Innovation, Technology and Inclusive Economic Growth:

Innovation, Technology and Inclusive Economic Growth: Presentation to the Queen s International Institute on Social Policy DATE

Overview 1. Tension Between Technological Innovation + Inclusive Economic Growth 2. Impact of Automation on Canada s Labour Force 3. Geographic and Regional Implications 4. Tasks and Skills At-Risk and In-Demand 5. Implications and Recommendations 6. Appendices 2

Part 1) Tension Between Technological Innovation + Inclusive Economic Growth Innovation drives long-run economic growth and prosperity. As Canada s resource industries falter and our population ages, it will be increasingly important to harness innovation to boost Canada s productivity and global competitiveness. The benefits of innovation are unevenly distributed. For example, new technologies could make many existing jobs obsolete, wealth creation through innovation is often highly concentrated, and new models of gig-based work could impact job security and benefits. This can create and exacerbate existing inequalities between the high and low skilled, urban and rural/remote communities, different industries and different demographic groups. To be a truly inclusive economy, Canada must simultaneously embrace innovation, help individuals successfully adjust to disruption, and broaden the benefits of an innovation economy to a wider swath of individuals. 3

Technology: an Engine of Growth In 2016, for the first time ever, the top 5 companies in terms of market value were all tech Apple, Alphabet, Microsoft, Amazon and Facebook. 4

Technology: an Engine of Growth Technology investments are contributing more to productivity gains than labour. 1.8 Capital and Labour Contribution to Productivity, 1961-2015 160 Capital Stock Growth Canada, 1961-2008 1.7 140 1961=1 1.6 1.5 1.4 1.3 1.2 1.1 1 Contribution of capital intensity to labour productivity growth Contribution of labour composition to labour productivity growth 120 100 80 60 40 Information and communication technologies machinery and equipment Non-information and communication technologies machinery and equipment 0.9 20 0.8 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 2015 0 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Source: Statistics Canada Cansim Table 383-0021 Source: Statistics Canada Cansim Table 383-0025 5

Technology: Doing More with Less Technology has been particularly effective at replacing workers routine tasks, which has been a major driving force behind the decline of middleskilled, middle-income jobs, such as those in the manufacturing sector. Source: Autor, D. (2015). Why Are There Still So Many Jobs? 6

Technology: Doing More with Less A 2017 US study showed that between 1990 and 2007, the introduction of one additional robot per 100 workers led to a decline of nearly 6 workers. One robot 6 workers Source: Acemoglu and Restrepro (2017). Robots and Jobs: Evidence from US Labor Markets. 7

Technology: Doing More with Less 250 Canadian Manufacturing Sector Output and Employment, 1976-2013 Jobs per $1M output Employment Real gross domestic product (GDP) 25 In 1980, it took 20 workers to produce $1 million worth of manufacturing output in Canada. By 2013, this had declined to 10 workers. Manufactuing Real GDP and Employment, 1976=100 200 150 100 50 21 20 17 16 13 12 11 11 10 20 15 10 5 Jobs per $1M output 0 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 0 Source: Statistics Canada Cansim Table 282-0008, Table 379-0031, Table 383-0032, BII+E Analysis Note: GDP Chained 2007 dollars 8

Part 2) Impact of Automation on Canada s Labour Force The Talented Mr. Robot: Released June 2016 Extended Frey & Osborne (2013) occupation-based methodology to Canada s labour force Examined automation potential for nearly all occupations (498) in Canada (six NOCs were excluded from the analysis) 9

Part 2) Impact of Automation on Canada s Labour Force Over the next 10-20 years: 42% of the Canadian labour force is at a high risk of being affected by automation 45% 42% 40% 35% 30% 36% Lower earning 25% 22% 20% 15% 10% Less educated 5% Source: BII+E, 2016 0% Low Risk (0-29% Probability) Medium Risk (30-69% Probability) High Risk (70-100% Probability) 10

Part 2) Impact of Automation on Canada s Labour Force Automation Potential and Average Income $70,000 $60,000 $61,928 $50,000 $40,000 $30,000 $36,294 $33,412 $20,000 $10,000 Source: BII+E, 2016 $- Low Risk (0-29% Probability) Medium Risk (30-69% Probability) High Risk (70-100% Probability) 11

Part 2) Impact of Automation on Canada s Labour Force Automation Potential and Proportion of Labour Force with University Education (bachelor or above) 50% 45% 45.6% 40% 35% 30% 25% 20% 15% 10% 5% 15.4% 12.7% 0% Low Risk (0-29% Probability) Medium Risk (30-69% Probability) High Risk (70-100% Probability) Source: BII+E, 2016 12

Part 2) Impact of Automation on Canada s Labour Force 1,800,000 Occupations in education, law and social, community and government services Management occupations Employment 1,600,000 1,400,000 1,200,000 1,000,000 800,000 600,000 Low Medium High Health occupations Natural and applied sciences and related occupations Business, finance and administration occupations Occupations in art, culture, recreation and sport Occupations in manufacturing and utilities office support & general administration, labourers, production and sales and service workers at high risk 400,000 Sales and service occupations 200,000 0 o.1 o.2 o.3 o.4 o.5 o.6 o.7 o.8 o.9 Probability of Being Impacted by Automation in the Next 10-20 Years Natural resources, agriculture and related production occupations Trades, transport and equipment operators and related occupations Source: BII+E (2016)

Part 3) Geographic and Regional Implications Automation Across the Nation Released June 2017 Used McKinsey & Co (2017) work activity methodology Mapped onto Canadian Census Metropolitan Areas (CMAs) and Census Agglomerations (CAs) covering roughly 83 percent of the country s labour force Used NHS 2011, will update using 2016 Census data 14

Part 3) Geographic and Regional Implications Accommodation and food services Transportation and warehousing Manufacturing Mining, quarrying, and oil and gas extraction Agriculture, forestry, fishing and hunting Construction Retail trade Wholesale trade Other services (except public administration) Utilities Arts, entertainment and recreation Management of companies and enterprises Finance and insurance Public administration Administrative and support, waste management and Real estate and rental and leasing Information and cultural industries Health care and social assistance Professional, scientific and technical services Educational services 52% 52% 51% 49% 46% 45% 44% 42% 42% 42% 41% 41% 41% 38% 37% 35% 30% 61% 61% 69% These industries have the highest proportion of work activities that are technically automatable (top quartile). About 62 percent of work activities could be automated within these industries. Equivalent to 2.5 million jobs These industries are least susceptible to automation (bottom quartile). About 35 percent of work activities could be automated within these industries. Equivalent of 1.6 million jobs Source: BII+E (2017), McKinsey & Company (2017) 0% 10% 20% 30% 40% 50% 60% 70% 80% 15

Part 3) Geographic and Regional Implications Concentration of work activities with the potential for automation (location quotient Wood Buffalo, AB 49% of work susceptible to automation Mapped the susceptibility to automation for all of Canada s CMAs and CAs (83% of population) Ingersoll, ON 50% of work susceptible to automation For an interactive map visit brookfieldinstitute.ca Source: BII+E (2017), McKinsey & Company (2017) 16

Part 3) Geographic and Regional Implications Limited variation between cities and towns, meaning automation can impact cities and towns across Canada Difference in Industry Proportions, CMAs and CAs Most Susceptible to Automation (Top 20) Compared to National Average, 2011 Small regional economies specializing in manufacturing or mining, quarrying, and oil and gas extraction are most susceptible to automation Areas less susceptible to automation include cities and towns with a large hospital, postsecondary institution or public sector presence Larger, more diverse labour markets, such as Toronto and Vancouver, are more likely to reabsorb displaced labour and weather potential automation impacts Source: BII+E (2017), McKinsey & Company (2017) 17

Part 4) Tasks and Skills At-Risk and In-Demand Overall, 46 percent of work activities in Canada have the potential to be automated, across all industries. This does not mean that these jobs will be automated. Most jobs comprise a mix of work activities, only some of which are automatable; however, this proportion of work activities is significant equivalent to 7.7 million jobs. Source: BII+E (2017), McKinsey & Company (2017) Tasks Most at Risk Data Collection Data Processing Predictable Physical Work Tasks Least at Risk Managing/ Stakeholder Interaction Applying Expertise Unpredictable Physical Work 18

Part 4) Tasks and Skills At-Risk and In-Demand Canada s labour force should be equipped with a broad suite of technical and soft skills that will be important for growth and that can not (yet) be automated, including: skills associated with digital literacy, entrepreneurship, and social intelligence. 19

Part 4) Tasks and Skills At-Risk and In-Demand A 2016 survey of 90 large Canadian private-sector employers identified teamwork, communication, and problem-solving capabilities as some of the most important skills for entry-level positions - skills that current technology can complement but not replace. A study from Burning Glass Technologies found that nearly 7 million job openings in the US in 2015 were for roles requiring coding skills, representing 20 percent of the total job market for career-track jobs that pay more than $15 an hour. The pace of change will also increase the value of entrepreneurial skills, not just for startups but for all Canadian firms. Entrepreneurs tend to score high on leadership, identifying opportunities, and managing uncertainty. 20

Part 5) Implications Negative Implications: Positive Implications: Potential for large displacement of workers Productivity gains, required to sustain aging population Specifically, low-skilled, low-wage jobs particularly at risk However, risks also cut across income ranges and all job categories. Likely to have disproportionate impact on less economically diversified areas of Canada Political instability / economic anxiety Workforce disruption and worker displacement Productivity gains and firm competitiveness Help address talent gaps in particular areas of the economy Creation of new jobs of the future At firm level, higher labour productivity can increase competitiveness especially, exportoriented firms Emerging market opportunities for firms who seize technology 21

Part 5) Implications The rise of populist movements internationally is at least partially in response to the tension between innovation and inclusive growth. A 2016 analysis of the US presidential election by FiveThirtyEight demonstrated that county-level results strongly correlated to the share of a county s jobs that are routine. President Trump s margin was close to 30 points in counties with more than 50 percent of jobs that are routine. 22

5) Implications Areas to explore: Investments in education and training, e.g.: Accessible digital literacy + coding education Work-integrated learning Lifelong learning / rapid upskilling New models for labour market information, e.g.: Improved signaling from employers to trainers + educators New sources of data Policies aimed at improving equity, e.g.: Basic income Changes to tax policy Changes to labour policy Policies to address cost of living in cities Test new approaches, e.g.: Co-design new models with users workers, employers, service providers, unions Collect data on program outcomes Pilot, evaluate, scale 23

6) Appendices 120 Business Sector Labour Productivity Growth, Canada 1961-2015 Information and communications technology (ICT) has been the driving force behind the acceleration of labour productivity growth in Canada and the United States since 1996 (Sharpe, 2006). 2007=100 100 80 60 40 20 0 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Source: Statistics Canada Cansim Table 383-0021 24

6) Appendices 250 Canadian GDP, Employment, Productivity and Income: 1981-2015 Brynjolfsson and McAfee (2014) argue that even 200 though technological innovations promise to radically increase 150 productivity, they are also having adverse effects on workers, particularly the lowand middle-skilled. They use 100 stagnant median wages and the falling labour share of 50 income as evidence of cheaper and better digital technologies substituting for labour. 0 1981=100 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Real gross domestic product (GDP) Median income (2014 constant dollars) Source: Statistics Canada Cansim Table 383-0008, BII+E Analysis Total number of jobs Labour productivity 25

6) Appendices Innovation (patents per capita) is driving higher levels of income inequality, accounting for around 17% of the total increase in the top 1% income share on average across US states between 1975 and 2010. However, innovation also enables entrepreneurs to create businesses and thrive. It is positively correlated with upward social mobility driven mainly by entrant innovators (i.e. startups). Source: Aghion, P. et. al (2015). Innovation and Top Income Inequality. NBER Working Paper No. 21247. 26