Como as sociedades reproduzem, geram e incorporam as tecnologias avançadas? Como fica o emprego diante da automação crescente na indústria e serviços? Espera-se mais ou menos oportunidades de trabalho? De maior ou menor qualidade?
As sociedades estão preparadas para lidar com eventuais impactos negativos gerados pelas novas tecnologias? Pesquisa realizada na Universidade de Oxford (Reino Unido) traça um perfil nada animador para o emprego.
TECHNOLOGY AT WORK v2.0
The Future Is Not What It Used to Be
The debate over the impact of automation has been going on for several centuries, from the Luddites in 1811 to John Maynard Keynes’ prediction of technological unemployment in the 1930’s, to Stephen Hawking’s recent warnings over Artificial Intelligence. Most recently, the topic has been chosen as the primary theme for the 2016 World Economic Forum meeting in Davos and the Financial Times awarded their 2015 Business Book of the Year to Martin Ford’s ‘The Rise of the Robots’.
In our February 2015 Citi GPS report Technology At Work we cited three primary reasons why we believed the impact of technology change on the economy was different this time: (1) the pace of change has accelerated; (2) the scope of technological change is increasing; and (3) unlike innovation in the past, the benefits of technological change are not being widely shared — real median wages have fallen behind growth in productivity and inequality has increased. In the report that follows, Citi again teams up with Carl Benedikt Frey and Michael Osborne from the Oxford Martin School to answer a number of the questions that were generated post the original report – on a range of topics from susceptibility to solutions.
There seems little doubt that the pace of technology change has accelerated. Whereas it took on average 119 years for the spindle to diffuse outside of Europe, the Internet spread across the globe in only 7 years. Going forward, as argued in the Citi GPS Disruptive Innovations III report, the cost of innovation continues to fall as cheaper smartphones will help bring 4 billion more people online. The next stage of connectivity will move from people to ‘things’ with Cisco estimating 500 billion devices will be connected by 2030, up from 13 billion in 2013. Increasing digital connectivity is fuelling a data boom, with data volume estimated to be doubling every 18 months, and computers are likely far better able to handle this volume than people. In 2015, the Edelman Trust Barometer found that more than half of the global “informed public” believe that the pace of development and change in business today is “too fast” with 70% citing technology as the driver of change. Over 96% of institutional clients who participated in Citi’s survey on technology and work believe that automation will accelerate over the next five years vs. the previous five years.
Our work also suggests the scope of technology change is increasing. The big data revolution and improvements in machine learning algorithms means that more occupations can be replaced by technology, including tasks once thought quintessentially human such as navigating a car or deciphering handwriting. Carl Benedikt Frey and Michael Osborne’s original study on employment2 suggested 47% of US jobs were at risk of computerisation. In the chapter ‘How Susceptible are Countries Worldwide?’, we use data from the World Bank to show that the risks of automation are actually higher in many other countries —- for example in the OECD the data shows on average 57% of jobs are susceptible to automation, this number rises to 69% in India and 77% in China.
Increased automation in low-wage countries, which have traditionally attracted manufacturing firms, could see them lose their cost advantage and potentially lose their ability of achieving rapid economic growth by shifting workers to factory jobs. In a survey that we conducted with institutional clients, 70% of our clients also believe that automation and the developments in 3D printing will encourage companies to move their manufacturing process closer to home — with North America gaining the biggest advantage from this development and China having the most to lose. A growing concern of ‘premature deindustrialisation’ in emerging and developing countries could require new growth models and a need to upskill the workforce.
Historically, new technologies have not only transformed regions, industries and companies, but also cities. In the US, the computer revolution has shifted the fortunes of many cities with some, such as San Francisco, experiencing rapid growth while others spiral towards bankruptcy. Cities that specialised in cognitive work gained a comparative advantage in new job creation, mirroring trends in population and wage growth over the same period. The expanding scope of automation in the 21st century may shift the fortunes of US cities once more, affecting a different set of cities than the ones historically impacted by computerisation and offshoring. Berger, Frey and Osborne in their most recent analysis3 suggest that the exposure of cities in the US to future automation ranges from 54% in Fresno to 38% in Boston. The best way forward for cities to reduce their exposure to automation is to boost their technological dynamism and attract more skilled workers.
The chapter ‘What Jobs do We See Ahead’ emphasises the importance of acquiring skills for future employment. The European Centre for the Development of Vocational Training (Cedefop) estimated that in the EU nearly half of the new job opportunities will require highly skilled workers. Today’s technology sectors have not provided the same opportunities, particularly for less educated workers, as the industries that preceded them.
This downward trend in new job creation in new technology industries is particularly evident starting in the Computer Revolution of the 1980s. For example, a study by Jeffery Lin4 suggests that while about 8.2% of the US workforce shifted into new jobs during the 1980s which were associated with new technologies; during the 1990s this figured declined to 4.4%. Estimates by Thor Berger and Carl Benedikt Frey further suggest that less than 0.5% of the US workforce shifted into technology industries that emerged throughout the 2000s, including new industries such as online auctions, video and audio streaming, and web design.5
Despite these low numbers to date, a number of forecasts suggest good opportunities for future job creation in the information technology, industrial (i.e. robot engineers and technicians) and green sectors. In addition, the health sector is set to create the largest job openings, estimated at more than 4 million new jobs in the US from 2012 to 2022. This is not surprising given that many people in advanced countries are living for longer. This, together with a reduction in fertility rates, is changing the demographics in many advanced and some emerging nations (e.g. Japan and China) as described in the chapter ‘Demography vs Automation: Will We Run Out of Workers or Out of Jobs?’ In these countries, automation is seen as a possible solution to the changing demographics, however we conclude that automation could pose more risks to jobs than demographic changes.
Incremental productivity gains through automation should offset some of the demographic developments; however we are currently also facing a ‘productivity paradox’ as described in more detail in the chapter ‘Impact of Automation on Productivity’. If, as we believe, progress in technology is so rampant, we should see healthy productivity improvements. However across advanced economies, labour productivity growth has slowed from 4% in 1965-75, to about 2% from 1975-2005 and further lower to 1% from 2005-2014. Rather than taking the view that the ‘lowhanging fruits’ of innovation have been plucked, we see limitations in the current measurement of productivity, a lag effect in measurement and a wider distribution of productivity across firms and workers. We found that institutional clients agree — 95% of clients in our survey expect automation and technology will drive an increase in productivity growth to some degree over time (60% to a significant degree and 35% to a minor degree).
While we are positive on productivity, the chapter ‘Impact of Automation on Inflation’ argues that the rise in automation and technology is likely to reinforce the current low inflation environment due to increased uncertainty and the narrow distribution of productivity gains. There is also good reason to suspect that current price deflators and measures of inflation are overstated. We may already be observing some of these effects given the Phillips curve has flattened. To close output gaps, demand stimulus will continue to be needed, with monetary policy remaining accommodative and more creative in most advanced economies.
Whilst the first half of the report discusses the effect that automation could have on countries and cities and on productivity and inflation, in the second half of the report we look at changes that are already occurring in a number of industries. We highlight a number of sectors such as the food industry, call centres and factories that are already automating a number of different processes which could ultimately affect (or are already affecting) a number of different jobs. In addition, with the cost of robots expected to decrease over time, economic barriers to the expansion of automation in different sectors are expected to fall. This is discussed in detail in the chapter ‘What are the Barriers to Automation?’ which shows that payback periods for some industrial robots would reduce to 1.2 and 2.6 years in China and Thailand respectively by 2017. Barriers still exist for small and medium-sized enterprises (SME’s) however the flexibility and affordability of co-bots could offer appropriate solutions for these companies.
Looking at the macroeconomic and microeconomic effects of technological change, we find further evidence that the pace, scope and benefits of this change will be a challenge to societies. The remainder of the report focuses on potential policy options. The chapter ‘Response: Policy and Job Risk from Automation’ details active labour market policies which could help people find jobs: from training, to earned income tax credits (EITCs) or lowering tax wedges, to incentives to support self-employment. We believe EITCs are more appropriate than current moves to raise minimum wages given the latter could lead to higher unemployment. If selfemployment is becoming a new norm, this also has policy implications — 34% of the total US workforce, or 53 million people, are currently freelancers (see Figure 34). Some argue a basic income or benefits (healthcare, housing, pensions) should be provided, yet we fear changes in taxation that may be required to pay for solutions will impede an effective policy response. At the same time, local, regional, and state governments could face challenges in maintaining tax revenues from income disparities and from the difficulties in effectively taxing the digital economy.
With technological change a key driver for the demand of both the level and type of skills needed ahead, it is no surprise that the number one policy suggestion in our client survey was for increased investment in education. Although it is clear that more information & communication technology (ICT) skills will be needed, significant differences currently exist in IT skills between countries, with the US and the UK lagging behind other countries on some measures. While many argue science, technology, engineering & math (STEM) skills are needed, intermediate level skills in STEM seem to be a riskier educational investment. Non-cognitive skills can be increasingly important, but malleability of these skills should not be assumed. These factors could complicate the ability of education to adapt to the pace and scope of technological change described above.
With careers likely to be more disrupted than at any other point in the past, individuals should anticipate the need to retrain in the future. A talent mismatch already exists in many countries, with many well-educated workers finding employment in lower-skilled jobs. To combat this, greater coordination will be needed between the educational, training and employment sectors.
One of the paths involving the least commitment for policy makers to adapt to accelerated technological change would be to look to shorter working weeks. This is starting to take place in some countries; for example Sweden is moving to a standard 6 hour week day with the aim of improving productivity. Annual hours worked per person have already fallen by 35% in France from 1950-2014 and by 10% in the US. Leisure time in the US has increased by 6-8 hours per week between 1965-2003 for men and 4-8 hours for women. When we add in longer lives and retirements, cumulative lifetime leisure time has increased 26% from the 1930s. Could automation increase leisure time further whilst also maintaining a good standard of living for everyone? The risk is that this increased leisure time may only become a reality for the under-employed or unemployed.
When we polled opinions on many of these issues in the third quarter of 2015, 85% of respondents felt automation posed a challenge to societies and policymakers, of which 64% said it was a major challenge. However, and encouragingly, 76% also said they were techno-optimists on the outlook for productivity and profits, with policy adapting to share increasing abundance. Only 21% were techno-pessimists on the outlook for growth, employment, inequality, and disruption of company profit pools. We agree with the optimists that the opportunities for increased innovation and productivity can be very beneficial, providing policy, education and workers can adapt to the challenges faced from accelerating technological change.
1 The authors are very grateful to Andrew Pitt and Professor Ian Goldin for guidance in framing this report, as well as to Anushya Devendra at the Oxford Martin School for advice and editorial support
2 Frey, C. B. and Osborne, M. A. (2013). The Future of Employment: How Susceptible are Jobs to Computerisation? Oxford Martin School Working Paper No. 7.
3 Berger, T., Frey, C.B., Osborne, M., “Cities at Risk”, Oxford Martin School Working Paper
4 Lin, Jeffrey, “Technological Adaptation, Cities and New Work”, Review of Economics & Statistics, May 2011, Vol. 93, No. 2, pp. 554-574. While the estimates by Lin are not necessarily directly comparable, they speak to further studies showing a downward trend in new job creation.
5 Berger, T. and Frey, C.B., “Industrial Renewal in the 21st Century: Evidence from U.S. Cities”, Regional Studies, forthcoming.
Você pode ver e baixar o relatório completo aqui.
Deixe seu comentário