The Economics of Artificial Intelligence

Home > Other > The Economics of Artificial Intelligence > Page 33
The Economics of Artificial Intelligence Page 33

by Ajay Agrawal


  much replication, etc.). Some attention to the “direction” may bring much

  larger returns.

  6.4 Concluding

  Remarks

  The historical record suggests that dismal prophecies about the economic

  and social impact of great technological advances rarely come to pass. Thus,

  with AI poised to emerge as the new GPT, we should not necessarily envision

  a future whereby humans will be rendered obsolete and mass unemployment

  will be the “new normal.” At the same time, as many occupations will indeed

  vanish, and many others will undergo signifi cant changes, it is important to

  inquire into what sort of strategies may ameliorate the detrimental eff ects

  of AI and enhance the positive ones. This is all the more important given

  that in the twenty- fi rst century the public at large has much less tolerance

  for bearing the costs of technical change and higher expectations for sharing

  into its benefi ts here and now.

  Therefore, we need to anticipate the required institutional changes, experi-

  ment in the design of new policies (particularly in education and skills de-

  velopment) in the professionalization of service occupations, and in aff ect-

  ing the direction of technical advance. Furthermore, economists possess a

  vast methodological arsenal that may prove very useful for that purpose—

  we should not shy away from stepping into this area, since its importance

  for the economy cannot be overstated.

  186 Manuel Trajtenberg

  References

  Bresnahan, Timothy, and Manuel Trajtenberg. 1995. “General Purpose Technolo-

  gies ‘Engines of Growth’?” Journal of Econometrics 65 (1): 83– 108.

  Gordon, Robert J. 2016. The Rise and Fall of American Growth. Princeton, NJ:

  Princeton University Press.

  Heckman, James, Tim Kautz, Ron Diris, Bas ter Weel, and Lex Borghans. 2014.

  “Fostering and Measuring Skills: Improving Cognitive and Non- Cognitive

  Skills to Promote Lifetime Success.” Report prepared for the Organisation of

  Economic Co- operation and Development, Paris. http:// www .oecd .org/ edu

  / ceri/ Fostering- and- Measuring- Skills- Improving- Cognitive- and- NonCognitive

  - Skills- to-Promote- Lifetime- Success .pdf.

  Hirschman, Albert. 1970. Exit, Voice and Loyalty. Cambridge, MA: Harvard Uni-

  versity Press.

  Mokyr, Joel. 2017. “The Past and the Future of Innovation: Some Lessons from Eco-

  nomic History.” Paper presented at the conference on the Economics of Artifi cial

  Intelligence, Sept. 2017, Toronto.

  Summers, Lawrence H. 2016. “The Age of Secular Stagnation: What It Is and What

  to Do About It.” Foreign Aff airs, Feb. 15. http:// larrysummers .com/ 2016/ 02/ 17

  / the- age- of-secular- stagnation/.

  II

  Growth, Jobs, and Inequality

  7

  Artifi cial Intelligence, Income,

  Employment, and Meaning

  Betsey Stevenson

  The evolution of artifi cial intelligence (AI) evokes strong emotions in

  people. Some imagine a dystopia in which people are replaced by machines.

  Machines will develop the content we read, and the entertainment we enjoy.

  Artifi cial intelligence will pick our friends and our politicians, and ultimately

  take away any sense of human agency. And worst of all, those machines

  will deprive us of work. Human beings will lose meaning and income, and

  perhaps ultimately, be driven to extinction.

  At the other end of the spectrum are those that envision the potential

  for utopia. With machines doing all the work, people will have plenty of

  income, yet very little unpleasant work to do. Instead, people will spend their

  days enjoying art and music. They will pursue their passions unburdened

  by the need to provide for their basic wants. They will feed their intellectual

  curiosity and fulfi ll the human demand for personal interactions. In short,

  people will be able to enjoy their lives with the freedom from time and money

  constraints that artifi cial intelligence provides.

  So who is right?

  7.1 Income Is Not the Problem

  Economists think that we know the answer, or at least part of it. Most

  economists believe that automation promises a future of higher income that

  Betsey Stevenson is associate professor of economics and public policy at the Gerald R. Ford School of Public Policy, University of Michigan; a visiting associate professor of economics at the University of Sydney; a research associate of the National Bureau of Economic Research; a research affi

  liate of the Centre for Economic Policy Research; and a research fellow of CESifo.

  For acknowledgments, sources of research support, and disclosure of the author’s material fi nancial relationships, if any, please see http:// www .nber .org/ chapters/ c14026.ack.

  189

  190 Betsey Stevenson

  stems from the higher productivity that artifi cial intelligence will provide.

  In September 2017, the Chicago Booth IGM Forum’s Economic Experts

  Panel asked forty- one economists from top universities in the United States

  whether they strongly agreed, agreed, were uncertain, disagreed, or strongly

  disagreed with the following statement: “Rising use of robots and artifi cial

  intelligence in advanced countries is likely to create benefi ts large enough

  that they could be used to compensate those workers who are substantially

  negatively aff ected for their lost wages.”1

  The answer was clear; no one disagreed with that statement. A few econo-

  mists—10 percent—were uncertain, and the modal answer was agree, rather

  than strongly agree. Yet, it is clear that economists believe that artifi cial intel-

  ligence represents an opportunity for substantial economic gains. Indeed,

  productivity gains have been at the heart of improvements in living stan-

  dards from the beginning of time. And so, it is diffi

  cult to imagine a world

  in which productivity gains do not generate benefi ts suffi

  ciently large that

  we could compensate the losers.

  Therefore, the relevant question is whether we would compensate the

  losers. Here economists are more skeptical. Economics tells us that there

  will be income gains, but our social and political structure help determine

  how they will be distributed.

  7.2 Who Gets the Gains from Automation?

  Much of the skepticism about being able to successfully redistribute

  income comes from a lack of trust that the political process will successfully

  manage redistribution in a world in which income is primarily generated

  by capital. The history of the last several decades has certainly not been

  encouraging on that front. The share of income held by the top 1 percent

  of the population has risen to nearly 20 percent, from around 10 percent

  in 1980, while the share going to the bottom 50 percent of the population

  has fallen to 12 percent from 20 percent in 1980.2 Currently we are failing

  to redistribute the gains from technological advances, and so the concerns

  that distribution will be a challenge are supported by our recent past.

  7.3 What Will We Do with Ourselves?

  Yet, the concern runs deeper than wondering whether as a society we

  could manage to redistribute income. Most economists are concerned
about

  how we will allocate jobs, and underneath that concern lies a belief that work

  matters independent of the earnings that are generated by the work.

  1. IGM Economic Experts Panel (2017).

  2. World Wealth and Income Database. http:// wid.world/ country/ usa/.

  Artifi cial Intelligence, Income, Employment, and Meaning 191

  Essentially, many people are skeptical that people could successfully fi nd

  engaging and emotionally rewarding ways to spend their time if they were

  not working. One of the IGM Forum panelists, Robert Hall, expressed his

  concern most concretely: “Those not in the labor force are unhappy and

  inclined to opioids.”

  So economists are fearful about what will happen if people lose employ-

  ment opportunities, yet economic history provides economists with opti-

  mism that employment will adapt. Which is why so many economists wonder

  what, if anything, will be diff erent about artifi cial intelligence compared to

  the industrial revolution or other important periods of rapid technological

  change.

  Economists’ intuition around the impact of technological change on

  employment comes from considering how employment has adapted fol-

  lowing previous periods of technological change. Here, once again, econo-

  mists have a united view: technological change has not historically reduced

  employment. This view of economists is seen in a February 2014 ques-

  tion posed to the Chicago Booth IGM Forum’s Economic Experts Panel.

  Forty- four economists from top universities in the United States were asked

  whether they strongly agreed, agreed, were uncertain, disagreed, or strongly

  disagreed with the following statement: “Advancing automation has not

  historically reduced employment in the United States.”3

  Economists are roughly united in agreeing with this statement, with only

  4 percent disagreeing and 8 percent uncertain.4

  Yet, when the IGM Economic Experts Panel was asked in September

  2017 whether they strongly agreed, agreed, were uncertain, disagreed, or

  strongly disagreed with the following statement: “Holding labor market

  institutions and job training fi xed, rising use of robots and artifi cial intel-

  ligence is likely to increase substantially the number of workers in advanced

  countries who are unemployed for long periods.”

  This is where economics lends a less clear answer and economists are

  divided on this question: 44 percent agree, 26 percent disagree, and 31 per-

  cent are uncertain. Is this a contradiction or a diff erent view about artifi -

  cial intelligence compared to other technologies? I don’t think it is either.

  Instead, I believe these answers refl ect the diff erence in what happens in the

  long run versus the short run. In the long run, technological change leads

  to prosperity and new jobs arise as we adjust to our new wealth, develop

  new skills, and come up with new ways to use human skills. In the short run,

  however, there is often a disruption.

  3. IGM Economic Experts Panel (2014).

  4. The fi gure of 88 percent is adjusted for respondents’ confi dence in their answer. Among all respondents, 76 percent agreed and 9 percent had no opinion.

  192 Betsey Stevenson

  7.4 The Long Run

  One of the confusions around what will happen to employment and

  unemployment stems from not separating short- run versus long- run eff ects.

  When most of us think about artifi cial intelligence and increased automa-

  tion, we are trying to think about what the long- run future holds, and our

  intuition comes from considering how growth has changed how people live

  across generations. It is not how it has changed our lives over the last fi ve

  years, but instead contrasting how we live our lives—and if you are reading

  this it involves large periods of intellectual contemplation—with how our

  own family members ten generations back spent their lives. In the 1800s,

  the vast majority of Americans worked in agriculture and very few of them

  spent their time thinking about ideas. Today, 2 percent of Americans are

  directly employed in agriculture. There are more people employed in the

  public school system than in agriculture. In sum, few of us are in the jobs

  or careers that our great- great- great- great grandparents were in and many

  of us work in jobs today that did not exist a single generation ago.

  One of the IGM panelists, Nancy Stokey, made it clear she was think-

  ing about the long run: “If this had been true over the last two centuries,

  almost no one would be working anymore.” When you take a really long-

  run view, it has to be true that automation has not reduced employment, at

  least not at as rapid a pace as the automation has itself occurred. In fact,

  many economists regard it as a puzzle that paid work has been remarkably

  stable even as nations have become increasingly prosperous, and its citizens

  might have been expected to use more of their higher income to choose to

  consume more leisure.

  7.4.1 In the Long Run, Employment and Hours Worked Have Declined

  Yet, despite our intuition, employment has tended to decline with techno-

  logical progress. The diff erence between our beliefs about how technological

  progress has impacted employment and what has actually happened refl ect

  two things. The fi rst is that hours worked and employment has not declined

  by as much as one might have predicted. The second is that economists tend

  to think about employment in a model in which people who want to work

  can fi nd jobs.

  Hours of work have declined in most countries with productivity growth.

  Figure 7.1 shows average annual hours worked in a handful of developed

  countries since 1970. Annual hours worked declined fairly steadily in France,

  Germany, and Japan. The United States and the United Kingdom had

  smaller declines. Yet in each country, the annual hours worked fell.

  To think more broadly about employment, childhood employment has

  been almost eliminated in developed countries. And employment of young

  adults, those age fi fteen to twenty- fi ve have declined as young people focus

  on investing in further human capital. On the other end of the life cycle, life

  Artifi cial Intelligence, Income, Employment, and Meaning 193

  Fig. 7.1 Average annual hours worked

  Source: OECD (2017).

  expectancy has increased while retirement ages have fallen in most developed

  countries.

  Work has declined in terms of the number and share of our life in terms

  of hours and days that we are going to spend working. The decline in work

  has occurred through the interaction of economic growth with government

  policies. For example, extended retirement has been facilitated by govern-

  ment pension and retirement programs. The dramatic reduction in child

  labor was facilitated by child labor laws. The demand for these programs

  and regulations is itself facilitated by the higher income that productivity

  growth creates.

  Decreases in employment because of childhood education and retirement

  are thought to be improvements in living standards and not something we

  need or want to fi x. However, they do require income redistribution. Older


  generations must support children, either through families or government

  redistribution (such as child tax credits, child allowances, child health care

  subsidies, etc.). Yet, most people agree that this is an improvement—few are

  trying to get kids back into the workforce to fi nancially support themselves.

  Something similar is true at the other end of the life cycle. While the elderly

  can save for retirement, redistribution allows those who are retired to share

  in continued economic growth.

  7.5 Short-

  Run

  Disruption

  The real uncertainty with artifi cial intelligence is what will the disruption

  be like and how will we manage people through it. Most economists think

  194 Betsey Stevenson

  there will be people who are hurt through decreased demand for their skills.

  There might be longer spells of unemployment and a larger need for worker

  retraining. There might be jobs that workers do not want or are not qualifi ed

  to do. While we can prepare a new generation for a world in which robots

  do many of the jobs, preparing a generation midway through their lives is

  harder. People are resistant to starting over, they mourn what they have

  lost, and they resent a defi nition of progress that leaves them diminished in

  status and income.

  The loss of income should be easier to solve than the loss of status. So how

  important is work and what do we know about it? Is work about the income

  that it generates or about the meaning and order it gives to our days? Much

  of the debate about the potential impact of automation on employment is

  really a debate about how we will spend our time. So it is useful to separate

  out the question of what will we do with our time if the robots take our jobs

  from the question of whether we can fi nd a stable and fair distribution of

  income in such a scenario. And it is useful to realize that the answers in the

  long run may be very diff erent to what happens in the short run. Yet, how

  we handle the short run will ultimately infl uence our long- run outcomes.

  7.6 There Is Work outside of Employment

  Work is a broader concept than paid labor. Paid labor is the result of

  a trade- off between leisure, home- produced goods, and market- produced

  goods. This matters from a measurement perspective because the 1970s was

 

‹ Prev