Skip to main content


A corporation announces it's seeking bids from local governments to build a new warehouse or move its headquarters.

Policymakers tend to swoop in with this mindset: Let’s entice that company with tax breaks to set up shop here. It’ll create new jobs and enhance the overall health of our economy. Heather Stephens

But economic development incentives may do more harm than good, especially for middle-class workers, according to new West Virginia University research.

“Incentivizing the wrong industries can fail to lead to growth in employment locally in those industries and our research suggests this may potentially be contributing to the decline of middle-class jobs,” said Heather Stephens, assistant professor of resource economics and management in the Davis College of Agriculture, Natural Resources and Design. “For this research, we examined what happens to the mix of jobs within communities when you incentivize certain industries.”

Stephens co-authored “Incentivizing the Missing Middle: The Role of Economic Development Policy” with Carlianne Patrick, of Georgia State University. Their findings are published in Economic Development Quarterly.

The researchers drew on several primary data sources, including data on taxes and incentives from the W. E. Upjohn Institute for Employment Research, an occupation-based typology from the University of Toronto and detailed industry-level wage and employment data from EMSI, Inc.

From the data, they could break down the incentivized industry classes and their impact on jobs. For instance, do economic development incentives – such as property tax abatements and job tax credits — for industries that contain a lot of middle-class jobs affect jobs only within those industries? Or could the impact trickle over to higher- or lower-wage industries?

Given that there is no one definition of “middle class,” they use two different definitions. One that classifies industries based on occupations as creative class, working class (middle class) or service class; and one that classifies industries based on average wages as high wage, middle wage and low wage.

The creative class can include a wide range of occupations in science, engineering, education, computer programming, business and the legal sector. These job types require high levels of cognitive and problem-solving skills. Working class occupations rely more on manual labor and skills, such as in the construction, factory and trade settings.

“This is the first time anyone’s tried to look at the distributional impact,” Stephens said. “In the past, we’ve seen studies on whether economic incentives lead to overall job growth. While incentives can lead to new jobs, those could be only low-wage jobs. Yet it still looks like you have more employment.

“Or you could be giving away substantial tax dollars to create a few high-wage jobs, but you might see overall employment losses.”

A tale of two cities

Comparing employment trends in San Antonio, Texas, to Birmingham, Alabama, shows anecdotal evidence of how these incentives play out in the United States.

From 2000 to 2015, the employment index steadily for San Antonio. During that same time, the city shifted tax incentives from creative-class industries to working-class industries, Stephens said.

“In the early 2000s, San Antonio was incentivizing industries in the creative class more, like your Amazon-types,” Stephens said. “Then they switched, and by 2015, working-class industries faced more incentives. This seems to have paid off.”

On the opposite end of the spectrum, you have Birmingham. At the start of the decade, the city incentivized the working class more but has since focused its tax breaks more on the creative class. There, total employment dropped, according to Stephens’ findings.

“It’s an anecdote, but it’s an accurate representation of the bigger picture,” she said.

‘Giving away the farm’

Though middle-wage jobs comprise 60 percent of employment in the country, high-wage industries face the lowest tax rates (or are incentivized more) – and the gap has been increasing, according to Stephens.

Her findings reveal that raising taxes on creative-class and high-wage industries do not lead to negative employment effects on any class of industries and doing so actually increases employment among the working- and middle-classes.

She hopes policymakers take note, as tax incentives emerge from state and local governments.

“Policymakers really need to think about whether these incentives are paying off,” Stephens said. “Not just whether they’re creating jobs—but are they inadvertently contributing to the decline of the middle class, which they say they want to support?”

She referenced Amazon as an example. In 2019, the Virginia Senate approved tax incentives of up to $750 million for Amazon to build a facility in Arlington. More recently, in February, policymakers in Delaware signed off on plans to give Amazon $4.5 million in subsidies for a warehouse.

“Our research suggests there's no reason Virginia should have given Amazon all of those tax breaks because they’re going to have the same amount of high-end jobs before and after,” she said. “If anything, it could be that Virginia may see a decline in middle class jobs afterwards. And they’ve just given away their tax revenues. It’s like giving away the farm.”


Title: Incentivizing the Missing Middle: The Role of Economic Development Policy


Read More: Tax incentives for businesses could contribute to the decline of the middle class, says WVU economist