Average is Over

New technologies are transforming the structure of the US economy but creating only modest numbers of jobs, according to the biggest official survey of businesses, conducted only once every five years.

The 2012 economic census shows how technology is creating a boom in output for new industries – such as shale gas and internet retail – but only a modest increase in their payrolls.

It highlights concerns that recent innovations in information technology tend to raise productivity by replacing existing workers, rather than creating new products that demand more labour to produce.

The FT link is interesting throughout and I believe these numbers vindicate what many of us have been arguing.  It also stresses the oft-neglected point that mining and drilling are relatively capital-intensive sectors:

Drilling is capital intensive, however, so even though the industry’s sales rose by $142bn, its annual payroll was up only $20bn to $61bn in total.

It also turns out that online retail is not very labor intensive at current margins.


Man, reading this is like pepper spray to the face.

Too often I read economists dismissing the idea that the rate of job automation can outpace the rate of job creation. Their argument is that is hasn't happened in the past, it won't happen in the future. That however, is a poor argument. I used to think that couldn't happen because the rate of job loss through innovation was essentially measuring the same thing as rate of job creating through innovation. They both are a measure of human creativity. Recently I've begun to believe that this too is mistaken. That in the age of automation and material abundance the job creating is harder, it requires new ideas, while automation often just requires the application of existing ideas to a known problem.

When thinking about this problem I often think of the poor horse, who numbered in the 30 millions in 1900, and now at just 7 million. Why didn't we think of new jobs for the horse?

Care to guess how many horses there were in North America in 1500?

Probably about the same number as white men.

Despite significant regulatory headwinds, the global horse slaughter business has never been better. Horses are exploited to make newer and better animal glues as well as a meat slurry that tastes very much like beef. This may be a preview of future trends in the demand for humans and their labor.

I'm long on Soylent Green. Huxley has been right about everything so far so "recycling" humans for their phosphorus is probably not out of the question.

mmmmmmm, meat slurry.

The key is to think of people as having the imagination and resourcefulness of horses. This comes naturally to top-down theorists.

I can't find a job so I guess I'll just eat some grass. Maybe somebody will show up and feed me some oats. Sure hope so.

Glad to hear you're ok!

Agreed. Take a couple days off, for goodness' sake.

Give me beans!

People who think the US economy needs high wages and mass consumer spending to keep chugging are going to be in for a rude surprise. Despite the purgiing of ZMP workers from payrolls we still see the stock market reaching new heights, BitCoin is up, many real estate markets are on fire right now. The US economy has never been in better shape and is generating wealth at an unprecidented rate. There is no downside here.

I like this blog but Tyler is incredibly ignorant about the oil & gas industry. How about a post comparing Silicon Valley to the oil industry? Which has more advanced technology? Which supplies more jobs? Which employs more scientists? Which is more important to the American economy?

"the oft-neglected point that mining and drilling are relatively capital-intensive sectors"

Pretty sure that the conventional wisdom is that mining and drilling are capital-intensive.

This post is quite timely. According to numbers released on Tuesday (3/25), Connecticut has the highest per capita income. Which state ranks second? North Dakota.

Drilling, and even refining have always had small labor forces, mining too, but drilling is in a class by itself, and has been since the nineteenth century.

If given a choice between more job creation and more productivity, its better for the economy to choose the later. American labor is still too expensive, which explains the trend towards more automation and outsourcing. Being above average is necessary to join the middle class. Being average won't cut it anymore.

What happens to the 150mil Americans who are average or below average? How long before being above average doesn't cut it?

Here's an if-by-whiskey take:

Either the USA will continue its relatively free trade policies, in which case middle class wage growth will slow until we get closer to the global average; or the USA will continue its relatively lop-sided trade agreements, in which case wage growth will continue as-is, but unemployment will keep creeping up.

Wage subsidies, aka the rich bribing the poor not to revolt.

You do realise what the 'middle' in 'middle class' implies, yes, from a statistical pov?

Why have we not reduced the number of hours people worked like we did when productivity increased before 1950 . Just think how many more people would not be able to find a job if we still had a 60 hour work week.

Hey, the Googlers I know would be glad to take 40% wage cut for a 3 day work week (assuming other take the same cut and they do not become less competitive in the housing market). But good luck teaching your gender studies majors how to code (even your average STEM major will not be good enough at it). It is not as easy as recruiting, training and managing 66% more workers (even if it was http://en.wikipedia.org/wiki/Brooks%27s_law still kills you). The whole point about the "average is over" (or shall we say winners take all) economy is that only the top 10-20% have the skills to succeed.

Like France did, for instance?

This also seems like the "lump of labor" fallacy, which I cannot adequately define in a pithy manner.

"It highlights concerns that recent innovations in information technology tend to raise productivity by replacing existing workers, rather than creating new products that demand more labour to produce."

Krugman, the one from the 1990s not the one now, wrote this interesting column about the fallacy of composition as it applies to technology "replacing" workers: [http://www.slate.com/articles/business/the_dismal_science/1997/01/the_accidental_theorist.html]. One could say that technology and productivity gains place workers into different jobs, but that's different from *re*placing them. As Krugman from the 90s says: "Productivity growth in one sector can very easily reduce employment *in that sector*. But to suppose that productivity growth reduces employment in the economy *as a whole* is a very different matter" (his emphasis).

As to why it *is* possible for the economy to suffer inadequate demand, Krugman from the 90s also has an explanation: "Such slumps are essentially *monetary*--they come about because people try in the aggregate to hold more cash than there actually is in circulation. (That insight is the essence of Keynesian economics.) And they can usually be cured by issuing more money--full stop, end of story. An overall excess of production capacity (compared to what?) has nothing at all to do with it."

Glad you are OK, Tyler.

Someone remind me again why we must have Third World immigration.

Tyler has admitted that we don't need them and they're just going to drive down middle and working class incomes. Alex's ideology of social justice is rendered illogical, given that so much labor competing for fewer jobs means they recreate their homeland conditions here.

It seems the real goal is just to elect a new people.

Unskilled workers are really handy for making the rest of us better off. I grew up in El Paso, in the sixties and seventies, when there was a free flow of labor across the river. My parents were just poor school teachers but we always had a maid and a gardener. Everyone had one. There are plenty of low skill jobs in our economy to fill. But there are no longer any low skill workers to fill them because crossing the border is now next to impossible.

There are plenty of low skill workers to fill them, but people aren't going to work for five bucks an hour.

It's pretty easy to imagine that technological advances are removing existing categories of jobs while government gumming up of the works is impeding the formation or expansion of new categories of jobs. Isn't it?

No, it's actually very very hard to imagine...

It's only hard to imagine if you get hard when you imagine government generally.

ad hominem?

I am just being not so ideologically blinkered as to ignore reality to make a point. I am sure the US and UK and Europe and Japan's systems of governments can be made better in a variety of ways and everybody should always try to improve its 'ease of doing business'.

And, sure, there will always be a tension between 'ease of doing business' and valid concerns about safety, work practices etc etc that requires legislation and control.

None of the above implies that I love big government. Actually, just like everyone else on Earth, I love small government. But I also want public goods.

You can have public goods without the massive government apparatus that forces something like 40% of workers to get permission from government to do work, and forces incompetent workers out of a job because their labor isn't worth a fixed price.

So assuming TC and all these articles are correct, what is the prescription to ensure sufficient incone for everyone? National basic income? Make work jobs?

Drop the cost of government by 10% of GDP. Too much overhead.

How will firing government workers help things? They pay taxes and participate in the econony. Many if them are above average (FDA, SEC, FBI, DOJ). Pushing government workers into the private sector will drag wages down.

Those above average government workers - where do their wages come from?

Reduce corporate income tax rate to 0%.

LOL. Tom Friedman cracking the whip. "Innovate! Create! Work! Harder, harder, harder!

I guess that was directed at me?

Let's say you are trapped on a desert island and your choice is to either start a fire and try to hunt for some food, or to try to survive using no tools, no fire, and only your bare hands. It's certainly difficult to do the former, but that doesn't make it any less necessary. Your (I guess?) solution of closing borders and receding into ethnic enclaves solves which problem exactly?

All the books and articles profiled on this site are spelling out a scenario where innovation and technology are reducing the need for labor. The data backs up the articles and books. So won't more innovation simply accelerate the problem? What new innovation will put 10mil out of work Americans into the labor force at a reasonable wage? I'd sooner kick out the illegal immigrants, freeze the borders and let the unemployed Americans become on the books dish washers, nannies and day laborers. That will buy us some time to see how other countries are handling the issue of technological unemployment.

Innovation and technology come from human minds. How is it possible on the one hand that there is a demand for greater innovation and on the other hand a reduced need for labor? Someone has to do the innovating, so why not those who find themselves unemployed in their current capacity?

I'm not saying that to be flippant, I'm asking what other choice they have? Beg the government to give them a guaranteed income? This is the better, more reasonable solution?

And note that this situation applies equally regardless of our chosen immigration policy. Saying, "Yeah, but more people will make things worse" is really just changing the subject.

"Innovation and technology come from human minds."

They come from 2% of human minds. In the past innovations led to new industries and new jobs. But that's not happening anymore. Innovations are helping businesses grow with fewer people. Truck drivers aren't going to get jobs at Google when they lose their jobs to the autonomous vehicles.

It may not be better or more reasonable but it might be a lot more likely than turning everyone into 'creators'. Creators of what?

But I don't quite get the issue. If Humanity reaches the point where it no longer has to work, doesn't it mean we 'won'? My take: http://theredbanker.blogspot.com/2013/01/robots-labour-and-sciences-fiction.html

... and also The Culture: http://en.wikipedia.org/wiki/The_Culture

Welcome to 'post-scarcity'...

"But that’s not happening anymore."

Okay, if you start from the assumption that no further industry-creation is possible through technological innovation, then you're right. I surely is a dilemma.

But I reject that assumption, and I don't see any particular reason why I should accept it.

Because innovation takes capital unemployed people don't have.

In hunter-gatherer and pastoral societies, concepts like borders and blood ties are taken very seriously.

But this is not an answer to my question.

At the least we seem to be discussing this.
@RPLong, industry-creation is certainly possible. But industry-creation is done by the "smart-fraction", and the smart fraction pie is not being grown by lower average IQ immigration.

If the population increases, more industry-creation is needed to keep the current and new population employed. This can only work if one of the following two assumptions hold:

1) the new population has the same smart fraction as the existing population, OR,

2) the new population has a lower smart fraction but this will increase becasue of better education.

The data shows 1) is false. You believe 2), and others here includingme think 2) is not valid.

That is the basis of our disagreement about where this is going.

ad*m - Hold that thought a moment. I agree that innovation is done by the "smart-fraction," but I disagree that they are the sole beneficiaries. Take the invention of the automobile as an extreme example: Sure, the Henry Fords of the world were the biggest immediate winners, but think of all the auto factory jobs he created. If we're really talking about industry creation then my only point is that it is not necessarily true that "not-smart-fraction" will be jobless in the future. At least not until we invent self-building robots, and I'm not ready to assume a self-building robot just yet...

One final point: The whole benefit of owning a dishwasher is that you don't have spend so much time doing dishes. With the time you once spent hunched over the sink, you're now doing something else. Some of what you do is leisure and some of what you do is labor.

If a robot displaces my job, that only means that I have more time to dedicate to something else. Maybe I won't invent the flying car, but most people who lose their jobs eventually find some other useful way to spend their time, other than being mindless welfare cases. Right? I'm saying that this will be as true in the robot future (ha ha) as it is in the stagnating present.

We can't all have high paying jobs. I sure don't. But that doesn't mean if I lose the one I've got, it's time to hang it up and go on the dole. C'mon! :)

The answer to your question is we shut the borders on low-skilled immigration that we don't need and can't afford. We have enough superfluous labor going on disability and welfare as it is.

I'm not sure what you're on about with your Robinson Crusoe-scenario.

Sure, additional jobs will be created when a new industry is born, but we don't know the timing or the magnitude.

If an economic oracle could accurately say today: "In 2017, a new company named Micrappleshmoogle will create an entirely new industry which provides 58 million jobs for low skill workers," we could really tailor our immigration policies to the demand.

But right now, our forecasts are not that good. Instead we have some ideologically bent economists prognostiguessing that we will have plenty of jobs for all of the unemployed. Some optimistic politicians looking for votes from a particular political constituency pass a super-idealistic, pollyannish immigration policy without allocating sufficient resources for administration/regulation/enforcement/social services/infrastructure build-out -- and here we all go. It's a great way to build a 21st century economy...

It is hard for my little brain to imagine any endeavor that would require a large number of unskilled workers. We are at the point where automation quickly becomes a better solution. Even if the cost of human workers is very low you still need to manage and train those people. In addition the lower the pay the less likely it is for any particular worker to show up to work. Finally if someone smarter than me did figure out a use for low cost labor why would they use labor in the US?

RE: The data shows 1) is false

The human race is getting stupider over time? What data shows that?

I certainly find the thesis of "Average is Over" more convincing than that of "The Great Stagnation" (http://theredbanker.blogspot.com/2014/03/the-great-stagnation-and-why-its-wrong.html) but, to reproduce a criticism first voiced by Matt Yglesias: "In Stagnation, Cowen reviewed the previous generation and concluded that despite substantial progress and catch-up in poor countries that the median household in rich countries had suffered stagnant living standards thanks to a slowdown in technological progress. In Average, Cowen looks ahead at the next generation and concludes that despite substantial progress and catch-up in poor countries, the median household in rich countries will suffer stagnant living standards thanks to a speed-up in technological progress".

i.e. both the slow-down and the speed-up of technology result in the same thing: the middle class getting shafted.

It shouldn't take a brilliant econo PhD to figure out that, whatever is going on, has little to do with the pace of technology....

In TGS, the emphasis was on tech not producing productivity gains that were comparable to previous eras holding all else constant. In TAIO, he's saying that EVEN IF we return to improved rates of technological change and productivity growth, the shift in the relative value of labor vs. capital and skilled vs unskilled workers will mean greater income inequality. He may be wrong but I don't see a contradiction there.

First of all, those lower productivity gains aren't entirely obvious. There seems to be a slow-down in nondurable goods TFP but 'general productivity' seems to have kept growing throughout the period considered.

Second, if there are shifts in labor vs capital or skilled vs. unskilled and that's more important than productivity, why isn't that insight used in TGS?

But, fair enough, maybe Matt Yglesias and myself were flippant and unfair...

...and everything to do with pounding the brave and controversial narrative of the middle class getting shafted, right?

I don't think I understand your objection.

I don't really pretend to originality (though, I'd hope to achieve 'some'). Plenty of people - except economists, it seems - understand that income inequality is directly responsible for our low growth/too-high-indebtness.

But something being uncontroversial doesn't mean it's wrong.

It does not. But it doesn't mean it's correct either. It is blithely assumed to be correct by everyone across the spectrum. And the "middle class" remains the darling of politicians of all stripes, who know how to count noses and how to play to people's fears.

How can the middle class get shafted for generations but somehow hang in there?

I say this as someone who spent a large portion of my life in the middle quintile, along with almost everyone I know.

Once again, in discussions of unemployment and concomitant lack of income nobody takes into account the techniques used by HR departments in weeding out applicants like the long term unemployed. Another unemployable demographic is made up of those who, for whatever reason, are unable to pass the increasingly meticulous screening process of background checks. Probably within our lifetimes, if the current nation-state structure survives, we'll see the development of parallel and mutually exclusive societies and economies, one a creation of the state itself, similar to what currently exists and a second laisser-faire and illegal one outside of the state purview whose members are rejected by the establishment. They gotta have some place to go.

"They gotta have some place to go."

I hope to God they knock down the gates that's keeping them out of the mainstream economy and push themselves back in.

All Americans should be able to participate in our economy.

We have enough people pushed aside and ignored as it is. What you are suggesting would mean making many bad things a permanent part of our economy and society. I think the current passivity of our population will change, once there really are only pennies being dropped to sustain the bottom 25% of our country, and nickels, dimes and quarters to the bottom 75%.

Unprecedented, long term deprivation will produce quite another different American from recent generations, with quite another viewpoint about accepting the status quo.

Suppose you have a civilization with one super person and 100 other much lower skilled people. This super person can supply all the needs of the others to survive in his spare time, while they can offer nothing econoic in return that the super person wants. How does this political economy work? What makes the super person do anything to support the others? Is this how the personalities of pets evolved?

The US gov't should be hiring more entry level people and especially long-term unemployed.
Room in the civil service should be made by making all "above median" pay jobs only part time -- 80%, 60%, 50%.

(Those in gov't who want high paid full time jobs ... should go out into the peaceful sector and get a real job.)
There is peaceful sector (private), and forced sector (gov't). The world would be better off with lower paid bureaucrats in the gov't sector.

It highlights concerns that recent innovations in information technology tend to raise productivity by replacing existing workers, rather than creating new products that demand more labour to produce

I'm always surprised how often people fail to realize there's an implicit inverse correlation to (the proportion of labor in consumed goods and services) and living standards. We could dig ditches by hand, but we don't because we couldn't afford nearly as many roads.

Comments for this post are closed