Labour markets across the globe continue to experience varying pressures, with one major issue being the recent weakness in wage growth. This section of the report looks at the issue of slow wage growth and related labour market trends such as the slowdown in productivity growth since the financial crisis; the decline in the labour share of income; and the impact of automation and machine learning. We explore the wider implications of these trends, and the ongoing labour market policy implications of automation.
Why are real wages growing so slowly?
A key feature of the labour markets featured in the Hays Global Skills Index is the relatively slow growth of economy-wide wages in recent years, after allowing for inflation. This comes despite falling unemployment rates, implying there has not been a large pool of unemployed workers competing for the available jobs. Some downward pressure on wages has been acknowledged from those in part-time jobs who would like to work full-time, but this alone can’t explain the widespread trend for slow wage growth.1
The issue of declining real-wage growth, particularly evident in countries such as the United Kingdom and the Czech Republic (see Fig. 1), is important to understand as it negatively impacts the rate at which each worker’s standard of living improves.
There appear to be two key explanations for this: the rate of growth of labour productivity – the value of goods and services produced per worker – which has slowed markedly since the financial crisis struck a decade ago; and the shrinking share of income that workers have been receiving in recent decades. In the latter case, a principal driver behind this declining ‘labour share of income’ is argued to be advances in technology and the automation of tasks.
The Global Productivity Puzzle
Productivity is a key determinant of wage growth. Unless economies are able to produce more output with a given set of inputs, there will be little growth in firms’ revenues or the real wages they pay.
If we compare growth trends in labour productivity – countries’ output per worker – before and after the financial crisis, we see that growth rates after 2008 remain well below pre-crisis levels. This is particularly the case for our Index countries in Europe, the Middle East, and the Americas (see Fig. 2). This was expected in the immediate aftermath of the crisis because in recessions output falls while firms tend to retain their labour. What was not expected was this slow rate of productivity growth to remain during the subsequent years of economic recovery.
Some of the enduring weakness can be explained by the large drop in both public and private sector investment levels following the crisis, meaning firms were unable to increase the productivity of their workers through the enhancement of the tools and machinery they use. The tighter credit conditions and heightened uncertainty also led to reduced investment in cutting-edge technologies, which is now weighing heavily on productivity growth.2 As a result, many countries appear to be caught in a ‘low-growth trap’, whereby weakness in productivity growth leads to lower investment that, in turn, further dampens productivity.
But while the financial crisis certainly appears to have been a contributor, some studies suggest this global decline in productivity is a longer-term trend, influenced by factors such as ageing populations globally; a downturn in global trade; less investment in education and training; and a slowdown in technological advancement. Despite the apparent flood of new digital and mobile technologies, there is an argument that productivity improvements from information and communications technologies (ICT) have not matched the exponential gains associated with the introduction of the steam engine, electricity or the telephone in earlier eras.3
Another factor highlighted by one recent study is that there has been a slowdown in technology and knowledge ‘spillovers’ from the world’s most productive (‘frontier’) firms to others within their sector (‘laggards’).4 Global firms including the likes of Amazon, Uber and Microsoft can more easily upscale their operations, expand internationally and generate synergies to create dominant positions in the market, whereas smaller firms are less able to do so. These are the firms that have experienced very weak growth in productivity.5
Finally, the increasing importance of the digital economy may also have led to productivity growth being underestimated in certain countries, as many information and communication technologies are not captured adequately in overall assessments of firms’ contributions to GDP.
Explaining the decline in workers’ share of income
While poor growth in labour productivity has weighed heavily on wage growth recently, there appears to be another, longer-term factor dragging on wage growth in many countries: a fall in the share of national income that goes to workers. The International Monetary Fund (IMF) estimates the share of national income going to labour in advanced economies has declined from a peak of 55% in the early 1970s to 51% in 2015 (see Fig. 3). In emerging markets and developing economies, it is estimated to have fallen from 39% in 1993 to 37% in 2015.
Some suggest this ‘decreasing share of the pie’ going to workers could be driven by globalisation, particularly in advanced economies, where many firms have moved labour-intensive tasks overseas to reduce costs (known as ‘off shoring’). This has made production processes more capital intensive, and thus reduced workers’ bargaining power.6 The dominance of ‘superstar’ firms such as Google and Apple may also have contributed as these companies have very high profits while hiring few, albeit well remunerated, employees.7
But it is widely argued that, prior to the financial crisis, technological advances and automation were the main contributors to this global decline in the labour share of income, because improvements in technology lead to faster productivity growth in capital, lowering the price of capital goods relative to labour, and therefore encouraging firms to substitute capital for labour.8 While the labour share has been broadly flat since the crisis, recent advances in AI and machine learning suggest the labour share may begin to fall again in future.
Robots or jobs?
Amid weak productivity growth and a declining labour share in many parts of the world, what does the future hold for workers? One of the most disputed labour issues is the advent of AI and robotics: on the one hand, this could herald a productivity revival that boosts wage growth, while on the other, it may threaten jobs throughout the economy.
Advancements in AI enable businesses to automate many jobs that have been historically undertaken by humans, which may have a negative impact on new job creation in the short term.
In the longer term, however, economic theory and the available evidence suggests there will be little impact on aggregate employment levels because technology cuts the cost of producing goods and services, which in turn will raise the purchasing power of consumers, whose extra spending creates new jobs. One recent study examining the impact of automation across a number of European countries, as well as Australia, Japan and the United States, found it has had a positive effect on aggregate employment.9
In addition, a recent study concluded that around 4.3 million workers in the United States will be displaced over the next decade due to the impact of automation. However, the report also found that the rise in prosperity caused by these new technologies would see millions of new jobs created elsewhere in the US economy. In short, while technological change will cause dislocation in the labour market in the coming years, new employment opportunities will also emerge and the main challenge for policymakers is to ensure people have the skills to take advantage of them.10
Of course, such a transformation of the global labour market will cause unsettling disruption for many workers. Further improvements in technology will likely lead to a ‘hollowing out’ of jobs distribution, whereby some middle-skilled jobs disappear and more jobs are created in both lower- and higher-skilled occupations. Indeed, relatively rapid change along these lines has been underway for many years: between 1995 and 2015, the largest decreases in middle-skilled jobs were observed in Austria, Switzerland and Ireland.11
The fast-changing workplace means businesses and policymakers face important choices. Adapting to a new workforce composition in the face of rapid technological change is challenging, as workers’ productivity and capacity to transition into different occupations will depend on their ability to develop new skills. Education and on-the-job training should therefore be a priority, enabling businesses and workers to boost their productivity, which will in turn help workers to retain good-quality jobs and boost wage growth.12
In addition, active labour market policies, such as publicly-funded vocational training schemes or job-search assistance, can improve employability and expand opportunities available to jobseekers.
Different countries have different attitudes concerning such policies, however. While a number of countries featured within the Index, including Denmark, Sweden and France, invest significantly in these labour market programmes, others, such as Mexico, the United States and Japan, exhibit very low spending as a proportion of their overall GDP.13
Either way, governments will need to think about how to best provide assistance for workers who find themselves displaced. More radical policies that have been suggested to mitigate the negative impact of automation include the introduction of universal basic income (UBI) and job guarantees, whereby the government becomes the employer of last resort.14