In my review of Norbert J. Michel’s Crushing Capitalism, which I quoted from yesterday, I didn’t have room to highlight 2 other striking parts.
First, in a section on why technological improvements in productivity have not caused net job loss, he has a discussion of ATMs. He writes:
ATMs automated some of the basic tasks that bank tellers performed and lowered the cost of running a bank branch, which allowed banks to open more branches. Thus, ATMs displaced some bank tellers and other bank employees. From 1990 to 2010, the number of ATMs installed in the United States went from 100,000 to more than 400,000, and the number of bank tellers employed rose from about 500,000 to almost 600,000.
Michel also argues, as do Phil Gramm, Robert Ekelund, and John Early in their excellent book, The Myth of American Inequality: How Government Biases Policy Debate, that the welfare state has sapped the incentives of many people to make more income. He writes:
An example from a Pennsylvania Department of Welfare study puts these disincentives in very stark terms: “The single mom is better off earning gross income of $29,000 with $57,327 in net income and benefits than to earn gross income of $69,000 with net income and benefits of $57,045.”
READER COMMENTS
Andrew_FL
Jun 21 2025 at 8:54pm
Bank teller employment has declined since 2010 though, probably due to increased use of online banking services and fewer banks.
Craig
Jun 22 2025 at 11:23am
I’m struck by the number of bank branches and mattress stores myself.
Monte
Jun 22 2025 at 12:39pm
Maybe over the period in question, but Michel’s conclusions conflict with more recent data on the impacts of automation on labor. Autor et al found that technological progress since the 80s generally led to net job loss in the U.S., even as it increased productivity overall.
See also: Silent layoffs are trending as A.I. takes over corporate America.
Robert EV
Jun 29 2025 at 5:14pm
500k to *almost* 600k is about a 20% increase. The total US population during that time went from about 250M to about 309M, which is about a 23.5% increase. And the average age increased as well. Doing my best to adjust for percent of the population that would use a bank I’m subtracting out the 19 and under age groups from the 1990 and 2010 population pyramids from the populationpyramid website:
1990 – 181.28M
2010 – 227.93M
Which is a 25.7% increase in the bank using population.
So the absolute increase in bank tellers is off by a relative 20-25%. As a bank user I’m not complaining as to the current state of things, but you need population adjustments to make this argument.
Robert EV
Jun 29 2025 at 5:17pm
And I’ll add, how much of this is the “welfare state” and how much is from benefits cliffs? Those cliffs are poison pills (and I’m betting they were put in as poison pills, at least in any bills passed into law since pocket calculators became common).
Comments are closed.