Automation, jobs, and the
future of work
A group of economists, tech entrepreneurs, and academics discuss whether technological advances will automate tasks more quickly than the United States can create jobs.
The
topic of job displacement has,
throughout US history, ignited frustration over technological
advances and their tendency to make traditional jobs obsolete;
artisans protested textile mills in the early 19th century, for
example. In recent years, start-ups and the high-tech industry have
become the focus of this discussion. A recent Pew Research Center
study found that technology experts are almost evenly split on
whether robots and artificial intelligence will displace a
significant number of jobs over the next decade, so there is plenty
of room for debate.
What
follows is an edited transcript plus video clips of a conversation on
this topic, moderated by McKinsey Global Institute partner Michael
Chui and MGI director James Manyika. The participants were Martin
Baily, senior fellow, economic studies, Brookings Institution;
Richard Cooper, Maurits C. Boas Professor of International Economics,
Harvard University; Curtis Carlson, former president and CEO, SRI
International; Reid Hoffman, cofounder and executive chairman,
LinkedIn; Tim O’Reilly, founder and CEO, O’Reilly Media; Matt
Slaughter, associate dean of faculty, Tuck School of Business; Laura
Tyson, professor of business administration and economics, Haas
Business and Public Policy Group, University of California, Berkeley;
and Vivek Wadhwa, fellow, Arthur & Toni Rembe Rock Center for
Corporate Governance, Stanford University.
Rethinking job creation
Tim
O’Reilly: There’s
this wonderful line from William Gibson, the science-fiction writer.
He said, “The future is here, it’s just not evenly distributed
yet.” So, yes, there’s all kinds of science-fiction things that
we can imagine in the future. But we can also just look around and
see what is happening today and then extrapolate forward.
Reid
Hoffman: If
you look at most of the automation, it comes down to man–machine
combinations. And all productivity means is that when you have
productivity increases, each person is doing more. And therefore, the
unit—the number of people to do this amount of work—goes down,
right? But that then creates resources for doing other work. The most
simple one was the transformation from an agricultural economy. We
used to have a huge percentage—
James
Manyika: Forty-one
percent of employment, right?
Reid
Hoffman: Yes,
was making food. And now it’s under 2 percent. What happens with
that 40 percent of the population? Well, they go on to other jobs.
Now, the reason this topic is urgent is whether exponentiating
Moore’s law changes the rule or not. There’s always painful
dislocation. Can we make that pain a lot less? Can we make the time
cycle shorter?
I
want to share one of the things I learned from a recent trip to
Shenzhen, because it’s one of the most interesting manufacturing
hubs. We went to Huawei. I was expecting, as a Silicon Valley
technologist, that it would be a complete line of robots. The whole
thing would be automated because that would obviously be the thing to
do.
Roughly
60 percent of it was automated and 40 percent of it was still people.
And it’s all a question of choice. You say, “is that just because
of low cost?” No, no. These are actually high-pay, high-skill jobs.
The answer is actually that, in the future, adaptability is key, and
people are more adaptable. So when they set up the machine line and
it’s all machines, there is a huge amount of retooling to shift
from line one to line two, whereas the people are much more easy to
shift.
Tim
O’Reilly: If
you look at this idea that it’s a combination of man and machine,
and you look at some of the examples that have really kind of
surprised us in just how they’ve taken off—like Uber, like the
Apple store—they are actually cases where humans are made more
powerful by this background. And that creates a better customer
experience, which creates new demand.Ultimately, we are going to be
focused on making better experiences for consumers, and that will not
necessarily be automatable. When we say we want to create jobs, that
takes away agency. It’s this notion that “I will hire you, and
you will work for the man.” That whole cultural ideal is an
artifact. It’s not written in stone.
Reid
Hoffman: I
think the optimistic scenario is, as Tim was describing, that we have
not only a creation of new industries and new jobs, which are
essentially a kind of full-time salary work, but also the creation of
a lot of different economic opportunities where people can be
microentrepreneurs—they can do all sorts of things. And that we can
facilitate because we’re in a networked age, with a faster ability
to have inventions and to scale up and double down on the inventions
that actually work.
You
know, the whole move from the agricultural age to the industrial age
actually came with a super amount of pain, too. The pessimistic
scenario is, roughly speaking, we have a serious youth-unemployment
problem today. And when a large percentage of unemployed youths think
they don’t have a future, that usually leads to some form of civil
instability. And so that civil instability can compound and create
reactions that then actually block out the optimistic future.
Who owns machines?
Matt
Slaughter: We
have a deep kind of risk-taking culture, a lot of institutions
concerned with how markets work—especially capital markets—and a
lot of public policies that have supported job creation in America.
My hope wavers a little bit if I add the adjective “good” to
jobs. And I think that’s a really important question.
It’s
quite clear, in the US in recent years, that we’re not creating
enough good jobs. People care a lot about their W-2s—what incomes
are they earning? If you segment this by educational attainment, 96.2
percent of the US workforce since 2000 is in an educational cohort
whose total money earnings, inflation adjusted, have been falling,
not rising.
That
includes even people with four-year-college degrees and
nonprofessional advanced degrees. The only ones that have been rising
are the PhDs, on average, and then the professional degrees: the
doctors, the lawyers, and the MBAs. So that’s a little sobering if
you think about whether we are going to create good jobs. And a big
open question that we’ll probably talk about—and our panelists
already rightly pointed to—is public policies.
Laura
Tyson: I
am with Matt on this. We live in a market economy. Supply and demand
ultimately determine the level of employment. So a number of jobs
will be created, but the quality of jobs is a huge question, I think.
What’s happening with the technology, which is skill biased and
labor saving, is that it’s eliminating middle-income jobs but is
complementary to high skills. The jobs are high-income jobs because
some smart people have to work with the technology. But there’s a
very large number of people who are being pushed down into
lower-income jobs.The second thing that’s really important—it’s
been with us for a long time—is the growing gap between
productivity and wages. And you can see this in the gap between
productivity, a measure of the bounty of brilliant machines, and how
it’s being distributed in terms of wages.
If
we had an inflation-adjusted, productivity-adjusted minimum wage
today, it would be something like $25 [an hour]. We would not be
arguing about $10. Public policy is, if anything, moving backward.
It’s certainly not moving forward at the level of the race. So the
policy makers lose the race, and a lot of displaced workers, a lot of
American families, lose the race. And that is my concern.
We’re
talking about machines—machines displacing people, machines
changing the ways in which people work. Who owns the machines? Who
should own the machines? Perhaps what we need to think about is the
way in which the workers who are working with the machines are part
owners of the machines.
Getting the workforce to adapt
Vivek
Wadhwa: So
what we’re going to see is automation. Right now, manufacturing is
trickling back to the United States. It’s not rushing back, because
of the infrastructure costs, because of the difficulty in retraining
a workforce. Give it 5 or 7 years and that trickle becomes a flood.
Give it 15 years, and now we have the robots going out on strike
saying, “stop the 3-D printers, they’re taking our jobs away.”
Because everywhere you go, you’re talking about decimation,
decimation, decimation. We cannot retrain the workforce. Now, what
are the solutions?
Curtis
Carlson: Today,
take any field—biotech, infotech, nanotech, energy, healthcare,
education. Every one, right now, is wide open to revolutionary
transformational developments. The only limitation we have is our
ability to exploit them; that’s the only limitation we have.
Maybe
we’re looking at the wrong symptoms as opposed to looking at the
fundamentals—we are not innovating at the speed of the economy. We
are not adapting fast enough. But just about everything you can
imagine can be automated. So what does that world look like? It’s
hard to know. But in the short term, the number of opportunities we
have in America is unprecedented. One problem is education. The good
news is, again, that technology is beginning to create curricula that
can transform education.
Job quality and fiscal policy
Martin
Baily: I
was struck recently by learning that in one of our largest banks, the
turnover rate for bank tellers is 50 percent a year. So, being a bank
teller now is no longer a sort of skilled job; it’s no longer
really a well-paid job. We’ve had this change in technology,
obviously. We’ve put a lot of the intelligence into the IT systems,
so we don’t need such skilled bank tellers. But if you ever go
inside a bank, you sort of long for the days when the bank teller was
more skilled.
The
banks obviously have decided, as have Walmart and many, many other
companies, that it’s more cost effective to use workers that don’t
have much training, that probably don’t have a lot of
education—although I think training is more important—but instead
to build productivity into the production system. They’re very good
at that. But it does create a huge number of not-very-good jobs,
together with a set of jobs for the conceptualizers, the people that
can take advantage of the technology, that have high incomes.
So
this has obviously created a problem of inequality in our society.
But also we’re seeing that people who cannot get or don’t have
the gumption to get—you can go both ways on this—a good job are
actually deciding not to work at all. So they’re ending up
unemployed. They’re ending up on disability. They’re ending up
leaving the labor force.
Maybe
there’s a way we can have a technological initiative that could
think about how we could change or adapt. I mean, we’re not going
to take the technology from this direction to that direction, to
change the direction of technology so that it is more friendly, more
complementary, to the mass of workers that are currently not
benefiting from technology. If workers that have been consigned to
lousy jobs suddenly see new opportunities opening to them, then I
think there’s more motivation, and they are much more willing to
seek out the skills.
Richard
Cooper: The
most rapidly growing category of jobs, in this large stub of
occupations that the Department of Labor records, was “other.”
That is to say, categories that were not big enough yet to warrant
their own line. And then of course under “other,” they did have
some nascent industries identified there.
What
people forget is that when there are innovations that destroy
jobs—and, as I say, we’ve been doing that for at least two
centuries, starting in Britain maybe two and a half centuries
ago—incomes are created for somebody else. And to close the logical
circle, you have to ask what happens to those incomes. It’s a very
important part of the process. And the incomes may be spent, but if
they’re going to be spent they have to be spent on something, and
that something creates new jobs. We may not know what they are ahead
of time, but that something creates jobs.
Where
the private market won’t do it—and there are lots of mechanisms
in the private market that contribute to creating new jobs—but
where it doesn’t create enough new jobs, then we can do it through
monetary and fiscal policy. That’s how we close the logical loop.
http://www.mckinsey.com/Insights/Economic_Studies/Automation_jobs_and_the_future_of_work?cid=other-eml-alt-mgi-mck-oth-1412
No comments:
Post a Comment