In today’s FT Edward Luce writes about the rise of the robots and the decline of the US middle class.
Luce argues that mechanisation is destroying middle income jobs and outlines an important paradox- more robots are better for growth but awful for middle income employment. He notes that:
The bulk of US jobs growth since mid-2009 has been in low-skilled areas, such as food preparation and domestic aides… low wage jobs (that pay between $7.69 and $13.83 an hour) formed 22 per cent of job losses in the recession but 58 per cent of recovery jobs since then.
As Luce explains US median income is almost 9 per cent lower than when Obama took office and this is as much an economic problem as a political one, as Luce writes:
At some point, policy makers will be forced to grapple with what is intuitively obvious – that sustained growth is inconsistent with declining middle class incomes.
This pattern of job creation in the low waged, low productivity sectors also fits the UK data. As Bill Martin and Robert Rowthorn have noted, in 2010 and 2011 low productivity firms created 550,000 jobs in the UK.
The ‘rise of the robots’ has been extensively debated on the blogosphere in recent months (with Alphaville’s Izabella Kaminska leading the way). The debate so far has mainly focussed on the idea of ‘capital-biased technological change’ – the notion that recent changes in technology have favoured the owners of capital and so far leading to an ever larger share of the proceeds of growth flowing to the owners of capital rather than the workforce. The fact that this is getting high level attention is quite a new thing, as Paul Krugman wrote back in December:
So the story has totally shifted; if you want to understand what’s happening to income distribution in the 21st century economy, you need to stop talking so much about skills, and start talking much more about profits and who owns the capital. Mea culpa: I myself didn’t grasp this until recently. But it’s really crucial.
Much of the previous discussion on income inequality focussed on the role of skills and shifts in income between different sections of the labour force. Whereas economists previously attempted to explain rising inequality by pointing to differing skills levels, this no longer fits the data. (This is the primary reason why I think the pre-distribution debate in the UK has to move away from a focus on skills policy – important as that is).
Whilst there is an important debate to be had about the role of technological change in explaining changing income patterns, I worry that too much attention is being given to the ‘rise of the robots’. It is undoubtedly a major issue but, in the UK at least, it is not the entire story.
Looking at the rising profit share (and the declining wage share) of the economy over the past three decades tells us that something really has changed but it doesn’t give the whole story. A TUC Touchstone Extra published last year (authored by Howard Reed and Jacob Mohun Himmelweit) looked in more detail at what was driving the increased profit share. As I blogged at the time:
In particular they highlight the rise of the financial sector. The chart below shows finance profits (as a percentage of GDP) alongside all profits and all profits-excluding finance.
As can be seen the finance sector’s operating surplus was negligible throughout the 1940s, the 1950s, the 1960s and the 1970s. It began to rise in the 1980s and reached almost 5% of GDP on the eve of the crisis. Over this period a rising profits in the financial sector explain the entire increase in the UK profit share.
Given that the rising profit share has been concentrated in the financial sector, I’m not at all sure that mechanisation and ‘capital-biased technological change’ can provide the full explanation. A potentially fuller explanation, which is a better fit with the empirical data comes from Mariana Mazzucato and William Lazonick a recent paper for Policy Network:
This paper by William Lazonick and Mariana Mazzucato offers a new perspective on why ‘smart’ innovation-led growth has not led to ‘inclusive growth’. It argues that there is a disproportionate balance between the ‘collective’ distribution of risk taking in the innovation process, and the increasingly narrow distribution of the rewards. It proceeds to set out a “Risk-Reward Nexus” framework for understanding the relationship between innovation and inequality, and concludes with key policy implications.
Whether we blame the robots or not, there is a clear trend in both the US and British economies towards stagnating (or even falling) median incomes and job creation in lower productivity sectors. One potential solution comes from Wendy Carlin who notes that what we commonly think of as low productivity sectors can also be thought of as ‘employment-intense sectors’.
We can all become better off as a result of falling prices for the goods and services where productivity growth is fastest. Because of the cheapening of the products of other services, we can more easily afford the products of the stagnant services. And if the stagnant services are financed through taxation, a higher tax burden is inevitable. Again, this is a sign of success not failure of the economy – the higher tax burden is paid for by the higher living standards that come from the higher productivity growth elsewhere.
In other words differential productivity growth need not be seen as a problem – instead society as whole can benefit from rising productivity in some high growth sectors by using the proceeds of this growth to pay for decent services. The problem comes when the benefits of higher productivity growth are not equally distributed. This, sadly, has certainly been the case in the UK over the past few decades:
Ultimately the challenge for policy makers is to find a way of boosting living standards for those in the middle and below and ensuring the rewards from growth are better distributed. But meeting that challenge means first identifying the drivers of stagnant median incomes and diverging rewards from growth.