Thursday

Automation and Robots: Fast Food, McDonalds, Starbucks, etc.


Millions of people hold low-wage, often part-time jobs in the fast food industry. Historically, low wages, few benefits and a high turnover rate have helped to make fast food openings relatively abundant. These jobs, together with other low-skill positions in retail, provide a kind of safety net for workers with few other options.

In the current economic environment, these jobs are, of course, much harder to get. McDonald's recent high-profile initiative to hire 50,000 new workers resulted in over a million applications---numbers that give McDonald's a lower acceptance rate than Harvard.

What about the future? Most forecasts assume that the fast food industry will continue to be a significant job creator. The Bureau of Labor Statistics ranks food preparation as one of the top four fastest-growing occupations, and that trend is expected to continue at least through 2018.

Is it possible that these projections miss the impact of technology? Could these jobs begin to disappear? For some insight into what could potentially happen, consider this article from the New York Times about the Kura sushi chain in Japan:
Efficiency is paramount at Kura: absent are the traditional sushi chefs and their painstaking attention to detail. In their place are sushi-making robots and an emphasis on efficiency.
Absent, too, are flocks of waiters. They have been largely replaced by conveyors belts that carry sushi to diners and remote managers who monitor Kura’s 262 restaurants from three control centers across Japan. (“We see gaps of over a meter between your sushi plates — please fix,” a manager said recently by telephone to a Kura restaurant 10 miles away.)
Absent, too, are the exorbitant prices of conventional sushi restaurants. At a Kura, a sushi plate goes for 100 yen, or about $1.22. 
Such measures are helping Kura stay afloat even though the country’s once-profligate diners have tightened their belts in response to two decades of little economic growth and stagnant wages.
McDonald's has already announced plans to install touch-screen ordering systems in over 7000 European locations. To me, it is not difficult to imagine many of the ideas being utilized at Kura eventually being deployed throughout the fast food and beverage industries.  If automated preparation and off-site store management work for sushi, then why not for burgers or lattes?

One important thing to take away from the sushi story is the way in which a stagnant economy can be a driving force behind increased automation. Almost any type of restaurant food is a discretionary purchase: if the price is too high, people can and will refuse to buy.  That presents a real problem if---as is the case now---businesses are seeing significant increases in the price of the food commodities they must purchase.  For a business that is squeezed between rising input prices and tepid demand, investment in labor-saving technology can represent one of the few viable paths to continued profitability.
Increased automation in fast food and beverage providers is likely to someday offer increased convenience, speed, and ordering accuracy. Robotic food preparation could also be viewed as more hygienic as fewer workers come into contact with food. And of course, price will ultimately be the determining factor. As one Kura sushi customer quoted by The Times notes:  “It’s such a bargain at 100 yen,” ... “A real sushi restaurant?” he said. “I hardly go anymore.”

If jobs in the fast food industry start to disappear, or even if the rate of job growth slows significantly, the implications for the workers that depend on these jobs of last resort will be dire. There may be few other alternatives for workers at that skill level, especially since other low-wage retail jobs may be similarly threatened.

China, Technology and Jobs - The Chinese Robots are Coming


The graph below, based on data from the Federal Reserve Bank of St. Louis, shows manufacturing employment in the United States as a fraction of all employment. As you can see, the line heads downward in an almost perfectly straight line beginning in the mid-1950s. Notice that the line doesn't become steeper as globalization takes hold after the passage of NAFTA in 1994 or the rise of China over the past decade or so. The line just slopes consistently downward.


This is primarily the result of technology, and in particular, automation. Manufacturing in the U.S. has become dramatically more productive and requires fewer workers. If we were to graph manufacturing output (rather than jobs), the line would slope upward, not downward. The value of U.S. manufacturing production is now far greater than it was in industrial era of the 1950s, even after adjusting for inflation. We just make all that stuff with a lot fewer people.

One of the most interesting things about the graph above is that, if technology is the primary driver, then employment in China must inevitably follow the same path. In fact, there are good reasons to believe that manufacturing employment's downward slope will be significantly steeper for China. The U.S. had to invent the technology to make manufacturing more productive, while in many cases China only needs to import it from more developed nations. It is also true that China is beginning its journey at a time when information technology (which is the primary enabler of automation) is many orders of magnitude more advanced than in the 1950s when U.S. manufacturing employment was at its peak. (See this recent article on skilled robots from the New York Times).

In the U.S. (as well as in other advanced countries), workers shifted out of manufacturing and into the service sector -- which now accounts for the vast majority of jobs.  Will China be able to pull off the same transition?

The U.S. had the luxury of building a strong middle class during an earlier time. Technology was advancing consistently and increasing productivity, but it was not so advanced as to create a mismatch between the type of available jobs and the skills of workers. Unionization was strong in the private sector and helped ensure that the lion's share of productivity increases ended up in workers' (rather that corporate owners') pockets. Those workers, in turn, became the broad-based consumer class that purchased the output from all those factories and kept the overall economy humming.

The situation in China is quite different. Consumer spending accounts for only about a third of China's GDP (as opposed to 60% or more in nearly all developed countries). While China has built a significant middle class in absolute terms, it remains small as a percentage of the country's huge population.
Workers enjoy few of the rights and protections that characterized the U.S. workforce of the 1950s. As I wrote in my book, The Lights in the Tunnel:
The [Chinese] government actively enforces discrimination that tends to drive wages even lower. Much of the work in China’s factories is performed by migrant workers who officially live in the countryside but are allowed to come to cities or industrial regions to work. These workers typically live in factory dormitories and do not have the right to bring their families to the cities or to genuinely assimilate into an urban middle class. Wages for these workers are far lower than for urban dwellers, and the money that they do earn is for the most part either saved or sent home to help support their families. These workers are not in a position to become major drivers of local consumption any time soon.
According to the New York Times, those worker dormitories apparently play an important role in Apple's (or Foxconn's) ability to bring production online at any time of the night or day:
A foreman immediately roused 8,000 workers inside the company’s dormitories, according to the executive. Each employee was given a biscuit and a cup of tea, guided to a workstation and within half an hour started a 12-hour shift fitting glass screens into beveled frames. Within 96 hours, the plant was producing over 10,000 iPhones a day.
Even that level of worker availability and efficiency isn't enough for Foxconn, which recently announced the introduction of huge numbers of robots. That may be a great way to drive production, but it's hard to see how China will succeed in dramatically shifting its economy toward domestic consumption.
And that has to happen before a shift to a service economy can take place. As consumers become more wealthy they begin to spend a larger fraction of their incomes on services -- things like banking, insurance, healthcare, education, entertainment and travel -- and that in turn drives service sector employment. At least that has been the path followed in other developed countries.

In the absence of consumer spending, China's economy remains highly dependent on manufacturing exports and, especially, on fixed investment. An astonishing 50% of China's GDP is driven by investment in things like factories, housing and infrastructure (the U.S. figure is around 15%). The problem is that all that investment has to ultimately pay for itself, and that happens via consumption. Once a factory is built it has to then produce something that gets sold at a profit. Homes, retail buildings and apartment complexes likewise have to be sold or rented out.  Obviously, no economy can indefinitely invest anything like 50% of its output without eventually finding a way to get a positive return on that investment.

Achieving  that return requires consumers -- either at home or abroad. China continues to rely heavily on consumers in the U.S. and Europe, but that's unlikely to be a sustainable formula for growth.  The debt crisis and the resulting austerity is cutting into economic growth and consumer spending in both Europe and the U.S.

As manufacturing automation increases (perhaps dramatically) in China, in the U. S. and other developed countries the most disruptive impact from technology will be in the service sector -- where millions of white collar jobs and service jobs in retail, distribution, food service and other areas may ultimately be at risk. After all, if robots can build an iPhone, then its a good bet that they will also someday be able to build a hamburger or mix a latte.  The result may be continuing high unemployment, stagnant wages and tepid consumer spending throughout much of the developed world.

The real problem China faces is that it is late to the party. Just as it reaches its manufacturing employment zenith, it faces a potentially disruptive impact from automation technology. And that will happen roughly in parallel with similar transitions in the service sectors of the countries that currently consume much of its output. In the face of that, can China succeed in re-balancing its economy toward consumption, increasing personal incomes, and building a vibrant service sector to keep its population employed?

Android / iOS Apps, Robo Shoppers and the Future of Online Shopping


Retail shopping -- and retail jobs -- are on the verge of being completely transformed by a range of new technologies. First among these is mobile, where shoppers are increasing using their devices to shop online for lower prices and to manage the entire shopping experience on their mobile phones.

One good example is RoboShopper, a new Android app that allows showers to scan barcodes and then easily look for better deals using multiple comparison shopping services. Since online stores can escape the fixed costs associated with maintaining retail locations, its a good bet that an app like RoboShopper will find a lower price somewhere. Shoppers can also maintain shopping lists on their phones and then email or share those lists, making it possible to scan barcodes with their phones and then complete the shopping process on a larger computer if they prefer.

As mobile phones increasingly become the focus of the shopping experience, the primary role for brick and mortar retailers is often degrading into that of a showcase for shoppers who want a physical connection with products -- before they purchase at lower cost elsewhere. Traditional retailers are trying to come up with strategies to combat "showcasing."  According to an article in Newsweek, one of the hardest hit chains, Best Buy, is planning to close dozens of stores, open smaller locations and invest in more luxurious retail settings in the hope that this will drive sales. Other major retailers have inked deals with their vendors to supply unique products (with unique barcodes), making it more difficult for their customers to comparison shop.

Online retailers are offering new delivery options designed to overcome the "immediate gratification" advantage that is perhaps the primary asset of traditional retailers. Amazon now has lockers in many locations where customers can take delivery of orders. eBay is experimenting with same day delivery, and probably has pockets deep enough to continue refining the service even if it means years of losses.
Even those traditional retailers who remain in strong positions are being transformed by mobile. 

Walmart, for example, has its own experimental program where shoppers are able to scan barcodes and then checkout and pay with their phones -- completely avoiding long checkout lines.  It seems clear that future shoppers will rely more and more on their mobile devices as a way to shop, pay and get help and information about products, even while in traditional retail settings.  This will surely create many opportunities for mobile advertisers to bombard shoppers with targeted offers at the exact moment they are on the verge of completing a purchase. And that would make things even harder for brick and mortar retailers (and especially smaller stores) to retain sales.

Retail has recently been one of the most important job creating sectors of the economy. Nearly every recent college graduate probably knows someone who was unable to land a job that required a college degree and is instead working in the retail sector. However, as technology and online competition transform the retail industry, there is certain to be a significant impact on jobs.
It seems almost certain that online retailers will continue to take a larger and larger share of the pie. In theory, this should not eliminate jobs but rather transition them from retail stores to warehouses and distribution centers. However, Amazon's recent purchase of the warehouse robot company Kiva Systems probably gives us a pretty good indication how things are likely evolve in the future: once jobs move to a warehouse it becomes easier to automate them.  At the same time, of course, technologies like self-service checkout lanes, mobile checkout -- and perhaps sometime soon even in-store robots -- will continue to drive down the need for even traditional retailers to hire as many people.

Big changes are coming to the way we shop and to the number and types of people employed in the industry. The only thing you can be really sure of is that your mobile phone, together with apps like RoboShopper will probably play a more and more important role.

Wednesday

Unemployment and Technology

My last post, Structural Unemployment: The Economists Just Don't Get It, generated some interesting comments. Here's an especially good one that was posted by an irate economist over at Mark Thoma's blog:
This is the second link I've seen to Mr. Ford's views on labor and technology. It needs to stop. Yes, we need to consider the interaction between the two. We do not, however, need the help of someone whose thinking is as sloppy and self-congratulatory as Ford's. Lots of work has been done on technology's influence on labor markets. Work that uses real data. Ford is essentially making the same "technology kills jobs" argument that has been around for centuries. His argument boils down to "this time it's different" and "people (economists) who don't see things exactly as I do feel threatened by my powerful view and are to be ignored."

There is a whiff of Glenn Beck in Ford's dismissal of other views.
Now, I think that saying "it needs to stop" and then comparing me to Glen Beck is a little over the top. It seems a bit unlikely that my little blog represents an existential threat to the field of economics.

The other points, however, deserve an answer: First, am I just dredging up a tired old argument that's been around for centuries? And second, have economists in fact done extensive work on this issue---using real data---and have they arrived at a conclusion that puts all of this to rest?

It is obviously true that technology has been advancing for centuries. The fear that machines would create unemployment has indeed come up repeatedly---going back at least as far as the Luddite revolt in 1812. And, yes, it is true: I am arguing that "this time is different."

The reason I'm making that argument is that technology---or at least information technology in particular---has clearly been accelerating and will continue to do for some time to come. (* see end note)

Suppose you get in your car and drive for an hour. You start going at 5 mph and then you double your speed every ten minutes. So for the six ten-minute intervals, you would be traveling at 5, 10, 20, 40, 80, and if you and your car are up to it, 160 mph.

Now, you could say "hey, I just drove for an hour and my speed has been increasing the entire time," and that would basically be correct. But that doesn't capture the fact that you've covered an extraordinary distance in those last few minutes. And, the fact that you didn't get a speeding ticket in the first 50 minutes really might not be such a good predictor of the future.

Among economists and people who work in finance it seems to be almost reflexive to dismiss anyone who says "this time is different." I think that makes sense where we're dealing with things like human behaviour or market psychology. If you're talking about asset bubbles, for example, then it's most likely true: things will NEVER be different. But I question whether you can apply that to a technological issue. With technology things are ALWAYS different. Impossible things suddenly become possible all the time; that's the way technology works. And it seems to me that the question of whether machines will someday out-compete the average worker is primarily a technological, not an economic, question.

The second question is whether economists have really studied this issue at length---and by that I mean specifically the potential impact of accelerating technical progress on the worker-machine relationship. I could not find much evidence of such work. In honesty, I did not do a comprehensive search of the literature, so it's certainly possible I missed a lot of existing research, and I invite any interested economists to point this out in the comments.

One paper I did find, and I think it is well-regarded, is the one by David H. Autor, Frank Levy and Richard J. Murnane: "The Skill Content of Recent Technological Change: An Empirical Exploration," published in The Quarterly Journal of Economics in November 2003. (PDF here). This paper analyzed the impact of computer technology on jobs over the 38 years between 1960 and 1998.

The paper points out that computers (at least from 1960-1998) were primarily geared toward performing routine and repetitive tasks. It then concludes that computer technology is most likely to substitute for those workers who perform, well, routine and repetitive tasks.

In fairness, the paper does point out (in a footnote) that work on more advanced technologies, such as neural networks, is underway. There is no discussion, however, of the fact that computing power is advancing exponentially and of what this might imply for the future. (It does incorporate falling costs, but I could not find evidence that it gives much consideration to increasing capability. It should be clear to anyone that today's computers are BOTH cheaper and far more capable than those that existed years ago.).

Are there other papers that focus on how accelerating technology will likely alter the way that machines can be substituted for workers in the future? Perhaps, but I haven't found them.

A more general question is: Why is there not more discussion of this issue among economists? I see little or nothing in the blogosphere and even less in academic journals. Take a look at the contents of recent issues of The Quarterly Journal of Economics. I can find nothing regarding this issue, but a number of subjects that might almost be considered "freakonomics."

The thing is that I think this is an important question. If, as I have suggested, some component of the employment out there is technological unemployment, and if that will in fact worsen over time, then the implications are pretty dire. Increasing structural unemployment would clearly spawn even more cyclical unemployment as spending falls---risking a deflationary spiral.

Consider the impact on entitlements. The already disturbing projections for Medicare and Social Security must surely incorporate some assumptions regarding unemployment levels and payroll tax receipts. What if those assumptions are optimistic?

Likewise, I think economists would agree that the best way for developed countries to get their debt burdens under control is to maximize economic growth. If we got into a situation where unemployment not only remained high but actually increased over time, the impact on consumer confidence would be highly negative. Then where would GDP growth come from?

It seems to me that, from the point of view of a skeptical economist, this issue should be treated almost like the possibility of something like nuclear terrorism: Hopefully, the probability of its actual occurence is very low, but the consequences of such an occurence are so dire that it has to be given some attention.

So, again, I wonder why this issue is ignored by most economists. There are a few exceptions, of course. Greg Clark at UC Davis had his article in the Washington Post. And Robin Hason at GMU wrote a paper on the subject of machine intelligence. I don't agree with Hanson's conclusions, but clearly he understands the implications of exponential progress.

Why not more interest in this subject? Perhaps: (A) Conclusive research has really been done, and I've missed it. or (B) Economists think this level of technology is science fiction and just dismiss it. or (C) Maybe economists just accept what they learn in grad school and genuinely don't feel there's any need to do any research into this area. Maybe something like this is so far out of the mainstream as to be a "career killer" (sort of like cold fusion research).

Another issue may be the seemingly complete dominance of econometrics within the economics profession. Anything that strays from being based on rigorous analysis of hard data is likely to be regarded as speculative fluff, and that probably makes it very difficult to do work in this area. The problem is that the available data is often years or even decades old.

If any real economists drop by, please do leave your thoughts in the comments.

___________________________

* Just a brief note on the acceleration I'm talking about (which is generally expressed as "Moore's Law"). There is some debate about how long this can continue. However, I don't think we have to worry that Moore's Law is in imminent danger of falling apart because if it were, that would be reflected in Intel's market valuation, since their whole product line would quickly get commoditized.

Here's what I wrote in The Lights in the Tunnel (Free PDF -- looks great on your iPhone) regarding the future of Moore's Law:
How confident can we be that Moore’s Law will continue to be sustainable in the coming years and decades? Evidence suggests that it is likely to hold true for the foreseeable future. At some point, current technologies will run into a fundamental limit as the transistors on computer chips are reduced in size until they approach the size of individual molecules or atoms. However, by that time, completely new technologies may be available. As this book was being written, Stanford University announced that scientists there had managed to encode the letters “S” and “U” within the interference patterns of quantum electron waves. In other words, they were able to encode digital information within particles smaller than atoms. Advances such as this may well form the foundation of future information technologies in the area of quantum computing; this will take computer engineering into the realm of individual atoms and even subatomic particles.

Even if such breakthroughs don’t arrive in time, and integrated circuit fabrication technology does eventually hit a physical limit, it seems very likely that the focus would simply shift from building faster individual processors to instead linking large numbers of inexpensive, commoditized processors together in parallel architectures. As we’ll see in the next section, this is already happening to a significant degree, but if Moore’s Law eventually runs out of steam, parallel processing may well become the primary focus for building more capable computers.

Even if the historical doubling pace of Moore’s Law does someday prove to be unsustainable, there is no reason to believe that progress would halt or even become linear in nature. If the pace fell off so that doubling took four years (or even longer) rather than the current two years, that would still be an exponential progression that would bring about staggering future gains in computing power.

Causes of Unemployment

Lately, there has been a fair amount of buzz in the economics blogosphere about the issue that I've been discussing extensively here: Structural Unemployment.

Several notable economists have weighed in. Paul Krugman touches on it here. Brad DeLong says this. Mark Thoma has a post in a forum focusing on structural unemployment at The Economist.

If you read through these posts, however, you won't see a lot of discussion about the case I've been making here, which is that advancing technology is the primary culprit. I've been arguing that as machines and software become more capable, they are beginning to match the capabilities of the average worker. In other words, as technology advances, a larger and larger fraction of the population will essentially become unemployable. While I think advancing information technology is the primary force driving this, globalization is certainly also playing a major role. (But keep in mind that aspects of globalization such as service offshoring---moving a job electronically to a low wage country---are also technology driven).

The economists sometimes mention technology, but in general they find other "structural" issues to focus on. One that I have seen again and again is this idea that people can't move to find work because their houses are underwater (the mortgage exceeds the equity). The emphasis given to this issue strikes me as almost silly. Are there any major population centers in the U.S. that have really low unemployment?

Even if people could sell their homes, would they really be motivated to load up the U-haul and move from a city with say 12% unemployment to one where it is only 9%? Have the economists lost sight of the fact that 9% unemployment is still basically a disaster? The few locales I've seen with unemployment significantly lower than that are rural or small cities (Bismark ND, for example)---places that are simply incapable of absorbing huge numbers of hopeful workers. Let's get real: playing musical chairs in a generally miserable environment is not going to solve the unemployment problem.

Another thing the economists focus on is the idea of a skill mismatch. Structural unemployment, they say, occurs because workers don't have the particular skills demanded by employers. While there's little doubt that there's some of this going on, again, I think this issue is given way too much emphasis. The idea that if we could simply re-train everyone, the problem would be solved is simply not credible. If you doubt that, ask any of the thousands of workers who have completed training programs, but still can't find work.

Economists ought to realize that if a skill mismatch was really the fundamental issue, then employers would be far more willing to invest in training workers. In reality, this rarely happens even among the most highly regarded employers. Suppose Google, for example, is looking for an engineer with very specific skills. What are the chances that Google would hire and then re-train one of the many unemployed 40+ year-old engineers with a background in a slightly different technical area? Well, basically zero.

If employers were really suffering because of a skill mismatch, they could easily help fix the problem. They don't because they have other, far more profitable options: they can hire offshore low wage workers, or they can invest in automation. Re-training millions of workers in the U.S. is likely to make a killing for the new for-profit schools that are quickly multiplying, but it won't solve the unemployment problem.

Why are economists so reluctant to seriously consider the implications of advancing technology? I think a lot of it has to do with pure denial. If the problem is a skill mismatch, then there's an easy conventional solution. If the problem's a lack of labor mobility, then that will eventually work itself out. But what if the problem is relentlessly advancing technology? What if we are getting close to a "tipping point" where autonomous technology can do the typical jobs that are required by the economy as well as an average worker? Well, that is basically UNTHINKABLE. It's unthinkable because there are NO conventional solutions.

In my book, The Lights in the Tunnel, I do propose a (theoretical) solution, but I would be the first to admit that any viable solution to such a problem would have to be both radical and politically untenable in today's environment. That's why I don't spend much time suggesting solutions here: what would be the point? (but please do read the book---it's free). I think the first step is to get past denial and start to at least seriously think about the problem in a rational way.

The few economists that have visited this blog and commented on my previous posts have generally barricaded themselves behind economic principles that were formulated more than a century ago (see the comments on my posts about the lump of labour fallacy and comparative advantage).

Most economists seem to be unwilling to really consider this issue---perhaps because it threatens nearly all the assumptions they hold dear. I wrote about this in my first post on this blog. We'll see how long it takes for the economists to wake up to what is really happening.

Update

I've posted a followup that addresses comments and poses some questions for economists: Econometrics and Technological Unemployment -- Some Questions

Jobless Recovery and the Jobless Future

January 2010

The official unemployment rate remains at 10%, and economists are projecting that the job market will take years to recover. Is it possible that, beyond the obvious impact of the financial crisis, there is another largely unacknowledged factor contributing significantly to the dismal unemployment situation?

I believe that there is, and I argue that case in my new book, The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future. In the past two decades, information technology has advanced dramatically and is increasingly being employed to eliminate jobs of all types. Job automation technology, together with globalization, has been the primary force behind the stagnant wages and diminished opportunities for less educated workers we've seen in recent years.

Because information technology accelerates (roughly doubling every two years), rather than increasing at a constant rate, we can expect that the coming years and decades will see even more dramatic progress. In the future, automation is no longer going to be something that primarily impacts low wage, uneducated workers. Technologies such as artificial intelligence, machine learning and software automation applications will increasingly enable computers to do jobs that require significant training and education. College graduates who take "knowledge worker" jobs will find themselves threatened not only by low-wage offshore competitors but also by machines and software algorithms that can perform sophisticated analysis and decision making.

At the same time, continuing progress in manufacturing automation and the introduction of advanced commercial robots will continue to diminish opportunities for lower skill workers. Technological progress is relentless, and machines and computers will eventually approach the point where they will match or exceed the average worker's ability to perform most routine work tasks. The result is likely to be structural unemployment that ultimately impacts the workforce at virtually all levels -- from workers without high school diplomas to those who hold graduate degrees.

Most mainstream economists dismiss this scenario. They continue to believe that the economy will restructure and create adequate numbers of jobs in the long run. Historically this has, in fact, been the case. Millions of jobs were eliminated in agriculture when mechanized farm equipment was introduced. That resulted in a migration to the manufacturing sector. Likewise, manufacturing automation and globalization has resulted in the transition to a largely service-based economy in the United States and other developed countries.
In the past, technology has typically impacted one employment sector at a time, leaving or creating other areas for workers to transition into. That's unlikely to be the case this time around. Accelerating information technology will offer a completely unprecedented level of work capability -- and it can be applied virtually everywhere. As technology providers compete and innovate, automation will certainly become more affordable and more accessible to even the smallest businesses. If a business can save money through automation, competitive pressures will leave it no choice but to do so. While there will certainly continue to be jobs that cannot be automated, the reality is that a very large percentage of the 140 million or so workers in the United States are employed in jobs that are fundamentally routine and repetitive in nature. Enormous numbers of these jobs are going to be vaporized by technology in the coming decades, and because that technology will be available across the board, there is really very little reason to believe that entirely new employment sectors capable of absorbing massive numbers of workers will be created.

The problem is not just one of unemployment. In my book, I use a thought experiment or mental simulation based on "lights in a tunnel" to illustrate the overall economic impact of relentlessly advancing job automation technology. As unemployment increases and wages fall, discretionary consumer spending and confidence will likewise plummet. The result will be a downward economic spiral that will be very difficult to arrest. Beyond some threshold, the business models of mass market industries would be threatened as there would simply be too few viable consumers to purchase their products. We would also likely see unprecedented levels of debt defaults, plunging asset values and financial system disruptions that might easily exceed what has so far occurred in the current crisis.

I believe that the impact of accelerating automation technology is likely to present an enormous economic, social and political challenge over the next ten to twenty years and beyond. Yet, the issue is simply not on the radar. In The Lights in the Tunnel, I suggest some possible reforms that might address the issue, but the reality is that the problem is potentially so disruptive that even progressive thinkers would probably find some of my ideas extreme. Conservatives will likely view my proposals as unthinkable. Nonetheless, if we are ultimately destined to progress into a world where traditional jobs are simply unavailable and where a huge percentage of the population has little in the way of marketable skills or opportunity to earn an income, there will be few if any viable solutions that would not be perceived as radical.

Martin Ford is the author of The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future

Friday

Artificial Intelligence - Economics and Job Market Impact

Robin Hanson, a professor at George Mason University (who also blogs at Overcoming Bias), is one of the few economists who has given serious thought to the potential economic implications of artificial intelligence. In Dr. Hanson’s 1998 paper, “Economic Growth Given Machine Intelligence,” he suggests several variations on a growth model which assumes that machines achieve sufficient intelligence to become complete substitutes for, rather than complements to, human labor. Dr. Hanson’s conclusions are very optimistic, and to me, quite counterintuitive. His models “suggest that wholesale use of machine intelligence could increase economic growth rates by an order of magnitude or more.” At the same time, however, he notes the obvious reality that as machines become affordable, and very likely more capable, substitutes for human workers, “wages might well fall below human subsistence levels. “

My immediate reaction to this is that economic growth at any level—let alone of an order of magnitude beyond what we are accustomed to—is fundamentally incompatible with wages that are falling dramatically for the vast majority of workers. We might, perhaps, have vigorous economic growth if the falling wages applied to only a minority of human workers, but it is very difficult for me to conceive of a way in which such growth would be compatible with wages falling across the board—or even for the bulk of workers. The reason is simple: workers are also consumers (and support other consumers). If wages fall dramatically, then consumption must likewise fall because the majority of personal consumption is supported by wage income.

Additionally, I would make the point that as artificial intelligence overwhelmed the labor market, the psychological impact on consumers would almost certainly amplify the fall in consumption. As the number of viable consumers in the market rapidly diminished, the mass market business models of most industries would be numerically undermined: there simply would not be enough willing and able buyers to support the high volume production of goods and services that characterizes most of the major industries that make up today’s economy. The result would not be extreme economic growth, but instead falling revenues and, quite probably, widespread business failures. It seems to me that the current economic situation offers a fair amount of support for my position.

In his paper, it appears to me that Dr. Hanson makes two separate assumptions to get around this basic problem of a reasonable balance between production and consumption. In his initial overview of his growth models, Dr. Hanson writes: “We assume that the product of the economy can be either consumed or used to produce more of any kind of capital (i.e., human, hardware, software, other).” I read this to mean that Hanson is assuming that private sector investment might “pick up the slack” left by diminished consumption. This strikes me an unsupportable assumption for the simple reason that business investment is not independent of consumption—or, at least, current investment is clearly a function of anticipated future consumer spending.

Businesses invest in order to better position themselves to reap the fruits of future consumption. As a business owner myself, I can really think of no other good reason for a business to make substantial investments in “human, hardware, software, or other” capital. In a world in which most workers’ jobs are being automated away and will never return, there would be very little reason to anticipate that future consumer spending would be anything but anemic. Therefore, there would be no reason for the private sector to invest. To me, it seems intuitively obvious that overall private sector investment would not increase, but instead would fall in line with plummeting consumption.

Dr. Hanson does, however, offer a second assumption that might help get around this problem:

Surviving the Economics of Artificial Intelligence: Everyone a Capitalist

Dr. Hanson suggests that his results “may be compatible with a rapidly rising per-capita income for humans, if humans retain a constant fraction of capital, perhaps including the wages of machine intelligences, either directly via ownership or indirectly via debt.” In other words, he seems to be saying that if consumers have an ownership interest in the economy of the future, then the resulting investment income will be sufficient to make up for the precipitous decline in wages. Presumably this would allow the population to continue consuming. Dr. Hanson fleshes out this view in another article on “Singularity Economics” that was published in IEEE Spectrum in June, 2008:

…human labor would no longer earn most income. Owners of real estate or of businesses that build, maintain or supply machines would see their wealth grow at a fabulous rate—about as fast as the economy grows. Interest rates would be similarly great. Any small part of this wealth would allow humans to live comfortably somewhere…

In other words, everyone (or at least most people) will have a piece of the action, and the returns on that ownership will be so fantastic that almost everyone will have a reasonable discretionary income—with which they can then go out and consume.

I’m going to leave aside the obvious problems with the distribution of wealth and income, as well as with any hope of social mobility, that this scenario implies, and instead focus on a more basic question: would asset values really increase at the extraordinary rate that Hanson suggests? Would they increase at all?

Dr. Hanson seems to be assuming that the stock market (and other productive assets) would increase dramatically in value because investors would recognize that businesses now have a fantastic new technology (intelligent machines) which will enable extraordinarily efficient production. The problem I see with this is that, according to modern financial theory, asset values are not determined by investors’ perceptions about technology. Asset values are defined by investors’ expectations for the future cash flows that will be associated with the asset in question. It seems clear to me that, in the midst of across the board job automation and plunging consumer demand, those future cash flows would be looking pretty minimal.

In order for asset values to increase, investors would have to reason that, because asset values would increase dramatically, nearly everyone would have access to an investment income which could then be used to consume—and thereby create the future cash flows which would justify the current value of the asset.

That strikes me as both circular and unlikely. Dr. Hanson seems to be assuming a perpetual asset bubble that somehow gets going even though it is not even remotely initially supported by fundamentals. In fact, the initial fundamentals would point—quite dramatically—in exactly the opposite direction.

Where would these courageous initial investors come from? During the height of the financial crisis last year, I happened to see a report on CNBC which noted that extremely wealthy people were buying gold and having it shipped directly to their estates. They wanted their gold in their homes, behind their gates and walls, buried in their underground vaults. Are those perhaps the investors that Dr. Hanson is expecting to step up to the plate and begin driving up asset values when intelligent machines arrive and destroy consumer demand?

It seems to me that Professor Hanson’s views are really quite unsupportable. Nonetheless, it is of course possible that I have made an error somewhere or I have misunderstood Dr. Hanson’s arguments. I look forward to Professor Hanson’s response to my thoughts.

Update: Dr. Hanson has responded on his blog. My response to his response is here.