Future of Work

How Digitalization of the Workforce Will Impact Europe's Social Protection Systems

Artificial intelligence, robots and platform-based work are reshaping labor markets in Europe. The authors of a new report share insights on the changing nature of work and its potential effects on the sustainability of Europe's welfare state.

How Digitalization of the Workforce Will Impact Europe's Social Protection Systems

July 25, 2019

by Guntram Wolff and Enrico Bergamini

The specter of automation can be unnerving—robots taking over jobs, it’s science fiction come to life. A Pew Research Center survey in 2017 found 72 percent of Americans were worried about a future in which robots and computers do many human jobs. Europeans are also worried; nearly three-fourths believe that because of robots and artificial intelligence (AI), more jobs will disappear than will be created, according to a Eurobarometer survey by the European Commission. Interestingly, fewer people feared that their own jobs would be affected; only 44 percent of Europeans thought their job could be done by a machine or AI, and only 37 percent of Americans thought their job was at risk of automation (Fig. 1).

So how worried should we be? In our new report, “Digitalisation and European Welfare States,” which focuses on the digitalization of the workforce in Europe, we uncover initial signs that not all automation may be threatening. But, to harness the positive aspects and prevent the disruption of workers and welfare states, we must plan smartly for these changes.

The crucial question is whether the efficiency gains from the new technologies will, as they did in the past, be realized this time or whether machines really will take away jobs. These technologies could also fundamentally change the nature of work, with far-reaching consequences for European welfare states, which are traditionally financed and organized in the context of traditional employment patterns.

The report tracks the effects on wages and employment of robots and information and computer technologies (ICT). The latter includes such things as big data analysis, cloud computing, early machine learning and software technologies. We also examine the potential impact of the next wave of automation from artificial intelligence and machine learning as well as the possible effects on Europe’s social welfare system from the rise of “gig” or platform-based work and other forms of nontraditional employment. 

Employment Rises but Wage Growth Slows with Industrial Robots

The industrial robotics market is quickly growing, but remains a small portion of the total investments in the economy (0.19 percent). Its advance also appears to be associated with rising employment rates. On average, as the presence of industrial robots in a region expands, the employment rate increases. A marginal increase in industrial robots is associated with a 1 percentage point increase in a region’s employment rate.

However, while employment rates rise, wages often do not. Adding industrial robots to the workforce corresponds with slowing wage growth; an additional robot per 1,000 workers corresponds to a 3 percentage point reduction in the growth rate of real hourly wages. Robots by nature reduce labor demand for specific physical tasks and therefore can potentially reduce wages.

Employment and Wages Both Rise with ICT Adoption

As with industrial robots, greater investments in ICT on the job are associated with increasing employment rates. Unlike with industrial robots, however, ICT is also associated with wage growth. ICT appears to be more complementary to labor tasks, increasing their efficiency and thus increasing employment opportunities. A bank teller may be replaced by an ATM, which lowers costs and allows more branches to open, but those efficiencies also allow more sophisticated non-routine tasks to be rolled out.

The Verdict Is Still Out on Artificial Intelligence

The impact of AI is uncertain. While industrial robots displaced humans working in low-skilled, physical tasks, AI targets cognitive tasks, such as preparing legal briefs or diagnosing patients. As such, its advances might benefit medium- and low-skilled workers by reducing the gap between them and high-skilled workers. This, in turn, could help reduce the job polarization we’ve seen from past technical advances. AI is also more likely to affect the service sector, given its focus on cognitive tasks.

Therefore, it is not at the moment clear that AI will have the same effect as prior waves of automation. If AI systems can provide expert advice that is more valuable to medium- or low-skilled workers, it might benefit them relative to high-skilled workers and reduce the gap between them.

Who Will Pay for the Welfare State as the Workforce Shifts?

These technologies could have far-reaching consequences for European welfare states, which are traditionally financed by taxes on workers. One of the key challenges of the 21st century will be to redefine the nature of welfare states—the way they are constructed and function—in light of the fundamental changes brought about by AI and other aspects of the so-called Fourth Industrial Revolution.

Employment rates in Europe are currently high—good news for a social security system funded largely from taxes and levies on employment. Welfare state spending is a substantial share of gross domestic product (GDP) in the European Union (EU), amounting to 31 percent in France, 28 percent in Italy and 25 percent in Germany.

Yet despite high employment, labor as a share of national income appears to be falling (Fig. 2). This raises the question of who should be taxed to fund the state and the welfare state in particular. Government revenue has remained fairly steady over the last 40 years. Accordingly, the average and the marginal tax rates on capital income (assets such as stocks and real estate) have fallen quite significantly. Overall, labor income and value-added taxation continue to make up the bulk of state finance.

The key question then becomes if and how capital income taxation should be rethought. It is exactly in this context that there is a growing need for a debate on the readiness of welfare states to adapt.

Nontraditional employees, including platform workers and the self-employed, often fall outside current social protection arrangements, which themselves vary greatly in different EU countries. The number of platform workers (those who arrange jobs via online digital app like Uber) is growing as a share of employment. Between 6 and 12 percent of adults in Europe have performed platform work, according to estimates by Pesole and colleagues. Yet, the researchers also find that revenue from such work is still a tiny fraction of labor income; only 2 percent have earned more than half of their income via platforms.

In many countries, unions, insurance companies and other organizations play a role in providing social protections. These arrangements already show signs of adapting to the changing world of work in the information era, a positive development that should be encouraged.

It is clear that the EU, with its advanced welfare states, will be studied carefully around the world. How successful the EU will be in adapting its welfare states to technological change will be the critical topic of the next decades. 

About the Authors:

Guntram Wolff is the Director of the Bruegel Institute and a Global Economy Fellow at the Center for Inclusive Growth.

Enrico Bergamini is a research assistant at Bruegel.