Saturday, August 29, 2015

The Productivity Paradox – a Possible Solution?

There is a great debate over whether computers have made us more efficient and how much (Paul Krugman, Noah Smith). I don’t propose to resolve that, but I do have some pertinent questions about it.

The first and most obvious one is, how is productivity measured? It seems like a holdover from the Fordist model where we can precisely determine how many widgets we can crank out with how many people. That’s an easy and straightforward calculation. But what is the “product” when most of our jobs are useless paper pushing? Most jobs in post-Fordism don’t make anything – they are service jobs or “supervisory” jobs (a euphemism for middle management). It’s easy to measure productivity in an industrial economy making widgets. But how do you do it in our economy? What is the “output?” If there is no output, then how do you measure it? I’m sure they’re using some sort of method, but I’ve never seen what it is in any article I’ve ever read on productivity. Maybe it makes sense, I just don’t know.

Most of our time is useless meetings. I had lunch with a friend of mine not long ago. He worked his way up into high-level IT without a degree. He now works for a medium-size insurance company (1,000 employees or so). He usually expresses shock that despite technically being an IT expert, all he does is “talk.” That is, he just attends meeting all day, because he is now high enough to be “management.” He has told me that he has been in meetings where the only goal is to schedule more meetings (seriously – I’m not making that up). But, as he acknowledges, he doesn’t really “do” much of anything.

That doesn’t sound very “productive” to me. Despite this, he makes a salary well over six figures. By his own acknowledgement, and to his unending bemusement, he really produces nothing at all! And this is someone in IT working directly with computers! Could this play a role? So how do you measure his “productivity?” Seriously, I would like to know. Without knowing how it is measured, how can we determine if it is accurate?

How do you measure the productivity of scientists doing research. Many of these scientists are engaged in “unproductive” activity. They must follow where the information leads them which might lead them down unproductive avenues or  to dead-ends. Yet it is extremely important. Just a few discoveries will enhance our lives tremendously. But you cannot tell that from being in the middle of the research. So how do you measure it? How about a novelist or film director? Is more output necessarily better? Roger Corman is productive, but could he produce a single Godfather film? Are all the works of Danielle Steel and Patricia Cornwell better than seven Harry Potter books?

It’s also worth noting that self-service is not counted. If a cashier can check out twenty people and hour and that goes to 25 people, the cashier is more productive. If 25 people use self-checkout in an hour, no increase in productivity is measured to my knowledge. Since our "job" is not being a cashier (or a travel agent, or an investment adviser, or a bookseller, or...), the productivity of increasing self service is not measured.

Is it possible the sheer volume of something like email engendered by ubiquitous computers is actually making us less productive? That is, we spend so much time answering emails that we don’t get anything done? If so, the efficiencies created by the computer have made us less productive. I don’t know if that’s true or not, but is anyone considering this? I have a hard time going through my email, but I know people who literally spend half their day doing nothing but answering emails. I suspect many of us know people like this. In the olden days, these same people would have spent their day not answering emails, (since it didn’t exist), but doing productive tasks such as drafting, calculating, etc. Thus it seems to me they are less productive thanks to the computer. The sheer ease of what the computer allows us to do overwhelms us and makes us spend our time in less productive tasks.

I wonder if computers have increased the amount of useless and needless work. We seem to have a need to create jobs just to employ people, rather than to do any useful work. As David Graeber points out, because we cannot reduce the amount of hours across the economy for political reasons, and because people must work a certain amount of hours to afford the standard living which is set be the amount of hours worked by the longest-working people, you get situation where people do 10-15 hours of work in 40-hour a week jobs. That is extremely inefficient, which as Graeber points out, was not supposed to happen under capitalism, only under socialism where full employment was a higher priority than efficiency. Yet it is happening.

If I got a bunch of computer programs and automated my job so that I could do my work in half the time, I wouldn't be able to come in 20 hours a week at the same pay, even though the same amount of work is getting done. I might be able to come in 20 hours with half my old pay, and with no benefits. Neither would I get a raise, since my pay is determined mainly by personality politics. So, no matter how "efficient" I get, it doesn't benefit me at all.

Have the complexity increases eaten up any benefit from automation? Just as work expands to fill the time allotted to it, complexity expands to maximally utilize the tools available to it. Often times when we work on existing buildings, we look at construction sets from the past few decades. A set of construction documents from the '50s through the '70s might consist of twenty sheets or less from each discipline. No telecommunications, no auto-operated doors, straightforward structural design of beams and columns, not many “custom” details. Today, it’s not uncommon to send out monster construction sets of over a 1,000 pages produced by dozens of specialists to document a single building. Moving columns around, which would have engendered a long series of recalculations, is done willy-nilly now thanks to computers. Design is done at the last minute. To coordinate this complexity, you need a lot more people, hence all the “supervisory” jobs that do nothing but go to meetings and answer emails. If our buildings were as simple as they were in the 1960’s, for example, we could produce those documents in a matter of weeks or months. Instead, these hyper-complex buildings take us years to design and document, even with 3-D BIM modelling. So if you measure, you would see no productivity gain at all, despite all the people working.

I think there is another important explanation. We know that Neoliberalism’s chief job creation engine is low-wage service occupations. See this: Technology has created more jobs than it has destroyed, says 140 years of data (The Guardian). But if you read the fine print, you’ll see why:
The report cites a “profound shift”, with labour switching from its historic role, as a source of raw power, to the care, education and provision of services to others. For example, the 1871 census records that there were 9,832 accountants in England and Wales and that has risen twentyfold in the last 140 years to 215,678. Technological progress has cut the prices of essentials, such as food, and the price of bigger household items such as TVs and kitchen appliances. The real price of cars in the UK has halved in the last 25 years, notes Stewart. That leaves more money to spend on leisure, and creates new demand and new jobs, perhaps explaining the big rise in bar staff, he adds. Concluding that “the stock of work in the economy is not fixed”, the report cites the surge in hairdressers as evidence that where one avenue closes in the jobs market, others open. The Deloitte economists believe that rising incomes have allowed consumers to spend more on personal services, such as grooming. That in turn has driven employment of hairdressers. So while in 1871, there was one hairdresser or barber for every 1,793 citizens of England and Wales; today there is one for every 287 people.
So, is it possible that the efficiency of computers has led to less people doing the same job. The wealth captured therein is channeled into the creation of a vast array of low-wage service occupations (i.e. McJobs) But it is impossible to raise the productivity of service occupations much by definition. Baumol’s cost disease explains why:
To understand the cost disease, start with a simple observation: whatever the economy’s average rate of productivity growth, some industries outpace others. Take car manufacturing. In 1913 Ford introduced assembly lines to move cars between workstations. This allowed workers, and their tools, to stay in one place, which cut the time to build a Model T car from 12 hours to less than two. As output per worker grows in such “progressive” sectors, firms can afford to increase wages.
In some sectors of the economy, however, such productivity gains are much harder to come by—if not impossible. Performing a Mozart quartet takes just as long in 2012 as it did in the late 18th century. Mr Baumol calls industries in which productivity growth is low or even non-existent “stagnant”.

Employers in such sectors face a problem: they also need to increase their wages so workers don’t defect. The result is that, although output per worker rises only slowly or not at all, wages go up as fast as they do in the rest of the economy. As the costs of production in stagnant sectors rise, firms are forced to raise prices. These increases are faster than those in sectors where productivity is improving, and faster than inflation (which blends together all the prices in the economy). So prices of goods from stagnant sectors must rise in real terms. Hence “cost disease”.

The disease is most virulent in industries where standardisation and automation are hard. The best examples are goods tailored to meet customer-specific demands, such as bespoke suits and haircuts…

But according to Deloitte, bespoke suits and haircuts are exactly what has absorbed the displaced workers!

The other problem is diminishing marginal returns. This is particularly pronounced in service industries, since their customer base is inherently limited, and expansion is difficult:
To illustrate this important law [of diminishing marginal returns], consider the production of Super Deluxe TexMex Gargantuan Tacos (with sour cream and jalapeno peppers). The table to the right presents the hourly production of Gargantuan Tacos as Waldo's TexMex Taco World employs different quantities of labor, the key variable input for short-run taco production. The first column is the number of workers, the second is the total hourly production of Gargantuan Tacos and the third column is the marginal product generated by each additional worker.
For the first two workers marginal product actually increases. This reflects increasing marginal returns and commonly results when the variable input is able to make increasingly effective use of a given fixed input.

For the third worker on, however, marginal product decreases. This reflects decreasing marginal returns and the law of diminishing marginal returns. The marginal product of the third worker is 25 tacos, compared to 30 tacos for the second worker. The marginal product of the fourth worker then declines to 20 tacos. For the fifth worker, the marginal product falls to 15. For each subsequent worker, the marginal product declines. Marginal product eventually reaches zero for the eighth worker and even declines for the ninth and tenth workers.

The decreasing values of marginal product exhibited for taco production by Waldo's TexMex Taco World reflect the law of diminishing marginal returns.

Computers aren’t going to increase the efficiency of taco makers, sales associates, sandwich artists or bedpan washers very much.

So if the Deloitte report is correct, and the new jobs being created by computers and automation are predominantly service jobs—like those at Waldo’s Tex-Mex World--then it is no surprise that the productivity is decreasing. It is possible that these types of jobs are eating up any productivity gains we get from computers and automation. So the growing unproductive industries would cancel out the growing productive industries, especially when we shift employees from one to the other, which we are doing and have done (manufacturing to service).

The mechanism is this – efficiency gains in manufacturing means less and less people involved in manufacturing. Computers mean more efficiency in “professional/technical” services pushing much of the workforce into the service economy and expanding that sector. Because that sector is vulnerable to Baumol’s cost disease and diminishing marginal returns, when productivity is measured as a whole across the entire economy, it appears to diminish, even where we can define what efficiency is (Sandwiches made? Bedpans changed?). If you measured it in just certain areas, we might see an increase, keeping in mind the caveats above (determining what output is, increasing complexity, useless busywork that is brought about because of the computer, etc.).

I think those might explain why, despite the fact that we all have computers on our desks more powerful than NASA used to get us to the moon, the measurements show relatively few productivity gains. This a problem primarily of economic organization. Paul Mason in Postcapitalism makes this point:
“By smashing the organized working class in the late twentieth century in all the centers starting with Japan, then Britain, then the United States then Northern Europe, what happened is that capitalism removed the impetus for a high wage, high technology solution to the problems of the late twentieth century and instead what we got was Neoliberalism. So Neoliberalism is not a new system, it is a cul-de-sac, and I argue it is a new kind of cul-de-sac which does not have a way out. And that is why we are living at a time when the fifth long wave of capitalism wishes to take off, there is a real rollout of technology going on, information technology is at the heart of it. You could look at collapsing prices of real technology, of sequencing DNA—the price of DNA sequencing falls in the exact same curve as the price of Wi-Fi or silicon chips. That’s the revolution. That’s the technological revolution. What should it lead to? Mass automation and the rapid reduction of the necessary work in society. But it’s not. Because Neoliberalism has no impetus to do that. It has only impetus to take people like you and put them in coffee bars to earn minimum wages and stay there forever. That’s the problem.”

Hence the low productivity, since productivity numbers are measured across all sectors of the economy, not just the ones that are becoming more efficient.

Note also that in the extremely productive sectors of the economy, people are being worked like dogs, which was the point of all the talk about So you have a vast, expanding sector of the economy of service jobs which are subject to Baumol’s cost disease and diminishing marginal returns. Then you’ve got a high-paid sector of people in upper management who produce nothing and go to meetings all day, but are working extreme hours to the point of burnout. Here’s the thing – overwork actually reduces productivity! See this:

Moreover, all studies of actual workers indicate that while working more might produce an initial boost, we burn out pretty quickly and don’t recover until we take a break. One study found that those who put in 55 hours a week performed worse at cognitive functions than those who worked 40 hours. Another found that workers can achieve a small boost by putting in more than 60 hours, but that it only lasts three to four weeks and then falls off. Putting in weeks of overtime eventually reduces productivity, which doesn’t bounce back. After eight 60-hour weeks, productivity is hurt so badly that it would’ve been better simply to stay with a 40-hour week the whole time. One woman who spoke anonymously about her time at Amazon described being forced to leave work by her fiancé every evening at 10 pm and doing work every day of her vacation. “That’s when the ulcer started,” she said. She no longer works there.

Sadly, Americans aren’t benefiting from all that research. Our average workweek is nearly a whole workday longer than the 40-hour one that researchers say is most beneficial. Ninety-four percent of professional workers log 50 hours a week or more.

So if Mason’s right, Neoliberalism is hitting us from both ends. A small sliver of prosperous, high-tech workers are being ground down into a sleep-deprived, low productivity psychosis, while the rest of us are stuck in low wage, low productivity McJob ghettos of diminishing marginal returns like slinging coffee and changing bedpans. I was once told by a worker at Target that she needed more hours and asked for them (she was only allowed 30), but instead Target kept hiring more people! This is extremely inefficient. It's all done so Target can avoid paying benefits. As Mason contends above, Neoliberalism is all about "getting tough" with labor, but this actually hurts productivity across the economy.

Meanwhile, with actually making stuff including the computers themselves, productivity gains are spectacular. The problem is, it just doesn’t employ that many people anymore, certainly not in the U.S. Note above that economies that still do employ their workforce in making stuff, like Germany, are the most productive, and prosperous. I’m guessing there are less Waldos Tex-Mex Taco Worlds in Germany as well.

No wonder productivity has crapped out in the U.S.

Now, if we had instead used productivity gains to reduce working hours, we would actually see those gains reflected in leisure hours. Instead we have chosen a path of ratcheting up complexity, creating bullshit jobs, and partaking of services that we could more efficiently do for ourselves (e.g. make our own coffee and drink beer at home). We would also see additional gains from, say, not spending as much time in traffic and lowered resource use (less electricity to keep office lights on, less pollution emissions to commute to jobs, etc). If we ate at home more, we’d probably be healthier, saving in health care costs. Instead we have plowed the gains into job creation in low-wage inefficient industries, destroying any productivity gains from computers. We did this to avoid mass unemployment. Do you want fries with that?

This wouldn’t be such a surprise to economists if they actually left their cloisters and spent any time in the real economy that human beings actually inhabit every day (like my IT friend). Instead they sit in their Ivory Towers and scratch their heads over the desiccated numbers and statistics, without asking how those numbers were arrived at, and without venturing outside and interacting with the ignorant, unwashed masses. Institutional economics, which describes how actual human interactions take place sans mathematical equations could provide an answer, but since it is an answer the people who employ  economists don’t want to hear, economists will most likely remain blissfully ignorant, since that appears to be their job nowadays.

Also, there was an article recently about awarding a contract to Oshkosh to build the Humvees (only 400,000 a pop) for America’s perma-wars across the globe (and probably domestic insurgency). Now this may not mean much to you, but if you’re from Wisconsin and lived here in the nineties, you remember that any cutback in the military budget, the so-called “peace dividend,” was accompanied by local reporters rushing off to Oshkosh and interviewing all the people who stood to lose their jobs. The employees would parade before the camera and lament how they would be unemployed and unable to find work, and what a horrible tragedy this would be if the military budget were cut. Somehow, all the private sector industries moving jobs overseas didn’t engender the same amount of consternation.

This is to say that one of the reasons American unemployment is so much lower than Europe’s is almost entirely due to the military!

A lot of people wanted the government to stimulate the economy building train cars and solar panels, and insulating old buildings. We didn’t do that. But we are building plenty of Humvees and jet fighters, including ones that don’t work and serve no purpose. This has sometimes been dubbed military Keynesianism. In other words, just like the military is somehow exempt from government bashing engaged in by Republicans, the amounts spent to employ Americans producing weapons is okey-dokey, but any money spent to employ Americans to do literally anything else is “waste.”


Consider how many Americans owe their jobs directly or indirectly to America’s global military. It’s a jobs program for poor rural Americans from Middle America’s ghost towns. But aside from the active-duty soldiers themselves, there are all sorts of ancillary industries, from the industries building weapons, to the contractors constructing the bases, to serving the food in the mess halls, to processing the payrolls, to providing health care, to the refineries providing the fuel for the Humvees and jets, the engineers who design the weapons, and so on.

I don’t know enough to calculate it all, but if you subtract all that economic activity out, what do you get? Considering that European defense budgets are minimal, would you have a similar unemployment rate, including youth unemployment rate (since the military disproportionately employs younger people), as Europe? My guess is the answers is yes, despite the higher minimum wages and evil welfare state.

Conservatives claim that the welfare state is why European unemployment is high. But isn’t it more accurate to say that government spending, which is allowed via military but nothing else, is why it is low here?

A final irony is that the Fox Valley where Oshkosh is located is a hotbed of far-right wing anti-government Tea Party sentiment. Joe McCarthy was from there, and the John Birch society is headquartered there.


  1. Nice ! I found your site very beautiful and distinguished too.
    Thank you for your good work and continue.
    aged care cleaning brisbane

  2. nother way the US hides its true unemployment rate is by its mania for imprisonment. A good many of the two million in our prisons would be on the unemployment rolls--that is, if they hadn't given up looking for work. In prison, they often ARE employed--at wages like 22 cents an hour--which is one hell of a sweet deal for prison industries, a deal that would make an eighteenth century slaveholder burn with envy--because HE had to pay the food and lodging costs for his slaves.

    1. Yes, the prisons also cover up the unemployment problem. People inside prisons are not counted, employed or not, and Europe does not jail even a fraction as many people as we do (nor does anyone else). This is especially notable since we dumped unemployment disproportionately on minority groups so we could just pretend that deindustrialization was no problem, despite entire swaths of the country becoming ghost towns overnight. They'll just move, we told ourselves.

      Also, consider all the money spent on prisons and just imagine if that money were used to provide jobs instead. Just paying a salary to prisoners would be cheaper than locking up ( or a negative income tax). Google "Prison Vendors See Continued Signs of a Captive Market."

      Mr. Robertson was one of 264 vendors in booths at the Indiana Convention Center for what is essentially a trade show for the prison industry...The convention is where those people window-shop. The United States currently imprisons about 2.2 million people, making it the world’s largest jailer. Those in charge of this immense population need stuff: food, gas masks, restraints, riot gear, handcuffs, clothing, suicide prevention vests, health care systems, pharmacy systems, commissary services — the list goes on. These outlays are a small fraction of the roughly $80 billion spent annually on incarceration, though precise sales figures are hard to come by because most companies in this niche market are private. Two publicly traded players, the private prison operators Corrections Corporation of America and the GEO Group, have a combined market capitalization of almost $5.8 billion. Both companies had booths in Indianapolis.

  3. You asked a strange half-a-question. You wanted to know how to measure productivity. To figure out virtually anything, you need to define just what you want to do with it, and then the answer to how to define the variables you need will be obvious. It's like asking how should I define the word "xyzabc". Well, it depends on what you want to say with it.

    Stan Erickson;

    1. So, what we measure and how we measure it has an outcome on the answer we are seeking, correct? Thus we can usually determine the answers we want by controlling what questions get asked (and which don't).

  4. Is the solution to go full Ludd? Because it sounds to me like the solution is to go full Ludd.

    In the seventies there was what they called Appropriate Technology, stuff like why use a leaf blower when a broom works as well. I really think if everyone started thinking like this, they could dump tech on its ass. Are we really living better than we were in the seventies, when we had dial phones, good handwriting, and free time?

    Here's another example, to so-called paperless office. The result of that is, people buy paper literally by the team. I remember when the average person bought writing paper in a thin little pack, and it was onionskin paper for airmail.

    1. Remember, the Luddites were less about technology than who benefitted from new technology. Those in power want to keep us from remembering that, so they conveniently redefined Luddites as people who were opposed to any and all technology, full stop (and covered up what was in effect a civil war - more English troops were used against English people than against Napoleon at times). This gave them the ability to argue that more technology is always better – technofetishism, which appropriate technology was designed to combat.

  5. "Here’s the thing – overwork actually reduces productivity!" Doesn't this make you wonder if there's something else going on, that it's not just about the money?

    There's ego satisfaction to be had from having more money than other people, a better job title, a higher rung on the corporate ladder, all that, sure. But there's also enormous ego satisfaction to be had from making the "little people" scurry around to serve your slightest whim.

    Making people work ever longer hours even though productivity suffers, their family relationships and friendships suffer, and their health suffers, can't be about the money alone. These corporate a-holes aren't entirely dim about how productivity suffers with high churn and too many hours worked.

    It can be about insecurity, that the business may fail unless the whippings escalate to match the competition, but for some it's about sadism, which is a powerful ego satisfaction.

    Anyone who's fallen under the control of a sociopath knows what I mean. And corporate capitalism today promotes and encourages that psychology.

    1. I think you hit upon something really important here. Any number of studies have shown that sociopaths are disproportionately represented in the upper tiers of management. As many as 10 percent of CEOs are sociopaths compared to 2 percent of the general population, a fivefold increase. Our system is explicitly designed to put sociopaths into positions of power and authority!

      It’s connected, I think, with tipping of all things. Tipping is a ridiculous practice. But the people who defend it tend to be the types who want to “lord it over” other people and get off on ordering people around and dangling money over them as a means of control. These tend to be management types. See this:

      …you have power over someone if they are dependent on you. So it makes sense that systems of tipping gives us power as consumers. Waiters, or other service people, are dependent on us for tips, which are often a large part of their overall salaries, so we have power over them.

      In fact, it gives us a kick, a real but short-lived kind of status in service situations. We get to temporarily play the part of the “little lord.” Some people exploit this role, demanding too much, asking for special favors, and enjoying having someone do our bidding. Others are over-sensitive to slights; if they feel like their status is being questioned at any level, they switch from little lords to little tyrants, demanding attention and extra work from their waiters. We’ve all seen this.

      There’s no “rational” defense of tipping, it’s just a power/ego trip. Just like working your employees to death.

      So it really has nothing to do with efficiency. It’s really about power and status. But wait, isn’t the economy designed to “maximize productivity” in the conceptions of economists? Yes, if like economists you remove all human social relations from economics, which is why it is a joke science that tells you nothing about the actual world we live in.

      It’s similar to the republican versus Democrats debate. Business men claim they support Republicans because they are “good for business.” But a century of data has shown that the economy does better under Democrats than Republicans in terms of jobs created and even stock market prices. So if they are hard-nosed businessmen caring about metrics, why don’; they tend to support Democrats? Because it’s rally about affiliation and ideology. The Republicans like to berate and kick down poor people and stroke the egos of the “job creators.” The actual “metrics” matter not one whit.

  6. Great info! I recently came across your blog and have been reading along. I thought I would leave my first comment. I don’t know what to say except that I have.
    healthy living

  7. Your blog entry is indeed helpful for my current research! Honestly speaking, I feel comfortable being here! Also, I managed to find - an author of this article points out essential tips on how to ideally determine working hours!


Note: Only a member of this blog may post a comment.