A few years ago, a couple of economists made a big splash by predicting almost half of occupations in the country had a big chance of being replaced by automation and technology in coming decades. Unemployment rates would soar to double digits and stay there. The economists’ forecast revived fears of “technological unemployment” posing a bigger threat to jobs than any future recession.

How out-of-touch those forecasts now look today. Companies are hungry for workers. Many are hiring unqualified people and then training them.

Unemployment rates are reaching historic lows among many groups and in many regions.

The job situation and facts have significantly improved from just a few years ago. A famous person (some say it was the 20th century economist John Maynard Keynes) once said, “When the facts change, I change my mind. What do you do sir?”

So, maybe it should be no surprise the dire projections of job loss due to automation and technology have been revised.

A new report now says only 20 percent of current occupations are severely threatened with extinction from non-human ways of performing work tasks.

Interestingly, one of the economists who was part of the dire prediction just five years ago is a member of the new, more optimistic, outlook.

What changed?

It’s easy to understand researchers are influenced by the times in which they work.

The research suggesting a potential 50 percent loss rate of occupations was originally done in 2013.

Although at that time we were a couple years past the Great Recession, the job market was sluggish. Many referred to the “jobless recovery.” It was easy to blame technology and automation.

In recent years, the economy — and especially the job market — have been performing much better. The previous labor surplus has been turned into a labor shortage. This certainly may be a factor behind the more optimistic occupation projections.

In this new reality, the updated analysis of the effect of automation and technology on occupations reflects a point many economists have made for decades. While technological advances can replace some jobs and occupations, they can simultaneously increase the need for others.

Automation and technology can be job-enhancing in two ways. First, the advances can lower business costs, thereby enabling companies to lower prices. The lower prices stimulate more buying and thereby motivate businesses to expand.

Here’s an example. Automation is coming to fast food restaurants with kiosk ordering. The kiosks replace workers at the counter taking orders (by the way, this was my first job 50 years ago).

But if the kiosks allow fast food restaurants to lower their prices, the restaurants will sell more, open more outlets and hire more workers — and potentially end up with higher employment than before the kiosks arrived.

Second, if automation lowers prices, consumers will have more to spend on other things, which in turn creates more employment opportunities.

So, the good news is the latest research suggests automation and technology may not be the job-killers they were thought to be — and they may even increase jobs.

Yet the research still predicts 20 percent of existing occupations are threatened by severe downsizing.

What are these occupations?

I can’t give you the entire list, but here are the tops ones: printers, financial clerks, food processing workers, manufacturing workers, sales reps, mechanics and several construction occupations. These are occupations where the increasing capability of technology will allow machines to soon replace many of the tasks done by people.

And — in the calculations done by the economists — they are occupations where the direct losses due to technology and automation won’t be counterbalanced by new spending from consumers.

Now let’s look at the other side — of occupations least threatened by technology and automation – and therefore most likely to expand. The top ones include teachers, engineers, counselors, entertainers and performers, scientists, architects and health professionals and technicians. These are occupations requiring human-to-human interactions that technology and machines are unlikely to soon replicate, or occupations requiring high levels of cognitive analysis and decision-making, or both.

The evaluation of how the evolution of technology will impact occupations and jobs is a continuing issue requiring continuing study. Individuals planning careers and educational institutions deciding on programs of study are both importantly impacted by this issue. The latest analysis of the interaction between advances in technology and automation and jobs is more optimistic than earlier studies.

Still, we shouldn’t be lulled into thinking no one will be harmed by our rapidly evolving tech-heavy economy. Even if only 20 percent of occupations will be taken over by technology — that is still a big number, and it will mean millions of workers nationwide and thousands in North Carolina will have to be retrained for different kinds of work.

Yet the news about upcoming technological unemployment is better than it was just a few years ago.

But will it continue?

You decide.

Mike Walden is a William Neal Reynolds Distinguished Professor and Extension Economist in the Department of Agricultural and Resource Economics at North Carolina State University who teaches and writes on personal finance, economic outlook and public policy.