> Wages reflect the labor portion's contribution to "input." We know that tech and automation's portion is continuing to increase in relation to labor. Therefore, with this understanding, wages must decrease in proportion with productivity.

Let's say I make widgets for $10 an hour, and make 100 widgets per hour. That's $0.10 per widget.

Then, a new tool comes along, and allows me to make 200 widgets per hour. The cost of a widget is now $0.05 per widget.

Who "deserves" this extra profit? Yes, my employer paid the expense for the new tool, but this is (more or less) a one-off expense. They will recoup this cost, and eventually the increased productivity will be pure profit,.

I, on the other hand, am still spending eight hours a day making widgets. My time is now worth twice as much as it was, but I'm still being paid the old rate.

The issue we're seeing with wage stagnation is that companies are making one-time investments and recouping increased profits forever, while employees are working just as hard, but seeing no increased benefit.

And that's just from a simple math perspective. This discussion entirely ignores the fact that when you employ someone, you are essentially renting a human being, and in order for our society to function, that cost must have a floor. Society as a whole suffers when you are able to rent a person for less than it costs to live.

> My time is now worth twice as much as it was,

no. your time is now worth half as much as it once was. the tool is what brings the increase in value, not you nor your time.

the owner can now replace you with someone who costs half as much and is half as efficent and get the same result as before the new tool was brought in. even with this decrease in time efficiency, the cheaper less efficient worker has the same value as you did before.

so lets say the owner still wants to increase efficiency but cut costs. ok, they fire you and hire someone else for 75% of your wage and 75% of your efficiency. the owner can still make the same amount of widgets as before in less time and for less cost.

your value has decreased, not increased.

I think a lot of people undermine the competitive nature and other risks that companies have to take in its pursuit of higher profitability only looking at a few with bad practices.

For example, a company may have to spend 1 year's worth of R&D to reach a new tool that would add 10% productivity to its current workforce. What % should go to increasing employees' wage? What are the risks of competitors finding similar productivity gains and not increasing employee's wage, leading to a competitive disadvantage to your company?

When a company hits a breakthrough that adds a multiplier to a part or all of its work, society, including its employees expects it to share more of their gains. What is the right balance? Who deserves the gains and at what %?

I do believe you have some fair points on the employee's side of things and that your last paragraph really hits the nail with a key problem society faces.

Let's say that you pay that hypothetical employee twice as much, so that his wage increases with the output. In effect, you've negated the technology's effect on productivity. In this case, wages would trend with productivity, but only because productivity would not increase.

The pot should get split three ways.

The toolmaker gets a cut. This usually happens in the process of buying the tool.

The buyer of the tool gets a cut. This happens, perhaps excessively.

And the worker should get a cut. The worker who knows how to use the new tool is worth more than the worker who only knows how to use the old tool.

It seems, however, that the workers getting a cut is where the process often breaks down.

That might be true, but it seems tech is not that meaningful. Usually technological change carries with it a lot of need for human labor. (Eg. first it needs to be evaluated, planned, implemented, supervised, maintained, repaired, etc.)

And it seems that tech displaces the middle income jobs. ( https://economics.mit.edu/files/11563 ) Which of course puts a downward pressure on wages, which probably contributes to the decline of labor share of income, but this effect is no as big as using a different deflator, nor is the technological change well correlated with the aforementioned decoupling.

It's weird how capital is assumed to take credit for automation. Automation is built by labor.

Yes, but not yours.

That is: when the Acme Widget Factory buys Jake's Improved Widget-Making Tool(TM), it's Jake's employees' labor that made that tool, not Acme's. It is therefore Jake's employees who deserve to be rewarded for creating the automation, not Acme's.

That's why I said "not yours" - not specifically you, crdoconnor, but the labor of the company that buys automation did not build that automation. They still should be rewarded for the the increased productivity that comes from knowing how to use it, but they didn't build it.

And when I said "they still should be rewarded", that doesn't mean that they get all the benefits from the increased productivity. The tool maker has to be rewarded, too, or there won't be many tools to improve productivity. And the capitalists who bought the tool have to be rewarded, too, or there won't be many such tools bought.

Yep but in the current economy it's not a shift of wealth from Acme workers to Jake workers that's going on. It's from workers of Acme and Jake to owners of Acme and Jake.

You're not wrong, but I think this is just ambiguity of terms.

'Productivity' in wage stagnation discussions is implicitly 'labor productivity', which is technically a "partial productivity" measure judging the amount of output generated by a unit of labor. It's not the same as overall output of value. It also isn't the same as "labor portion" in the equation you sketched, though, because capital expenditures like space or technology that increase worker efficiency still constitute labor productivity gains.

(The 'labor portion' there seems closer to what Marxists look at in the labor theory of value - the amount of time and effort expended by labor, isolated from other factors. You're right that that doesn't end up tracking productivity or wages.)

The idea that wages trend with labor productivity, then, is that anything increasing labor productivity increases the output value of an hour's work, which ought to increase the amount of money a worker can demand for their time. They won't capture the full rise in value output, because some of it will be consumed by capital expenses promoting that efficiency, but their wages should trend in the same direction.

Less hypothetically: let's say Cyberdyne's powered exoskeletons become standard for bricklayers, who can then work 20% faster. Since there are plenty of bricks to lay, buying an extra hour of bricklaying now creates more value, and remains profitable at a higher cost. Bricklayers won't simply collect all the extra output value because the amortized cost of the suits has to be covered, but we'd still expect them to get a raise. And if the new tech increased productivity at a lower capital cost per unit than existing tech, the percentage of output captured by labor could even rise.

Now, there can be out increases that don't involve labor productivity, anything from automating away jobs to finding cheap new sources of raw materials. People sometimes confuse the terms and compare GDP or overall productivity to wages, in which case you're right that there's no intrinsic linkage. (Although in the long run, labor's percentage returns tend to outweigh fixed-value changes like trade agreements.) But looking specifically at labor productivity, which should track wages, we still find a close trend seems to abruptly break around 1970.

Because we would usually expect employers to compete for workers up to the wage-level of marginal productivity. If you're underpaying your workers and we both have the same technology, I can lure them away by paying them more, until they start demanding wages higher than our tech can allow. To compete with me, you then have to raise wages too, and the market evens out at marginal wages equaling marginal productivity.

One of the Big Questions not covered by the article is: how competitive have labor markets been over time?