maintaining the system, then those workers will be in a strong position to bargain for good pay and conditions (since so much has been invested both in their training and in the machines). By contrast, if the system is run as the Sports Direct warehouse, with thousands of untrained workers running about, man-handling the goods, then they have no bargaining power at all. They are readily replaceable; even if the warehouse was brought to a standstill by a strike, little has been invested in it and another one could spring up with a new set of workers.

  8. From a pure macro view I would expect real wages to be at least somewhat affected by immigration.
    - As migrants bring usually neither a lot of capital nor natural resources with them, migration leads to an on average lower capital intensity and correspondingly lower productivity. Lower productivity corresponds in first order with lower real wages.
    - migrants usually have a skill set, which is less useful for the country they move to than the local population, which again reduces average productivity. This doesn't necessarily mean, that the real wages of the native populations shrink, but in the real wage estimate, the lower wages of the migrants themselves are used within the calculation, so the real wages of the total population can shrink, even if the native population sees wage increases - in a country with a very small wellfare state the real wage development might not be very relevant for any political argument.
    - as you already noted, your arguments largely hold in a perfect market situation without any monopoly or oligopoly power. However, in the real world, there is some such power and indeed unions help and large migration is an impediment for any attempt by workers to get a share of the monopoly gains.

    Of course for the UK, as you demonstrated, the dominant source for the lousy real wages development are the terms of trade, but there are situations, where immigration and lack of unionisation can be the source for low real wage development. In a political context of course real wage development for individuals is imporant rather than for the population in total, where both unions and migration can have larger positive and negative impacts than on the total real wage/capita.

  9. On 31 July Philip Hammond, seemingly allaying fears that the UK was going to make its living post-Brexit by becoming a tax haven, said there would be no cuts to taxes and no cuts to regulations after March 2019. "Number 10" made no response to this, although that might have been because Theresa May was on holiday.

    For the lower-orders becoming a tax haven means low wages, weak unions/workers' rights and minimal public services. For the upper-class there are low taxes and lax financial regulation (which is code for "easy tax avoidance and money laundering").

    So why would Hammond make things harder for Tory donors? Has he been threatened by the EU that they will make life difficult for the UK if it goes down the tax haven route?
    Or has he realised that if the Conservative policy is to cut regulations then people might ask: "What regulations are you going to cut? Would they include fire regulations on, say, the type of cladding that goes on tower blocks ?"
    And would that policy be a vote-loser? And was the silence from "Number 10" an indication that they too had realised that ?

  10. I think you are making the mistake of seeing competition as a driving force for business. It is really a nuisance at best and the reality is that the main driver is profit. I like to illustrate it with this question: if a company makes a product for £1 and sells for £10, then when another company comes along and can also make the product for £1, then which of the following is more likely. a) that it sells the product for £2, or b) that it sells the product for £9? Modern business is all about maximising profit rather that outselling the competition.

    1. I agree with Richard Koo that we are in a balance-sheet recession and business is not in profit-maximising mode any longer. Instead, it is trying to minimise debt, because the private sector is drowning in its own debt.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.