What you aren’t taking into account is secondary jobs created.

Uh, yea I am. Let’s look at the macro case first, then we’ll explode the rest of your fantasy.

In the scenario you explain, the tech explosion would lead to an increase in GDP (tech jobs are at the high end of the productivity food chain), THUS a decrease in unemployment WITH a substantial increase in mean household income.

So, three economic indicators, one of which can only be interpreted in light of the others. Where are they over the last decade?

  • GDP isn’t up; it’s in the trash. GDP under Bush, which was lackluster, still boasted a year at 3% growth. Obama was the first President since Hoover to never have a full year over 3% GDP growth, and most commentors have stated that getting back to the 3% level is….unlikely. (And they don’t tell you the worst part. To be able to sustain these low unemployment levels over time, we probably need to AVERAGE 3% — or close to it — going forward.)
  • Mean household income…..enough said:


  • The unemployment rate is down, but this is largely a function of the flaw in the algorithm; the rate drops when more people go to work (good), but also when people leave the workforce (bad). Plus, the metric doesn’t take into account if the job is full time, part time, highly compensated, or lowly compensated. So, if (a) the rate is down, and (b) wages do not rise, that means (c ) that the jobs created are either part time or low wage. Since we KNOW from the BLS data that the majority of created jobs have been full time, then we know that the improvement in the unemployment rate is a function of (d) workers leaving the workforce and/or (e) the creation of low wage job.

Conclusion: The rosy scenario you paint is incongruous with your theory. That said, let’s move to the specifics:

For instance, a company uses tech which replaces a certain number of jobs. Now consider that they will hire a few more tech people but probably not enough to displace the amount lost.

Understatement. Obviously it depends on the operation, but if you’re manufacturing and robotics, for every 100 people you idle you’re going to create perhaps one or two people-jobs who can program and fix the robot. Companies don’t buy robots if they have a record of breakdown. Know why? Because if a human breaks down (illness) I can bring somebody in, pay them overtime, and the line keeps moving. If the robot breaks down, the line stops, and lost revenue starts piling up. So, these things don’t go in unless they are highly dependable, and if they’re highly dependable, they don’t need much in the way of care and feeding.

If we’re talking about a small operation (say, a fast food franchise automating the burger making) you’ll get to dump a dozen workers perhaps, and they’ll be one guy working for the robot manufacture running around between maybe a dozen franchisees fixing the thing. Do the math.

If it’s IT, it’s even worse. I’ve consulted on no shortage of huge IT projects, and the sales guys always sell the project based on the opportunity for “worker reallocation”. Everyone in the room knows what that means; they get to fire loads of customer service and IT people. I’m no Jeff Bezos, but I am ashamed to tell you how many tens of thousands of jobs I’ve assisted in killing over the years.

However, there are third party vendors that provide software/hardware and as well as contractors. The more business is brought to them means more people they have to hire not just for the service they provide but also the sales team, licensing team, legal team.

Yes, I work for those guys. If you’re talking about VENDOR jobs, I would guess that we hire people at about the same ratio; maybe one new hire for every hundred or so jobs we kill.

I’ll also add that when we (and I work for a household name IT vendor) add marketing and administrative people, we don’t add them in the US. We add them in India and Southeast Asia, for pennies on the dollar.

Adding all of those type of jobs will also increase jobs at a service level including restaurants to handle more business lunches, increased revenue and likely jobs in hotel service for tech conventions. So on and and so on.

Yep. Since this is service industry, the jobs are nothing to write home about, you do add more. Instead of one or two people per hundred, maybe another five or six low paying job per hundred of jobs lost.

Personally, I think if you add 10 jobs for every 100 you lose due to automation, you’re lucky.

BTW……..don’t get too comfortable. We’re about 10 years from not needing anyone called a “database administrator” anymore. Machine learning will idle about 80% of those jobs as well.

Written by

Data Driven Econophile. Muslim, USA born. Been “woke” 2x: 1st, when I realized the world isn’t fair; 2nd, when I realized the “woke” people are full of shit.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store