Notes from Econtalk with James Besson (May 23, 2016)

Link here.

TL; DR: Employers raise wages only when they must do so to keep workers. Workers threaten to leave only when their skills are useful at other firms. Thus standardizing skills across firms gives workers more options and bargaining power. This forces employers to raise wages to keep workers.


Wages did not rise in textile mills for a while after productivity rose.

  • Even though the workers were producing more stuff with the looms and machines, their wages weren’t rising.
  • First textiles mills were in 1817, and wages did not go up for textile workers until after the Civil War (roughly half a century).
  • Sidenote: Even though the machines were technically producing the higher output, the human tasks were still critical. If the machine broke down, humans had to fix them. Humans were bringing something different, and even more valuable, to the production process. They brought an essential technical skill, which they had to learn on the job, to the process.

Why didn’t wages rise immediately?

  • Because workers’ skills were not portable; they could not threaten to leave because their skills were not valuable at other firms.
  • Each firm had their own system, and even experienced workers at new firms had to learn from stractch. Since workers could not threaten to leave, employers did not need to raise wages to entice them to stay.

After the Civil War, standardization started taking place.

  • Mills coordinated their machinery and training schools developed. Thus a robust labor market developed.
  • In 1830’s, very few new hires had any experience. By 1880’s, almost all new hires had new experience.
  • With the developed labor market, Mills would bid up the wages. In the competitive market, workers could apply their skills at many other Mills. Workers could threaten to leave for another firm if they weren’t paid enough.
  • Therefore, mills bid up the wages to attract and retain workers. This set wages on their upward trajectory.

The underlying labor economics:

  • What sets your wages is your next-best alternative, combined with your productivity.
  • If there are lots of people who can work a power loom, and it doesn’t take very long to learn how to do that, then you’re not going to get paid a lot.
  • If operating a loom becomes more complicated, then workers with the skills to operate the loom could command higher wages.

But operating a loom didn’t get harder. So why did wages rise?

  • Workers were being paid what they were worth. In 1870, workers could threaten to leave, because they could find jobs elsewhere.
  • In 1830, you couldn’t leave a mill and find work elsewhere. Other mills did things differently, so the experienced workers had to be trained from sctrach. Workers had firm specific human capital.

But really, what changed that made workers so much more valuable?

  • The industry became more competitive. A residnetial labor force grows; large dorms on-location at the mill.
  • RH: I don’t see the answer.

Present Day

Manufacturing in the US is thriving in terms of output, but declining in terms of employment. This is driven largely by productivity; machines allow output per worker to be sufficently high, and demand for workers is relatively low.

But the fact that machines make workers more productive, or eliminates them, doesn’t always decrease employment in an industry.

  • For weavers, as the price of clothe went down, people demanded more clothe; the demand was elastic. People demanded more clothes (they had very few clothes before).
  • The rising demand for clothe increased enough to offset the labor-saving effect of the new technology. Even as machines replaced human work, the demand increased enough to still increase employment overall.
  • This ended around the 1930s, when demand for textiles gets saturated. People had enough clothing, even if the price dropped further.
  • Then we see a long decline in the relative employment in textiles, then absolute employment around the 1970s, driven almost all by technology.
  • We see this in agriculture too; productivity is going through the roof, and after a while just don’t need as many farmers. Which is fabulous, unless you have a lifelong dream of being a farmer.
  • For the rest of us, who like to eat, it’s great (until we reach an obesity problem).

A similar pattern happened for bank tellers and ATMs.

  • ATM machines appeared around the mid 1990s, and everybody assumed this would eliminate the bank teller job.
  • But it didn’t. Since 2000, teller jobs rose, and rose faster than the labor force as a whole.
  • This is because the overall demand for banking services increased.


Non-compete agreements prevent employees from working in a similar field for a certain amount of time after they leave their firm. Employees sign them before they leave their current firm.

Why employers want non-competes:

  • Employers train and invest in the employee in some specific skill. Employers want the employee to stay and use the skill at the firm, not take the skill to another firm.
  • Non-competes raise the incentive for employers to provide the training.


  1. Employess have less incentive to learn a new skill, because they cannot bring the skill with them to another firm.
  2. Non-competes prevent knowledge from spreading from firm to firm. This hinders newly emerging technologies from developing.
  3. Non-competes make it difficult for new firms to start up because they cannot hire the skilled workers they need because those workers are prevented by non-competes. This hurts innovation.

Route 128 is an example of a high-tech area which fell behind partly due to non-competes. Silicon Valley, on the other hand, enforced non-competes much less, leading to greater innovation.


Advice for non-technical, non-mathematical workers:

  • Not everyone can do advanced math or code at a high level. So how what the advice for them? And what should our school system be doing for them?
  • Need to be comfortable with computers, be able to think quantitatively, and need to be able to learn new systems as the systems continually change.
  • Learn new systems as the systems continually change.
  • People need to be life-long learners.
  • And schools don’t know how to make students like that.

There will be some non-technical skills that are going to be more important:

  • Human skills like empathy.
  • Jobs will increasingly require human aspects, and at the same time, be able to relate to computer systems. That’s the combination needed.