Would it surprise you to know growth in startup companies in the U.S. not only has stagnated but has declined since 2000? Covering Silicon Valley as I do, it certainly surprised me. But that’s precisely what Fortune’s Geoffrey Colvin reports in the current issue of the magazine. In his article, “Startup Stagnation,” Colvin explains that headline-grabbing “unicorn” companies like Uber and Airbnb are the exceptions, not the rule.

The trend is disturbing. Startups exert a powerful influence on the economy because they generate a disproportionate number of jobs. Also, as Colvin writes, a decline in startup activity is “bad news because the rise of new companies is an important way in which capital and labor get reallocated from low-productivity uses to high-productivity uses.” One of the researchers Colvin cites recently gave a presentation called “Where Has All the Skewness Gone?” Clever, right?

A couple of other things surprised me about this report. The first is that Silicon Valley’s resilient startup scene doesn’t make that big of an impact on the macro-level numbers. The second is that the researchers don’t know what is causing the decline. That doesn’t stop Colvin from speculating on the outcome. “One can’t help imagining,” he writes, “that maybe these trends are contributing to several intractable problems in the U.S. economy: stagnating wages, long-term unemployment, low productivity growth, and overvalued unicorns as VCs compete for ever fewer high-growth startups.”

Startups and philanthropy.

Were this trend more widely known it could easily become a political football, especially in the hands of the anti-regulation crowd. And they might have a point. It isn’t likely Donald Trump would pay attention to something as nuanced and sophisticated and fact-based as academic economic research. But it’s quite possible this data explains the resentment of the voters who support him.

A healthy economy needs oodles of startup businesses, not just the mega-cap lottery-ticket winners of Silicon Valley. An economy without vibrant startups, no matter how attractive the headline-level data, harbors giant pockets of discontent.