Do 2018 US Productivity Numbers Show Start of a New Boom?

The Us Bureau of Labor reported a 1.0% labor productivity gain in 2017-2018. This number was better across the country than the 0.4% per year gain from 2007-2017. However, labor productivity was still only 1.3% versus the 2.8% labor productivity of the PC-Internet productivity boom from 1995-2007. The lack of a labor productivity boom that matches the multi-factor boom indicates that it is not technology efficiency of the PC-Internet scale. The Nextbigfuture thesis is that the worker on demand, services on demand and cloud services that people and businesses use regularly has reached the scale to move the overall productivity number. However, labor is not really that much more efficient. People are being worked hard. Uber and Lyft drivers are working on weekends and we are working assets harder with X as a service. This level of stronger productivity number has room to run for many years. It will then merge into a larger real AI-robot-self-driving automation wave. It is a real productivity shift because it takes a lot to reach the scale to move these big nationwide numbers.
First line shows strong but not PC-Internet great multifactor productivity. Labor productivity in the second line did not move. Capital and Labor look like they are worked harder.

SOURCE- Bureau of Labor Statistics, Analysis by Brian Wang of Nextbigfuture

35 thoughts on “Do 2018 US Productivity Numbers Show Start of a New Boom?”

  1. You’d be right if the worker would do the same task (i.e. pressing a button) with Machine A and Machine B that produces 10x more, but that’s rarely the case. Blue collar work used to be physically challenging back then, now it’s all about computer skills. Usually operating more complex machines require a better training, that’s why they say that education is a driver of (labor) productivity.
    Your insistence on differentiating the inherently related labor/capital seems a little naive.

  2. To reiterate on my past posts, I believe the 20s will be boom times for capital based income to increase as a percentage of all earned income, with a corresponding decrease in wage based income.

    In other words, it’s better to have a lot of wealth as opposed to a decent job. Duh, right? The difference is the degree to which this will be true. Of course, declining wage-based income also has a negative impact on capital based income. This gets really complicated trying to figure out how badly.

    Stepping back, I’m still a hopeful fan of technological singularities. The next one usually appears in about half the time it took for the latest one to occur after the one previous to it. Bear in mind that singularities aren’t instant but, instead, can span a substantial period of time. The dates will NOT be exact!
    PAST SINGULARITIES (approx. dates for most recent)
    1545 – Printing Press, 256 years later we get
    1801 – The Industrial Revolution, 128 years later we get
    1929 – Electronics, 64 years later we get
    1993 – World Wide Web, 32 years later we get ?

    2025: Full Automation (Cognition based) – A perfect storm of technologies create automation technology that reduces or eliminates a multitude of occupations. Eventually leads to fusion power and cheap energy. 
    2041: Synthetic intelligence
    2049: Biological singularity (indefinite lifespans)
    2053: Man-machine?
    2055: Mind-to-mind? Nanobot replacement of biological cells?
    2056, 2056, 2056, 2056, 2056 . . . .

  3. Economics didn’t destroy slavery. Slave prices were still rising at the start of the Civil War, and economic returns were quite high, around ten percent.

    Same thing with the British. When they banned slavery in the Empire, sugar prices tripled.

  4. You’re arguing that the definition of labor productivity isn’t the definition of labor productivity. That’s a pointless argument, because it is what it is: labor productivity is a measure of how much (or little) labor goes into the output.

    If you want to argue that multi- or total-factor productivity (which usually includes capital inputs) is a better measure, that’s fine. But you don’t get to choose your own definitions if you want anybody else to understand you.

  5. I was under the impression that the economic obsoletion of the Southern US slave economy was the fundamental CAUSE of the civil war

    No. While it was starting to happen, it wasn’t the fundamental cause of it. It was the new political realignment that started with the creation of the GOP and the election of Lincoln that triggered the Succession. The war didn’t start until several months later when the South gave the North casus belli with the attack on Fort Sumter. The North — and especially Lincoln — thought the division of the US to be a long term threat to continental security. Didn’t help that Brits jumped in and did with the South what the US has done in recent times in Ukraine.

    In a democracy, that means that the political power of the slave owners also began to shrink and shrink.

    1) US is not nor ever has been a ‘democracy’.
    2) The plantation owners would always be in charge no matter what, as proven by how they swiftly took back power after Reconstruction was over.

    To keep the slave economy at all…that meant leaving the arrangement where factory workers in New York had political power over South Carolina.

    Not really:!/articles/1/essays/6/three-fifths-clause

    But in the Senate, yes as the Missouri Compromise was repealed. That contributed to the South’s secession more than anything.

  6. Initially the cotton gin actually made slavery more profitable, because it increased the economic return from raising cotton, while still leaving a lot of other labor necessary to get it to the gin.

  7. “If you’re arguing that labor is a diminishing input in overall output, that’s true. But that’s just another way of saying that labor productivity is going up.”

    No, my point is that it’s not labor productivity, if labor isn’t responsible for the additional production. It’s capital productivity.

  8. Agenda drivers can use something real to push their agenda. It’s more effective to use a half truth than something that is a complete lie.

  9. I was under the impression that the economic obsoletion of the Southern US slave economy was the fundamental CAUSE of the civil war.
    Once the slave economy was no longer competitive, the proportion of the population that was operating in a slave economy began to shrink and shrink.
    In a democracy, that means that the political power of the slave owners also began to shrink and shrink.
    Given that the non-slave economy voters both didn’t like slavery for ethical/religious reasons, and didn’t like how slaves could compete with them for low wage jobs*, they started to vote against slavery.
    To keep the slave economy at all (and there were parts where it was still highly beneficial to the people on top) that meant leaving the arrangement where factory workers in New York had political power over South Carolina.

    *. A slave could compete with a free worker for a job (such as assembling a widget) even if the free workers were overall a more economically competitive arrangement, because the marginal cost of the slave was lower even if the overhead was sending the slave owner broke.

  10. Yes. It drops when the labor market slackens too. But the main differences is between what happens in the short term (daily or even monthly) vs the long term (at least a year’s time frame).

    When labor markets are depressed, capital is not invested as much into productivity enhancing investments. So, productivity suffers over the long term. In the short term, it seems that productivity increases because those left with jobs have to do the jobs of 1.7 people or even more.

    Whereas when labor markets tighten, capital (in general) is incentivized to invest in productivity investments. In some economic sectors less than others and vice versa. Long term productivity increases, as those investments — either in machinery or better management or both — tend to live on after the boom. In the short term, it suffers because those poor bastards doing the work of 1.7 people discover that they have greener grass to go work on instead. Greener grass involving less work for more pay.

    Sometimes technology suddenly disrupts things even when labor markets are depressed that capital invests in it anyway. This was why slavery in the US was on it s way out even w/o the Civil War. Why? Because Eli Whitney’s cotton gin was such a huge (and disruptive) boost to Southern plantation productivity that only cultural and legal standards kept the institution of slavery an operating business for as long as it did.

  11. Or…(Occam’s Razor) it is all BS, which there is ample proof backing up that conclusion. It is those who refuse to see/acknowledge that proof who are the deniers.

    Similar behavior is now being exhibited about Trump’s Russian Collusion even tho that was long discredited and now officially discredited.

    In both cases, agenda drivers have used it to purposely push outright falsehoods because of motivations that have nothing to do with saving the planet or actually being concerned about ‘Russian collusion’.

  12. Labor productivity, the amount of production due to a given amount of labor, rarely budges much.

    Sure it does. If one worker makes one widget worth $1 per hour, the labor productivity is $1/labor hour. If somebody makes a capital upgrade so that the one worker now supervises a robot that makes 100 of the widgets an hour, labor productivity is now $100/labor hour. That’s a pretty big budge.

    If you’re arguing that labor is a diminishing input in overall output, that’s true. But that’s just another way of saying that labor productivity is going up.

    If you’re arguing that labor productivity is now less relevant as a measure of overall economic wellbeing than multi- or total-factor productivity, then… OK… but now you’re having a different argument. Labor productivity still is what it is, and the fact that capital investment enhances it is irrelevant.

  13. I have no confidence in this measurement. I know the CPI, and the GDP measurements are goal seeked, and misleading. This is likely the same.

  14. Interesting statement in the middle of that essay
    “People refuse to recognize a problem, he said, until they see its solution.”
    I think that is a big part of global warming denialism. The fact that nuclear power is the solution has been deliberately obscured, so many who see through the false solutions of wind & solar, don’t want to admit that there is a problem.

  15. There used to be a lot of talk about automation making for 4 day work weeks, six hour work days, etc. but that hasn’t happened.

    The reality is that it is much cheaper, given benefits packages and labor saving devices, to have one employee work as much as they were or, typically, even more, as opposed to keeping two employees on the payroll. Granted, increasing automation tends to mean they need more skills, some in automation, of course, but also in different areas, even if the total knowledge and expertise they need in each of those areas is less that it was without automation.

    This has the net effect of making these employees more valuable, which might or might not mean they make more money but, typically, they still won’t cost nearly as much as the two lower-paid employees together. The rest of that difference goes to the capital providers (the folks that own the automation). Which fuels additional wealth and income inequality, some of which is necessary but, at extremes, is as destructive to society and the economy as extreme socialism. Humans can’t handle extremes.

    With more labor freed for unskilled jobs, their cost stays low and can, for a time, compete with automation, delaying the inevitable. Try to get a job doing bookkeeping in a paper ledger, and refusing to use a calculator. Even if you worked for free, you could not.

    This full employment thing going on right now could very well just be the last high tide before the sea level starts dropping.

  16. Interesting essay, parallels much of my own thinking on the topic.

    Humanity was born in a world where labor was the sole input to production, and is headed towards a world where capital is the sole input to production. Seriously, we’re within a stone’s throw of being able to build Von Neumann replicators, factory complexes that can handle the entire productive process from resource extraction to shipping the final product, (Including more such factories!) without human labor.

    At that point the fiction that human labor is the only input to production is going to get pretty hard to sustain, along with the economy that’s built around it. The only real, sustainable answer is for us to all become capitalists.

    It’s how to achieve THAT, that is the pressing challenge that faces humanity in this century.

  17. No, labor productivity rarely improves much, because humans aren’t changing much. You can improve it by introducing specialization, or by training, but it’s still the same old humans, and they’ve got a hard ceiling on how fast they can move, how much they can lift, how fast they can think.

    You can get around that with machines, but then you’re talking machine productivity, not human productivity.

  18. The problem with automation and labor productivity is the tendency for labor to migrate from work that automation made more productive to work that automation has not impacted as yet. Because of this labor productivity rarely improves by much.

  19. Um, the 1.0% growth from 2017-18 seems to be multifactor productivity, not labor productivity. (Also, the 0.4% figure is 2007-18, not 2007-17.)

  20. Short term vs long term. In the short term, a tight labor market reduces productivity because employers can’t be as selective about who they hire, or pressure their employees as much to work harder.

    In the LONG term, this leads to investment in production equipment, that raises productivity again. But not, (See above!) “labor” productivity. Machine productivity.

    But you can hire people this afternoon, the capital improvements take time.

  21. Basically because electrons don’t have rights, and desert your chip if they don’t like the result of your designing.

  22. Speaking as a tooling engineer, I’ve always disliked this “labor productivity” term, it’s very deceptive, and often leads to wildly incorrect reasoning.

    Labor productivity, the amount of production due to a given amount of labor, rarely budges much. Occasionally it will increase as a result of increased specialization, (Adam Smith’s classic riff on making pins.) or training, but not a lot.

    What we’re usually calling “labor” productivity is actually capital productivity. The “capital” is producing more, resulting in a higher ratio of production to labor, but labor had nothing to do with the added productivity.

    If I add water cooling to a stamping die, so that the machine can run faster, it’s not the operator being more productive, it’s the machine.

    This is actually one of the reasons for wage stagnation: Labor is a declining input into production, and so gets a shrinking fraction of the output.

  23. To my knowledge as the labor market tightens, wages go up and employees are in short supply so employers are under pressure to do more with less employees.

  24. At a guess,

    • because bits can’t vote
    • transistors that never get used don’t decay and become useless
    • throwing the entire city away and buying a new one with architecture matching the current requirements is a bit more expensive than with a chip.
  25. I was under the impression that productivity often dropped as the labor market tightened up.

    The people getting hired now as unemployment shrinks are the people who couldn’t get a job before, when employers had any choice. Therefore they are likely to be the most inefficient, least productive of the workers.

  26. The suburbs & outskirts should be understood to be feeders for the city, and therefore need the throughput of a fast cache, so that people and goods can move rapidly between these places.

    Most city planning out in the Western US during the 60s and 70s tried to do just that. But businesses and workers had other things in mind. Their politicians followed those minds.

    This can be alleviated with PRT systems, which are scaleable over vast tracts of land, thus being the only option for suburbs for example.

  27. Why can’t we organize cities and their transport networks the way we organize microprocessors? The suburbs & outskirts should be understood to be feeders for the city, and therefore need the throughput of a fast cache, so that people and goods can move rapidly between these places.

    Driverless technologies can be phased in: drivers would mostly be in control of their vehicles, but for high congestion situations, a centrally-controlled autopilot would take over to allow coordination and overall faster flow of traffic. Perhaps such vehicles could initially be rented for commuter use, and then gradually evolve into a completely driverless Uber system with pay-per-use.

Comments are closed.