In short
How the Bureau of Labor Statistics’ Occupational Outlook Handbook predicts “future-proofed” jobs, how those projections are made, and how accurate they’ve been over time.
Guests/backgrounds
Asa Hess Matsumoto, application security engineer and Georgia Tech teaching assistant; Sarah Mattson, BLS economist; Maxime Massenkopf, former academic now at Anthropic, author of a study on BLS projection accuracy.
Key claims
BLS projections start with broad labor-market trends (e.g., demographics), extrapolate historical industry/job patterns, then adjust for emerging tech/social changes (e.g., AI). BLS has a strong track record: correctly predicts whether occupations grow or shrink about 70% of the time; long-run forecasts correlate with actual outcomes.
Notable examples
Future growth leaders include nurse practitioner, solar photovoltaic installer, and wind turbine service technician; projected declines include tellers, desktop publishers, and bill/account collectors. Historical “miss”: telegraph operators were predicted to avoid heavy layoffs.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VOListener Inquiry - Job Prospects in AI
1:49 to 4:52
Discussion about a listener's questions on job prospects in the AI era, referencing the BLS's Occupational Outlook Handbook.
“My name is Asa Hess Matsumoto, and I work as an application security engineer.”
Evaluating BLS Job Projections
4:52 to 9:20
Analysis of the accuracy of BLS job growth projections and their historical reliability.
“Yeah, the BLS is bullish on both scams and cybersecurity students.”
Transcript
Automatic transcript. May contain errors.0:01NPR.
0:05This is The Indicator from Planet Money. I'm Waylon Wong. On this Labor Day Monday, we're naturally thinking about the American workforce. With so much economic uncertainty in the air, from AI to tariffs, it can be hard to know just what the future holds for our careers or the broader labor market. Luckily, the Bureau of Labor Statistics is something of a crystal ball in what's called the Occupational Outlook Handbook. Today on the show, we're re-airing one of our favorite episodes all about this obscure government report. After the break, Darren Woods and I get to the bottom of whether we can actually predict what jobs are future-proofed.
0:51This message comes from IXL. The first few weeks back at school set the tone for the rest of the year. IXL is designed to help kids build confidence and get up to speed fast. Covering pre-K through 12th grade, IXL offers personalized interactive practice in math, language arts, and science. One subscription covers everything for all the kids in your home. Receive an exclusive 20 % discount on an IXL membership at IXL.com slash NPR. This message comes from NPR sponsor Raymond James, a financial firm offering wealth management, banking, and capital markets services that are inspired by people. Before Raymond James financial advisors build plans, they build relationships so they can craft individual strategies designed to achieve priorities and pursue what's possible.
1:41That's the power of personal. Disclosures at RaymondJames.com. Raymond James and Associates, Inc., member NYSE, SIPC. The other day we heard from a listener. My name is Asa Hess Matsumoto, and I work as an application security engineer. That means Asa keeps our apps safe from hackers and viruses. And he also moonlights as a teaching assistant at Georgia Tech. He's used to fielding questions about job prospects. Fair, in this new AI vibe coding world. You have career changers, students, new graduates. And, you know, I try as best as I'm able to to provide like sort of empirically backed guidance.
2:22Asa says he finds one section of the BLS website quite useful for those hard numbers. It's the BLS's Occupational Outlook Handbook. There he can look up jobs like the one he does, Information Security Analyst. He sees things like its median wage. $125 ,000. And he can see where these jobs are most plentiful. Virginia and California. He can also see the projected job growth to the year 2034. 29 % growth. That's a lot of analysts. Yes. His job is safe. Well, this lodged a question in Ace's mind. How is that done? How is that estimate made? And has there been any follow-up to determine how accurate these estimations and projections have aligned with the real-world data once time catches up?
3:10So at the top of the BLS's list for future job growth is nurse practitioner, solar photovoltaic installer, and wind turbine service technician. Ooh, you did that once for a day, for an episode, remember? You put on your steel-toed boots, you climbed up a windmill. It was the height of the Statue of Liberty. Oh my gosh. And where I didn't go was the bottom of the BLS list, which is roles like teller or desktop publisher, bill and account collectors. Those jobs are projected to decrease. Sarah Mattson is an economist at the BLS. Sarah does the work looking at trends in the labor market. But, you know, she's not like a psychic.
3:48If I could make every exact prediction in a crystal ball, I would not be working here. I would have won the lottery by now. Sarah says the way the Bureau calculates job growth starts at looking at overall trends in the U.S. Factors such as the general aging of the population and general population change and stuff like that. Then Sarah's colleagues extrapolate the historical trends for jobs within the different industries. Like maybe home health and personal care aides are growing at a faster rate than, say, physicians and surgeons. But there's no amount of historical data that can project trends that don't yet exist in the historical data.
4:24AI, of course, is a great example of that. So that's where my team comes in and we apply our research to those initial estimates to kind of move them as we see appropriate for the emerging trends. So for our listener Asa's job, information security analyst, that's expected to grow. The BLS writes, cyber attacks have grown in frequency and these analysts will be needed. Bad for the world, but good for Asa's cybersecurity students. Yeah, the BLS is bullish on both scams and cybersecurity students. Yes. But that gets us to Ace's second question. How reliable have the BLS projections been? Economist Maxime Massenkopf has written a paper answering this very question.
5:07Maxime recently moved from academia to AI. He now works for Anthropic. With all the developments in AI, I got really curious about, will this job still be around? and how good would advice be based on these kinds of job forecasts? Maxime's research began by looking just after World War II. That's when millions of young veterans had returned home from Europe in the Pacific. The U.S. government wanted to help them decide what careers to pursue. So it made the first Occupational Outlook report in 1946. It had a really wide circulation. The circulation in the U.S. approached that of like a New York Times bestseller.
5:46And that popularity remains. It's the most visited section of the BLS website today. The report has been updated every few years since, and it highlights trends that might affect jobs, like when the U.S. highway system was being built out from the 1950s. These huge construction projects would unearth a lot of artifacts, and so the BLS predicted that demand for archaeologists would go up. In the 50s and 60s, the handbook predicted more veterinarians as more people moved to the suburbs. They'd have a lot more space and they'd be more likely to have a pet dog. Even trends for men having longer hair in the 1970s.
6:29The prediction for decreased demand for barbers. People didn't need to get their hair cut as often. Some assessments haven't aged as well. Like in the 1940s, the report offered a fairly benign outlook for telegraph operators. It said the downward trend of employment expected in the occupation over the long run will probably not be sharp enough to cause heavy, prolonged layoffs. That sounds like a missed prediction for me. So Maxime tallied up the overall scoreboard. He looked at all of the projections from the 1940s through to the 1990s, and then he checked them against how the jobs actually did down the line.
7:05These predictions are really strongly correlated with the employment trajectories. If you choose a job that the outlook was really optimistic about, there would be much higher growth in that occupation in the next 10 or 20 years. What Maxime found was that the top third of jobs that were projected to grow the fastest did grow a lot. They rose by 57 percent over two decades. The bottom third of projected jobs actually grew by only 12 percent. How much of this accurate record could have just been done by looking at how fast jobs are growing now and then assuming that they'll continue growing at the same rate?
7:40Maxime found that existing job trends did account for a lot of that forecasting power. Right. So if warehousing jobs are growing a lot last decade, they're likely to continue growing next decade. That might be due to the continued rise of online shopping. These things, at least historically, have just unfolded slowly. That said, Maxime says the BLS did add value. When you compare them head-to-head, the occupational outlook wins out a little bit. So if you had to just choose one source, you would still go with the BLS. The BLS does its own backwards-looking assessments of how accurate its projections have been.
8:15And Sarah Mattson from the agency says that the most recent evaluation shows that the BLS correctly projected whether an occupational group would grow or shrink almost 70 % of the time. And that's all available on our website for anyone to take a look at. People like our listener Asa Hess-Matsumoto. As we heard, he was wondering if he could rely on the BLS projections when talking to the next generation of cybersecurity analysts. Well, he can rest assured, the BLS has a pretty good track record. We told Asa what we learned. And I actually have an answer for you. Great. Yeah. We explained that the BLS takes long-term job trends and extends them.
8:57Then it boosts these up and down depending on new technological or social changes. And evaluations afterwards show they're reasonably accurate. What are your thoughts? I'm pretty impressed that it's that accurate given how far out that they cast and how even looking back at the last few years things have changed. That's incredible. That's really incredible. Do you want us to take your question? email us, indicator at npr.org. This episode was produced by Corey Bridges with engineering by Jimmy Keely. This is fact-acted by Sierra Juarez. Kate and Cannon edits the show and The Indicator is a production of NPR.
9:40This message comes from Capella University. That spark you feel? That's your drive for more. Capella University's FlexPath learning format lets you earn your degree at your pace without putting life on pause. Learn more at capella.edu. This message comes from Bowl & Branch. Change the way you sleep with soft, 100 % organic cotton sheets from Bowl & Branch. Designed to help you fall asleep faster. With airy blankets, cloud-like duvets, and breathable sheets. Experience pure comfort on night one and feel your sheets get softer with every wash. Discover the difference with 15 % off your first order at bowlandbranch.com with code NPR.
10:21Exclusions apply. See site for details.
From the publisher
Today on the show, we flip through the Occupational Outlook Handbook and answer your questions about the future of work.
This episode originally aired on May 8th, 2026
Fact checking by Sierra Juarez.
Your Next Listen
— How AI is shrinking the job market for teens
Connect with The Indicator
— Sign up for The Indicator’s weekly newsletter!
— Buy the Planet Money book
— Find our socials, YouTube and more!
— For sponsor-free episodes, subscribe to NPR+
Support public media with NPR+ and enjoy perks for over 25 podcasts like this one. This show’s perks include sponsor-free listening. Learn more at plus.npr.org.
See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.
NPR Privacy Policy




