It’s no secret that we’ve observed a rise in wage inequality over the past two decades. Although blue-collar wages and employment in manufacturing experienced a substantial bump over the past three years, the middle class has been hollowed out: Workers in the middle of the skill distribution have seen the weakest growth in wages and jobs.
Understanding the source of these changes in the labor market is a prerequisite for producing effective public-policy prescriptions. Otherwise, any “solutions” may end up being counterproductive.
In a recent working paper released through the Mercatus Center at George Mason University, Georgetown professor Alberto Rossi and I investigate changes in the labor market for financial services, focusing on the rise of science, technology, engineering, and mathematics (STEM) workers. STEM employment grew by 22 percent between 2011 and 2017, exceeding growth in any other sector besides professional services. STEM jobs are associated with large wage advantages and poised for further growth over the next decade — almost three times as much as non-STEM jobs.
These figures represent a substantial shift in the demand for skills within the financial-services sector. Employment for bank tellers, on the other hand, declined by 10 percent during that time — from 533,650 to 481,490 workers.
These patterns raise a question: Were changes in financial-services jobs the natural result of technological development and competition, or something else? Although we investigate three potential theories in the paper, we find that the rise in STEM employment is linked with a rise in regulation. Financial services, more than any other industry, experienced a surge in regulatory restrictions between 2011 and 2017, driven largely by the passage of the Dodd-Frank Wall Street Reform and Consumer Protection Act in 2010.
Using data spanning every occupation over time, we show that a 10 percent rise in regulatory restrictions is associated with a 5.3 percent rise in STEM employment. Increases in regulatory restrictions are also associated with declines in lower- and middle-skilled jobs. That’s important, given that non-STEM jobs have historically served an important role for the middle class, creating opportunities for upward mobility and family stability. This marks one of the important unintended consequences of greater regulation.
Unlike prior studies that have sought to quantify the effects of regulation, our analysis uniquely isolates the responsiveness of STEM employment, relative to its non-STEM counterparts, to changes in regulation within the same sub-sector over time. This helps avoid concerns about spurious factors like overall changes in technology or a growing demand for the digital workforce.
What explains the link between regulation and STEM employment? Not surprisingly, we show that increases in regulation are associated with greater compliance costs. In this sense, the data suggest that firms, especially in financial services, hire STEM workers at least in part to automate more of their organizational activities, which reduces the scope for human error and raises the overall value of the business. In fact, according to some estimates, the market for regulatory technology (or “RegTech”) is expected to grow from $4.3 billion in 2018 to $12.3 billion by 2023.
In sum, the surge in regulation accelerated the shift toward STEM employment in financial services, adversely impacting many lower- and middle-skilled workers who traditionally relied on these jobs.
These results highlight the importance of thinking through the unintended consequences of regulations before enacting them. Indeed, even if your priority is to mitigate inequality, these results show that the rise in regulation adversely affected the very individuals that it aimed to help. On the other hand, regulatory reform that focuses on removing unnecessary costs works in the other direction: It leads to increases in economic growth and wages across the distribution.
If the financial-services sector is going to reap all the benefits of emerging technological change, such as the application of artificial intelligence, it needs to be agile enough to draw upon the right mix of skills. Regulatory policy marks one of the important differences between the Trump administration and a potential Biden administration. Whereas regulatory restrictions grew rapidly under Biden during the Obama administration, they have declined for the first time ever under Trump.
As this election season moves along, we need to think more deeply about how public policies ultimately affect workers. Policymakers can sometimes dress legislation up to look very nice on the outside, but we only find out about the negative effects in the years that follow. Let’s try to avoid the latter.