Carleton University, Ottawa March 2-4, 2017 Canada s Standard of Living in 2042: Will Policy Keep Pace with Technology? Christopher Gunn, Carleton University Hashmat Khan, Carleton University Conference Sponsor(s): Faculty of Public Affairs Partners: Presenting sponsor: Version / Deposit Date: 2017-06-15
Canada s Standard of Living in 2042 Will Policy Keep Pace with Technology? Christopher Gunn Department of Economics Hashmat Khan Department of Economics Visions For Canada 2042 Imagining the Canada of the Future March 2nd-4th, 2017 Carleton University Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 1 / 26
HOW WILL TECHNOLOGICAL ADVANCE IMPACT CANADA IN 25 YEARS FROM NOW? Economic Social Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 2 / 26
Labour Market Employed Population = ( ) Employed Labour force ( ) Labour force Population E-Pop RATIO = EMPLOYMENT RATE PARTICIPATION RATE Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 3 / 26
0.64 Canadian Labour Market Cansim Table 282-0002 E-POP 0.62 0.60 0.58 0.56 64 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 E-Rate E-POP 62 60 58 56 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 E-Rate P-rate 0.67 0.65 0.63 P-rate 0.61 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 4 / 26
REAL INCOME PER PERSON A measure of Canadian Standard of Living It simultaneously reveals and masks many things It reveals how an average person is doing relative to previous generations, and relative to other countries It masks the distribution of income in the society Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 5 / 26
ECONOMIC GROWTH = the growth in REAL INCOME PER PERSON Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 6 / 26
Log of Per Capita Real GDP (2002 dollars) Figure 1: Real income per person in Canada, 1914-2014 4 3.5 3 2.5 2.3% per year 2 1.5 1900 1920 1940 1960 1980 2000 2020 Year Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 7 / 26
Figure 2: Real income per person in Canada, 1981-2014 Average growth = 0.9% per year Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 8 / 26
A Useful Decomposition Level REAL INCOME = PRODUCTIVITY HOURS PER PERSON PER HOUR PER PERSON Growth rate ECONOMIC = PRODUCTIVITY + HOURS GROWTH GROWTH GROWTH Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 9 / 26
Figure 3: Annualized Growth Rates of Output per Capita, Output per Hour and Hours per Capita 2.5 2 1.5 1 0.5 0-0.5 1976-1980 1981-1990 1991-2000 2001-2010 2011-2015 Growth Rate Y/N Growth Rate Y/H Growth Rate H/N Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 10 / 26
List of considerations Physical capital and investment Human capital (educational attainment) R&D, Knowledge, Ideas Economic policies Demographic shifts (baby boomers, immigration) Climate change Health care Globalization Institutions In today s presentation we will mainly focus on the four items in bold Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 11 / 26
MODERN IDEAS-ORIENTED GROWTH THEORY The stock of ideas drive growth (Romer 1990, Aghion and Howitt, 1992) PRODUCTIVITY GROWTH { K, H, Ideas, n} Y K Y : the capital-to-output ratio H: OECD defines human capital as the productive wealth embodied in labour, skills and knowledge. R&D: The global intensity, ideas are non-rivalrous, so global idea creation matters (global researchers producing new ideas) n: The higher the population rate, the higher the number of researchers producing ideas Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 12 / 26
Sources of Canadian Productivity Growth This modern ideas-oriented theory implies (Jones, 2002) PRODUCTIVITY = TRANSITIONAL + LONG-RUN GROWTH GROWTH GROWTH {K/Y, H, Ideas} n Based on Hasanzadeh-Khan (2017) Canada: 1981-2014 1.09% = 1 0.09 %age point %age points Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 13 / 26
CAN WE COUNT ON TECHNOLOGICAL IDEAS TO DRIVE LONG-RUN GROWTH IN THE FUTURE? Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 14 / 26
Technological change Productivity growth varied considerably in past Distinct technological eras of ideas (David 1999, Jovanovic 2005) General purpose technologies - Deep inventions that are pervasive, improving over time, give rise to spin-off inventions IR1 - Steam engine, railroads (1750-1830) IR2 - Electrification, internal combustion, indoor water/plumbing (1870-1900) IR3 - Informations and communications technologies: computers, internet, digitization, robotics, AI (1960 -?) Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 15 / 26
GPT s, frontiers and productivity Where does a small country like Canada fit into these technological eras? Spill-overs from global technological frontier of ideas (Gunn 2013) IR2 had substantial impact on global technological frontier and thus productivity in frontier country US from late 1800 s to 1970 s (David 1991, Gordon 2015) IR3 behind late 1990 s 10-year surge in productivity (Jovanovic & Rousseau 2005, Gordon 2015) Huge debate in literature about future impact of IR3 on productivity Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 16 / 26
Will IR3 create high future growth? The pessimists (Gordon 2015): Late 90 s surge from IR3 was it IR2 transformed work and cultural life, happen only once Recent innovations centered on entertainment The optimists (Brynjolfsson & McAfee 2014): Late 90 s surge was just the beginning Digitization technologies not subject to as tight physical constraints as IR1, IR2 muscle power innovations Inflection point in innovation, very significant developments in very short time Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 17 / 26
Moore s Law Figure 4: Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 18 / 26
IR3 and inequality Much of debate has focused on predictions about average productivity in future But researchers tend to be in much more agreement about implications of IR3 for inequality Not a new worry. Past IR s caused significant labour displacement: machine automation of muscle power performed by humans But automation anxiety was largely misplaced as new categories of jobs created (Akst 2013, Acemoglu 2016) Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 19 / 26
But is this time different? Machine automation of mental power Humans thought to retain comparative advantage at extremes of skill distribution (Autor 2015), complex cognitive and communicative tasks (Acemoglu 2016, Brynjolfsson & McAfee 2014) But ICT threaten moving into areas previously thought safe domain of humans Result: increased likelihood of rising inequality among labour Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 20 / 26
IR3 and capital-labour inequality Digital/digitized goods non-rival, can be reproduced at very low marginal cost As result, scale of production of new product can be enormous, and can replace entire industry of workers. Ex. TurboTax software (Brynjolfsson & McAfee 2014) Income that previously flowed to the workers now flows to the creator of the idea/product Large amount of income of new products concentrated in relative few capital owners Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 21 / 26
Figure 5: Average and median income in Canada Real GDP per capita vs. median real income 200 180 160 140 120 100 80 60 40 20 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 median real inc gdp per cap Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 22 / 26
Figure 6: Intergenerational Inequality (Corak (2015)) Figure 1 Comparable estimates of the intergenerational elasticity between father and son earnings for the United States and twenty one other countries Peru China Brazil Chile United Kingdom Italy Argentina United States Switzerland Pakistan Singapore France Spain Japan Germany New Zealand Sweden Australia Canada Finland Norway Denmark.19.18.17.15.34.32.29.27.26.52.5.5.49.47.46.46.44.41.4.58.6.67 0.2.4.6.8 Intergenerational Earnings Elasticity Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 23 / 26
CONJECTURE It is likely that the gap between average and median incomes will continue to diverge over the next 25 years. Why? Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 24 / 26
Because machines are becoming better at substituting humans out of their jobs Wait...if machines are so good at replacing humans, why do we still have jobs? - The Polanyi s Paradox (Autor, 2015) It must be because We can know more than we can tell (Polanyi, 1966), and machines are not good at this But this view is challenged (i.e., machines are becoming become better at replacing humans) Machine Learning Environmental Control Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 25 / 26
Economics: Scarcity Distribution Economic policies aimed at REDISTRIBUTION OF INCOME are likely to become extremely important for the Canada of the future. Will Policy keep pace with technology? Christopher Gunn & Hashmat Khan Canada s Standard of Living in 2042 2017/03 26 / 26