Melanie Zaber studies everything from the rise of artificial intelligence to apprenticeship programs in rural Appalachia. Her goal is straightforward: to help more Americans find meaningful work and a path to the middle class.
Zaber is a labor economist at RAND and co-director of the RAND Lowy Family Middle-Class Pathways Center. Her work has helped improve workforce training programs across the United States; helped steer young workers away from dead-end jobs; and helped show that the middle class is not always what people think it is.
Her latest projects have taken her to West Virginia. She’s working with state leaders there to shape new training programs, new career pathways—to give more people a shot at the middle class. It’s part of a new effort at RAND to join forces with rural leaders and residents, to address the issues that matter most to them. Zaber is helping to show what the RAND Rural America Partnership can do.
Your research has shown that making it into the middle class, and staying there, has become much more challenging. What would you change to help people who are just hanging on?
People don’t understand how precarious the lower end of the middle class can be. There are a lot of shocks that can kick people out: rapid inflation, unemployment, caregiving obligations, injuries, incarceration. You know the kid’s game Chutes and Ladders? Those are the chutes. We focus a lot on building ladders to the middle class. We don’t focus enough on keeping people away from those chutes.
What policies could help? Expanding leave for caregivers. Increasing pathways back into the workforce for folks who have had a career break. Expanding opportunities to work for people who are disabled. All of those are strong options for mitigating the impact of those chutes.
Speaking of potential shocks, what’s your advice for workers trying to make sure they don’t get left behind by AI?
Sometimes technology automates away jobs—but sometimes it expands productivity in a way that leads to more people getting jobs. The classic example is the sewing machine. Before it was invented, most people had one or two sets of clothes. They thought the sewing machine was going to completely decimate jobs for sewers and seamstresses. But instead, it made making clothes so cost-effective that employment actually expanded, because people could now afford to buy more clothes.
The challenge is trying to figure out where those opportunities are going to be. AI can’t do much in the physical world yet. Some of the roles that might be more resilient are technologist or technician jobs in health care or manufacturing. They don’t require four-year degrees, they pay pretty good wages, and they involve physical interactions that AI can’t do. AI might enable folks in jobs like those to take on more complex roles without putting them out of work.
People don’t understand how precarious the lower end of the middle class can be. There are a lot of shocks that can kick people out.
What’s one finding from your research that changed how you think about the middle class?
Some colleagues and I looked a few years ago at how we define the middle class. It’s usually defined based on income—it’s the folks at the middle of the income distribution, or folks between two income-level bookends. But when you think about somebody who’s middle class, it’s somebody who’s doing ok, who feels stable. They’ve got a little bit of savings. So that’s exactly how we defined it. We looked at how much of their income people have to spend on necessities like food, housing, and transportation, and how much they have left over to save.
And when you do that, you get a very different answer about who is in the middle class than if you just look at income. Our middle class was older and had significantly fewer Black and Hispanic people in it. We found that many people might look middle class based on their income. But when you look at how they’re really doing, based on the lifestyle they’re able to afford, then they don’t look middle class anymore.
If you could answer one question with your research, what would it be?
Do I have magic data? There’s been an explosion of workforce training options in recent years. It used to just be college, trade school, the military, or the school of life. Now we have noncredit programs, short-term certificates, long-term certificates, industry programs. Just imagine being a 17-year-old kid trying to navigate all that.
My dream project would be to identify the labor market returns for all these different pathways. But if I have magic data, I’m going to go a little further and get into working conditions, too. Do you find meaning in your work? Are there opportunities for you to learn and grow? I’d want to show what these different options lead to, not just in terms of ‘Did you get a job within three months,’ but in terms of quality of life.
How are you working to address some of these issues in West Virginia?
We’re working with folks in West Virginia to really understand some of the barriers that people face. Everyone there is very clear-eyed about what the challenges are and very motivated to try to address them. We’re working with them to strengthen apprenticeship programs, technical education, and career pathways. Our goal is to provide state officials with information that helps them get more people working. That makes this really rewarding. We know that our research findings are going to inform policy and get more people into jobs that are growing, that they find meaningful—and that aren’t overly exposed to AI risks.
