Is Technological Innovation Making Society More Unequal?

, ,

  • 2016•12•21

    Wim Naudé and Paula Nagler

    Society has perhaps never been more unequal than at present, in terms of the distribution of income and wealth. Within-country income inequality (as measured by the Gini coefficient) is, according to the UN Development Programme, “more unequal today than at any point since World War II”.

    These inequalities, and the resulting societal divisions, were one cause of the 2008 global financial crisis — and were, in turn, amplified by it. The economies of the West, the epicenter of the crisis, remain stagnant a decade after the crisis. The political rise of Donald Trump in the United States and the Brexit vote in the United Kingdom — not to mention the rise of nationalism across Europe and the political failures and tensions in Brazil, South Africa, and Russia — can all be argued to be outcomes of, or related to, these inequalities.

    Much has been written about the causes and consequences of high and rising income and wealth inequalities. Two years before the outbreak of the financial crisis, back in 2006, a UNU-WIDER study on the global distribution of private household wealth found that “the richest 10% of adults in the world own 85% of global household wealth; the bottom half collectively owns barely 1%”.

    Are robots to be blamed?

    Why is inequality rising? The main narratives connect higher within-country inequality to institutional and governance weaknesses, especially in the US and Europe. These have allowed policy and regulatory capture by an entrepreneurial elite, resulting in a greater share of GDP accruing to the owners of capital.

    In recent years, though, technological innovation has been receiving increased attention as another possible cause. In particular, information and communication technology (ICT) advances — such as in robotics, automated processes, machine learning, the Internet of Things (IoT), “big data”, and artificial intelligence — are being blamed for making workers (primarily those with mid-level skills in medium-wage jobs) redundant.

    It has been argued that the world is at the start of a “New Industrial Revolution” (e.g., by Peter Marsh in his book The New Industrial Revolution). However, whereas the First Industrial Revolution created thousands of jobs in manufacturing, the new Industrial Revolution (see also “Industry 4.0” or the WEF’s “Fourth Industrial Revolution”) may see humans largely replaced by robots. A much quoted example is of Rethink Robotics who has produced a robot named Baxter, which it claims can replace all humans in routine manufacturing tasks, such as “material handling, line loading and unloading, product inspection, light assembly, sorting, and packaging”.

    Robots are becoming “smarter” over time. Many new models have the ability to learn, and advances in artificial intelligence will only accelerate the value of these robots. As a result, robots are increasingly preferred over humans, especially for routine, difficult/dangerous, or menial types of jobs Indeed, robots are better at work that are “dull, dangerous and dirty”. The “hourly wage” of a robot such as Baxter has been reported to be around US$4.32 per hour, less than the average hourly wage of US$23.32 paid to humans in US manufacturing. Robots already perform about 80% of the world’s automobile manufacturing.

    With autonomous (self-driving) cars likely to be a regular feature on roads within the next decade, the huge transport industry, including taxi and truck companies, will need fewer human drivers. Transport services giant Uber, for example, has ordered 100,000 autonomous cars from Mercedes-Benz.

    Many other industries will be affected, too — at least in terms of the types of tasks performed in a particular occupation. Some jobs, especially those involving routine tasks, are more susceptible to being made redundant by machines. As for the overall impact on employment, Frey and Osborne estimated (in 2013) that 47% of current jobs in the US are likely to be replaced by computers over the next twenty years.

    This trend is not confined only to advanced economies like the US, but also is affecting emerging economies. Frey et al (2016) estimate that up to 66% of all jobs in developing countries are at risk. And developing countries are not only passive players in the New Industrial Revolution; they are increasingly taking part in the production of robotics. The International Federation of Robotics estimates that there will be more robots in China than in the US or Europe by 2017. India is approaching the top ten in markets for sales of robotics. Indeed, according to Sirkin, “the countries moving ahead most aggressively – installing more robots than would be expected given their productivity-adjusted labor costs – are emerging markets”.

    Four decades of ICT

    The question is: To what extent have technological innovations been responsible for the historic rise in income and wealth inequality — especially since the 1970s, when the current upward trend started?

    Numerous authors have noted that the 1970s coincided with the start of the ICT revolution. In 1974, a pizza was ordered for the first time through a computer. By 1989, the World Wide Web was launched. The ICT revolution introduced technologies that complemented the labour of high-skilled workers. Those who were able to use it experienced a productivity and wage increase relative to less-skilled workers, which set off a self-reinforcing cycle: more ICT raised the productivity of skilled workers, creating a higher demand for skilled labour, resulting in more investments in education and, thus, a growing supply of skilled workers.

    The wider availability of skilled workers, in turn, created an incentive for further innovation in ICT, and so on. Where the supply of skills could not keep up with the demand, highly skilled workers commanded a “skills premium” — and wage inequality increased. In economic parlance, technology is not neutral, but “skill-biased”.

    Since the 1970s, the ICT revolution has also led to proportionally more medium-skilled jobs being replaced or automated by computers. This caused “job polarisation”, wherein the demand for high- and low-skilled workers increased, while the demand for medium-skilled workers decreased, resulting in a wage gap.

    What has been the effect of technological innovation?

    Empirical evidence leads credence to both the skills premium and job polarisation concepts. However, the magnitude of these effects has presumably not been large. Various bits of inconsistent empirical evidence suggest that other factors, including globalisation and the “off-shoring” of jobs, changes in the labour market, and the decline of social welfare institutions in the West, have had an even more significant impact.

    First, most of the rise in income inequality has actually occurred at the top 0.1% of the income distribution. In the US, the income share of the top 0.1% of income earners grew by an impressive 324% between 1979 and 2006. This is reflected in the fact that CEO wages rose from a ratio of 20:1 in 1965 to 352:1 in 2007, compared to the average worker compensation in the US.

    Second, in many countries in the West, a large number of highly skilled workers have been experiencing “declining returns to education, less employment security, and shorter job tenure” over the past ten years, and particularly after the global financial crisis. Income inequality, however, has continued to increase — suggesting that it is not an excess demand for highly skilled workers that is underpinning rising inequality.

    Third, in countries where the adoption of industrial robots is most advanced, such as Japan and Germany, inequality is much lower than in many other less-automated societies.

    Finally, in the case of advanced economies, such as the European Union, it remains unclear why average labour productivity has been declining since the 1970s if technological innovation has been so rapid as claimed. In actual fact, investments in ICT have been declining in the recent past in various countries of the Eurozone.

    Weak governance, weak entrepreneurship

    The conclusion from these considerations is that while technological innovation may have contributed to rising inequality, its actual contribution has probably been dwarfed by the institutional-governance weaknesses. The rise in the income share of the top 0.1% of income earners has been fueled by reductions in social security, an unravelling of the power and rights of the labour union movement and collective bargaining, and regulatory capture by the rich. This has resulted in reductions in taxation and transfer payments, and in lower regulations on globalisation and the activities of multinational enterprises.

    Instead of blaming technology, there is actually the possibility that technological innovations can lead to more job and income opportunities, and to more equal societal outcomes. New technologies have the potential to improve economic and social policymaking and result in new business opportunities.

    This, in turn, can result in new markets and new business models, including business models that promote the circular economy, address health needs, and allow older people and persons with disabilities to lead longer and more productive lives. This, however, will require more and better entrepreneurship than ever — which may be a tall order, as entrepreneurship itself is in crisis in many countries and regions.

     – – – – –

    This article draws on a background paper for the UNIDO Industrial Development Report 2016. For the complete paper, see “Industrialisation, Innovation, Inclusion”, UNIDO Working Paper No. 15/2015, Vienna: United Nations Industrial Development Organization.