Millennials May Be the First Generation to Lose a Majority of their Jobs to Automation.

The consensus of most people who study labor automation is that the total number of jobs is not going to keep up with population growth as we have seen in the past – at minimum. However, job prospects will likely get a lot worse for those between the ages of 18-34 as time goes on, mainly because of the types of jobs that are easiest to automate.

Increasing industrial efficiency allows robots and other duty-specific machines to perform the work of 100 1960s factory workers with an average of about 23 workers in 2017. In the past, the loss of these industrial positions were more than offset by the rising service industry. Today, even these service industry positions are being taken by machines, leaving little else for the younger generation to grasp onto.

Half of All Jobs Gone

Oxford University researchers have postulated that almost half of U.S. jobs could be in jeopardy due to automation within the next 20 years. Those jobs aren’t going to disappear overnight, but they are already leading to businesses needing to hire fewer people to meet their staffing needs.

20 years ago, true labor automation was largely confined to heavy industry. Robot arms spot welding vehicle parts were novel and “self-serve kiosk” might as well have been synonymous with a vending machine. Labor efficiency had improved considerably since the middle part of the century, but new jobs sprung up to replace them. The internet was in its infancy and was mainly used to find limited information about your interests.

Today, the internet has put the world at our fingertips and eliminated countless jobs in the process. For a company like Amazon to provide what they do (at their relatively fast speed) to anyone in the U.S. would mean they would probably have to employ at least triple the number Wal-Mart already employs – 1.4 million – but they currently made do with about 269,000.

Even when you do go to Wal-Mart or Target, you have the option of going through the self-checkout lane. During most of the day, there is only one person in those lanes doing the work of 8-12 people. When these large retailers get their hands on machines that can stock shelves with close to the accuracy of a human (it is in development), we may see the evaporation of the position that makes up between 25-40% of employees at big box stores – millions of people out of a job.

Clerk, cashier, stocker, customer service representative, entry-level support, many jobs that younger people take while in school or just after graduation may mostly go the way of the Dodo. Ask retailers like Kmart, Sears and Macy’s how their 2016 holiday season went, automation and the internet are going to keep putting downward pressure on both major and micro retailers.

This is very likely the future we are going to be facing in the next couple of decades. While the number of jobs in the U.S. probably won’t be cut in half from 145.3 million to 72.7 million, we will almost certainly see the number of unemployed individuals rise, even as the economy overall will strengthen. The critical question becomes at what point will action be taken? Not action to stop automation but action towards developing what will likely be a universal income.

Great Depression

During the Great Depression, a period of time that brought about sweeping economic reforms, unemployment hit 25%. In October of 2009, unemployment was at 10% after the most recent recession. As history has shown us, American society has adapted and weathered a multi-year period of extremely high unemployment figures. The 2008 recession happened in close to the same societal and technological timeframe as we are in now, which can give us clues as to what automation-driven unemployment might look like.

December 2016 unemployment figures were released a few days ago and the numbers on the surface look relatively good. 156,000 jobs were added by U.S. employers (a .1% increase) but the unemployment rate went up by .1%. This likely signals a widening of the gap between the number of people and the workforce and the number of jobs available.

Population growth in the U.S. is currently at about .7%/year but going back 18-22 years (for millennials entering the job market today), growth was at about 1%/year and stayed close to that for nearly a decade. Coupled with immigration, we can probably expect to see about a 1.1% increase in workforce participation.

The U.S. economy has essentially hit a floor as far as the unemployment rate goes (likely in November 2016 at 4.6% thanks to increased hiring for the holidays) even as the economy overall is doing fairly well. What we probably should have seen for December 2016 was a hold at 4.6%, the small but significant uptick may be looked back upon as a sign of things to come.

Domino Effect

It only takes one domino to go down to start a cascading effect on the economy. One bad unemployment report could trigger a market sell-off and get money managers digging into the root causes. What they find they probably will not like, leading to further deteriorating market conditions. When things get tight for businesses many of them are going to start looking at what they can do to cut costs, what some will ultimately decide is that they can do without as many people working for them.

Millennials will likely bear the brunt of what has started off as a drop off in hiring compared to workforce population. Those who are older may be able to ride this out into retirement, however with people working into an advanced age, the lack of positions opening up due to retirements will become apparent. Once again, it will be the younger generation (teenagers already have a 14.2% unemployment rate) who are going to be facing dwindling job prospects.

Read More: The average Millennial credit score is 640 and people wonder why new college grads can’t buy houses. Take a look at your credit score for free.