Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday 22 June 2014

Real wages, monetary policy and innovation

In my drive into Oxford I pass a petrol station that offers a car wash service. Ten or twenty years ago you would have expected this to involve a large degree of automation. However in this particular case it involves a few workers with hoses, mops and buckets. Now anyone who has seen my own car will realise that I know very little about car cleaning technology. But with this caveat, it seems to me this garage offers a nice illustration of how labour productivity is a function of relative prices. If labour becomes expensive relative to capital, it is worth the garage investing in a car washing machine, but if the opposite happens, once the machine reaches the end of its life it goes back to the old labour based technology.

In technical terms what I describe above is just an example of factor substitution. This is one explanation of the UK’s productivity puzzle, investigated at the aggregate level by Joao Paulo Pessoa and John Van Reenen. They “argue that ‘capital shallowing’ (i.e. the fall in the capital-labour ratio) could be the main reason for [the productivity puzzle]”. Although initially the US did not see productivity fall, there are indications a milder form of this may be happening there too.

So why does this not happen in every recession? One answer, provided by Pessoa and Van Reenen, is that the behaviour of real wages in this recession has been very different. Real wages have been much more responsive to unemployment in this recession compared to the recessions of the 1980s or 1990s. Pessoa and Van Reenen suggest this could be the result of a combination of weaker union power and welfare reforms that keep effective labour supply high even when demand is low. You could also add greater availability of cheap labour from members of the EU where unemployment is high.

A counter argument might be that earlier recessions have been caused by tighter monetary policy, pushing up the cost of capital, whereas in the Great Recession interest rates have been at the zero lower bound. However there is evidence that the Great Recession has increased the cost of capital for large firms in the UK, and the impact on small firms will have been even greater. In addition a recession that involves a financial crisis is likely to leave firms feeling particularly reluctant to invest, because any borrowing will involve a long term financial commitment that leaves them more vulnerable. In the past they could have relied on their bank to see them over any temporary cash-flow problems, but now they are less sure. In these circumstances, labour intensive rather than capital intensive forms of production seem much less risky.

Indeed in a world of certainty the capital intensive form of production might actually be more efficient. In economic jargon, switching to people with buckets and hoses has actually reduced total factor productivity. But in a world where financial risk has increased, the firm may still choose the labour intensive form of production.

This process of factor substitution will also lead to a steady decline in survey measures of excess capacity. As the recession hits, the car wash business with a large machine will report excess capacity as people economise on car cleaning. However when the machine reaches the end of its useful life, it is replaced by people with buckets and hoses, and the firm reports no spare capacity.

What happens when demand begins to rise? Initially not much - the firm just hires more labour. Productivity does not increase. The situation becomes more interesting if labour becomes scarce. Does the firm start paying higher wages to attract more workers, or push up prices to choke off additional demand? Or does the firm now think that maybe it is time to invest in a car washing machine, which would in a more certain future allow the firm to reduce costs and prices (and lead to a reversal in the fall in productivity)?

Perhaps all of the above. But suppose that at the moment real wages or inflation begin to rise, the central bank tightens monetary policy. This would raise the cost of capital, and could be interpreted as an attempt to prevent real wages rising. In other words, a strong signal to the firm to stick with its labour intensive production methods. We enter a kind of low productivity, low wage trap. Monetary policy, which in theory is just keeping inflation under control, is in fact keeping real wages and productivity low.

Monetary policy makers would describe this as unfair and even outlandish. A gradual rise in interest rates, begun before inflation exceeds its target, is designed to maintain a stable environment. As the owners of the garage begin to appreciate this, they will eventually decide to invest in that car washing machine. On the other hand, if they sense that inflation might rise above target, they will not invest, however strong short term growth might be.

I’m not sure I believe this. As Chris Dillow argues here, investment may be particularly prone to confidence or animal spirits. Would these animal spirits be stimulated more by strong demand growth, even if it was accompanied by forecasts of 3% or 4% inflation, or by monetary tightening to prevent this inflation ever happening?  

I also have another concern about a monetary policy which tightens as soon as real wages start increasing. What little I know about economic history suggests an additional dynamic. As long as the firm is employing labour rather than buying a machine, there is no incentive for anyone to improve the productivity of machines. The economy where real wages and labour productivity stay low may also be an economy where innovation slows down. The low productivity economy becomes the low productivity growth economy.


  1. “If labour becomes expensive relative to capital, it is worth the garage investing in a car washing machine…” There is problem with that argument, namely that the cost of capital equipment is largely made up of the cost of labour. Ergo it’s not possible to alter the price of labour relative to price of capital equipment. (To be more accurate, the cost any anything is ultimately composed of 3 items: labour, interest and profit I seem to remember. But even interest and profit can be regarded as the “wage” or cost of the labour of money lenders and entrepreneurs.)

    Instead, I suggest it’s an increase in the RANGE of wages paid, e.g. the very low wages paid to immigrants mentioned by Simon. Certainly my experience is that hand car was establishments are run and staffed by immigrants – Middle Eastern and Indian sub-continent in my experience.

    Assuming the above increased RANGE reduces unemployment (which I think it does) we could exploit that phenomenon by implementing a large scale workfare scheme: i.e. hire out the unemployed at say 1p/hr to employers, with employees getting their usual pay in the form of benefits.

  2. Do wages fall below the minimum wage? Or does that 'substitution' occur above the minimum wage?

  3. RM;

    "we could exploit that phenomenon by implementing a large scale workfare scheme: i.e. hire out the unemployed at say 1p/hr to employers, with employees getting their usual pay in the form of benefits."

    Roll up! Roll up! Roll up!

    Watch as the man striving to be treated as a 'Very Serious person' stretches logic past its useful point and advocates a state slavery system. See the invention of modern indentured servitude before your very eyes!

    1. The dictionary definition of a slave is along the lines of someone who gets no wage, apart from enough food to eat plus shelter, plus who is FORCED to work. Workfare employees DO GET A WAGE. Moreover, workfare employees are NOT FORCED TO WORK, an option they’ll go for particularly where they have some other source of income or support (e.g. a spouse whose partner goes out to work).

      Next, the wage that workfare employees get (i.e. the benefits they’d get if they DIDN’T WORK) is little different to what many on minimum wage get. So riddle me this: what’s the big difference between minimum wage work and Workfare?

      And finally, while I’m well aware that about half the population thinks that the money government spends grows on trees, the reality is that money for government spending comes from tax. Thus if one person is allowed to live on benefits, another person is being “forced” to work. So riddle me another question: which of those two is the slave?

    2. some workfare "employees" are paid similar to other low-paid jobs, but we have to bare in mind that those in low-paid jobs are sometimes relying on benefits too. (Unlike like my Dad, who is on a zero-hour contract, and so needs the hours; and could be paid 50 pounds more than me per week. I am on ESA)
      I know that some people like to view us as 'customers' when actually we make up the Labour force. The cunning thing about workfare is that they (im not in one) also have to be looking for paid work as well. Paid work being part of the economy. I think to look at benefits as a wage is to overlook the exploitation inherent in capitalist relations: so damn right you should be taxed.

  4. Yellen has told the media she want to see real wages rise. My guess (and hope) is that she will hold off raising rates until they overshoot on inflation. I won't be that surprised if it doesn't happen given the political push to raise rates now, now, now.

    "Savers" and creditors don't want inflation to eat into their savings and returns and they command immense political power.

    What happened the last time wages shared in productivity in the U.S. in the late 90s - something Obama mentioned on the campaign trail - was that Greenspan allowed unemployment to drift down to 4 percent even as others, like Yellen, wanted to raise rates. Globalization and the internet raised productivity and kept inflation down. But what happened instead was a Tech stock bubble which burst and then morphed into a housing bubble. There was an unsustainable inflation in assets as wealth was redistributed upwards and labor no longer shared in productivity gains.

  5. I have noticed in the last two years or so a garage close to me (west midlands) start employing manual car-washers, and anecdotally I'd say its more popular than the nearby machine washers. They use pressure sprayers.

  6. There is a post at interfluidity from 2012 along similar lines: might be interesting.

  7. No, no, no - not knowing much about car cleaning technology is not a minor caveat. Simply put, automated car washing is generally not as good quality as manual washing. The rise of manual washing probably correlates with an increase in the trust of the 'group of men' approach and is a substitute for having a valet service or doing it yourself.

    The washing machines themselves appear to require a large amount of servicing (people-based task) and also seem to break down a lot, which must affect the rate of return for the owners.

    I'm sure the general thrust of your economic argument stands, but the car washing observation doesn't quite work for me

  8. I read this sort of academic, theoretic argument, and think, well, he could be right, but geez, there are some simple answers that need to be done:
    tax the rich
    short term financial trading tax
    go after off shore sheltera

    I mean, geez, why bother with this complex high falutin stuff when you got easy stuff to do ?

    In a way, you could say that the author is acting as agent of the rich, by deflecting energy away from what we need to do; if we go down the author's road, we will spend endless hours debating the paid toadys of the rich, and nothing will happen

  9. The idea seems to be to stop a wage-price spiral by stopping wage increases - but doesn't this leave out ant productivity gains that might still be occuring?

    For example, if the inflation target is 2% and there is normally a 1% increase in productivity because of education, training, technological innovation and normal levels of capital investment, then wages can rise by 3% and yet companies need only raise prices by 2% to maintain profitability.

    One mans wage inflation is another mans increase in income levels... if wages do not rise faster than inflation, the standard of living/incomes never improve unless people work longer hours...

  10. Related to your point, Calvo, Coricelli and Ottonello (Jobless recoveries) also raise the issue of whether credit constraints favor capital intensive technologies.

  11. Not that it takes anything away from the article, but there's far more water metering now, and less fixed cost water rates. I'm guessing a bloke with a bucket uses less water. And in SE England, a garage probably faces the risk of a machine car wash ban every few years, during which an expensive capital item would be sitting idle.

    So the increase in hand car washing might instead be telling us (a) what happens when businesses are forced to pay for the true cost of the resources they use, or (b) the effect of regulation, or (c) what happens when there is a lack of infrastructure (not building enough reservoirs.

  12. The article deals with the shift in capital intensity of the economy.
    The first argument is related to the relative cost of capital and labor that can be split in four terms: the real price of investment goods, the real interest rate, the financing premium and the real wage. During the crisis, price of investment and the interest rates has decreased (but not enough due to ZLB): these developments should have sustained capital intensity. But at the same time, the financing premium went up (a lot) and the real wage dropped (more than in previous recessions). As a result, medium run capital intensity have been reduced, causing a astonish labor productivity.
    Now, assume a recovery. According to the analysis, capital intensity will depend on the financing premium. One may expect a normalization.
    The second argument is a kind of “low investment trap”: investment is low because the real wage is low; and the real wage is low because capital intensity is low (the story is surely more complex).
    Does this post suggests that the financial premium and the ZLB has pushed the economy in a trap ? This would be exactly a so-called “lost decade”, i.e. a phenomena where an initial shock is amplified / prolonged by an inaccurate policy or the absence of an accurate policy.

  13. Interesting post. You may be interested in the situation in Switzerland, which doesn't seem to fit so neatly into your model. Swiss wages are high, relatively, while the cost of capital is low. Yet when there are road works, you hardly see traffic lights; instead there is a team of men wearing dayglow gear controlling the traffic, 24 hours a day. They aren't completely without machines though; they do use walkie talkies...

  14. Suggest you to read Das Capital volume I by Karl Marx chapter on machinery. You should get a nice flow of idea as well as answers to your many questions quite well.


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.