Why Economists Are Bad At Economics
by Sandwichman
Why Economists Are Bad At Economics
If you want an informed, thoughtful analysis of the effects of robotization on productivity, investment, employment and wages, ask an art school professor. Jason E. Smith’s two-part series, “Nowhere to Go: Automation, Then and Now,” (part one and part two) at Brooklyn Rail cuts through all the statistical aggregation category errors to highlight the accumulation dilemma that economists just don’t seem capable of either grasping or admitting.
Current speculations on both the promise and threat of automation are confronted with an ongoing crisis of accumulation. In this climate, a fragmentary implementation of automation is unlikely either to liberate large fractions of humanity from work, or produce mass unemployment of the sort envisioned over and again by commentators for the past century.
Smith has a PhD in comparative literature, which perhaps explains why he can tell the difference between an arbitrary catch-all term like “service sector” and the very different and distinctive components that are lumped together in it. Both promise and threat evaporate when the components are dis-aggregated and the jumble untangled. Definitely worth a read.
Good articles (Part II especially).. thanks.
It has been no mystery that productivity gains in goods production have shifted employment to services at lower wages/benefits.. this has been on-going since at least the 1940’s (post war) in the US.
It has also been no mystery why outsourcing of peripheral overhead jobs by business to smaller specialized businesses (payroll, maintenance, HR) has been occurring either — they are lower productivity growth rate jobs so not the best bang for the capitalist buck to increase profits… e.g. low roi contribution.
Services employment in turn has been increasingly displaced as well by automation: switch-board operators, travel agents, secretarial services, legal staffs, bank tellers (contrary to some measures.. which I’ll get back to below).
Yet it is quite true that many services aren’t amenable to *direct* (emphasis, direct) automation — hair stylists, barbers, message therapists, sexual services, manicurists, physical therapists, medical doctors — general practitioners, Ob-Gyn, chiropractors, dentists — nurses, dental assistants, technicians, teachers, cooks, dishwashers, waiters/waitress’s, etc.
But I’d point out that barbers have automated to the extent that most shops are now much more efficient, more hair cuts per hour per hair cutter utilizing trained and licensed barbers (mostly females in part-time employment at far lower wages than traditional barbers, yet I pay nearly as much for the same hair cut I used to pay (net of inflation).
Restaurant dishwashers use highly automated machines and methods today to clean & dry far more plates, glasses, and utensils per unit labor hour than ever, with less water consumed, and in less floor space too boot. Dental technicians — who clean teeth &/or take x-rays — use vastly superior tools today for cleaner teeth in half the time. Dental x-rays take 15 seconds of a dental tech’s time than just 5 years ago, much less 10 years ago.
Which get’s me back to bank tellers.
I go back several decades to my direct banking experiences… mid 1960’s when I began college. Tellers were personalized clerks for each customer’s specific needs. There were always at least 5 – 10 tellers active with a long que waiting. Fast forward.. two or at most 3 tellers, a que of up to 3 at most 4 people waiting. ATM’s replaced tellers. That Banks now employ former tellers as agents for other banking activities —taking loan applications mostly used to be done by banking loan officers.. who not only took your application but helped fill it out for you, and then in the same room the loan was approved with rates and terms by another loan officer, manger. The former teller’s as loan application takers are not doing with the loan officers did before, nor does the local branch manger have any authority to authorize a loan — that’s all handled now by a central computer at regional headquarters… taking several days (I know, I just got a new home equity loan with near top of line credit and a 40 year history of both loans, checking, and savings with the same bank).
I spoke with my 90 year old uncle about this .. he was a loan officer at a major bank in the early 40’s through 1950’s. He said a typical loan took him at least 3 hours of his or his secretary’s elapsed time and often much longer. He could process an average of 2 loans a day… so call that 4 hours labor per loan officer. Now it takes an hour or two for a person to fill out their own application, & provide all the supporting documents just to start the loan application process. It takes 15 minutes of a loan “agents” time to review it and type the critical information into a computer, which is transmitted wirelessly to the central office. where it’s processed by a computer and a legal “agent” looks over the details in 10 minutes before hitting the ‘ok’ or ‘not ok’ button. So 4 hours labor (minimum) is now down to 15 minutes of a low wage former teller’s time, + 10 minutes of a legal “agent”.. not a lawyer who probably get’s paid an average median income/benefits wage… equivalent now to ~ 20 minutes of a former loan officers costs… that’s approximately 1/12th the cost it used to be.
Is my loan rate lower though? No. Did I spend less of my own time and effort? No. Did my loan get approved / denied faster .. No.. it took a week this time.. twenty years ago I got a home equity loan at the same bank for the same total loan amount that took 10 minutes of my time (my last year’s tax return form 1040 was all I needed to provide then).
So economists can try to “measure” the impact of ATMs on number of “tellers” but total number of bank employees including loan officers and tellers per capita is a lot lower than it was pre-ATM’s.
So the upshot of the article is that services will be replaced by automation in many more than just 6% or 9% of jobs as tasks are separated more and more into the tasks comprising the jobs, reducing therefore the time per job or stating it differently providing more “services” per unit time. This will stimulate employment for more low wage / low skill jobs to be sure… which sort of reminds me of the 17th and 18th century aristocrats’ personal house-hold servants, where at least then household servants were well fed, clothed, and boarded, if they weren’t slaves (another issue entirely).
Somehow this just doesn’t add up to an improving standard of living except perhaps for the top 20% .. or top 10%… while the remainder suffer reduced standards of living..
And don’t forget retail service jobs… Amazon and on-line retailing in general… reduces retail jobs while increasing delivery services (UPS for example), though not by a 1:1 ratio. so the net is fewer retail and retail dependent jobs all around.
If you measure sales/employee for Amazon or any on-line retailing business there’s not even any comparison… it’s far more cost effective and efficient for on-line v brick-mortar sales… including real-estate costs, employees, and transport/storage. Jobs reductions and skill reductions overall have been huge… and that’s still ongoing.
More and more retail outlets are using cashier kiosks… self service cashiering… one retail clerk watches 6 kiosks. That’s 5 fewer retail clerks at peak shopping hours… probably averaging 2 fewer at least overall.. 1/3’rd fewer employees as cashiers. (not yet as much in grocers, but coming soon to a store near you.. in a few years)..
How about skill levels of cashiers. They don’t even have to know how to add or subtract now, and if they do have to do it they call their supervisor over to do it for them. .
And now let’s bring in future autonomous vehicles;.. so those UPS derivers are out of a job, as are every other delivery service and transport of goods and people jobs.
Take it a little further into the future (and Amazon’s already experimenting with this so it’s not pie in the sky futureistic). Imagine walking into a store, slipping a card or wireless device into or passing over a reader (which ID’s you as a unique shopper and also has your credit card, home address, name and telephone numbers (cell and home if there are then even any such things as a home landline phone), the pointing to or touching the objects you want to purchase and then just walking out of the store after “oking” wirelessly your purchases.
The purchases can be picked up on your way out or just delivered to your home via autonomous vehicle while you continue your shopping or other activities. Imagine some aspect of this at grocers for example.. perhaps fresh veggies and meats excluded (because people like to pick an exact piece of fruit or vegetable or meat). None of this is out of the range of present technology… autonomous vehicles excepted for a couple more decades or maybe even 50 years until it’s the only form of transport (or nearly so).
So how will services jobs be then? And that’s just in one segment of it.
My point is that we’re not being proactive in how society will deal with these things… or even considering that we need to be paying closer attention to how or when to begin dealing with them…so the restructuring will be helter-skelter, major disruptions to at least half of societies over a relatively short span of time (relative to legislative processes)..
What we’re doing societally is simply arguing about whether these things will occur with increasing frequencies and effects… it’s the bau crowd believing we can keep things as they are now or make the adjustments on the fly without having to worry about negative consequence.
Reminds me also of Climate Change denial.