In what was arguably the most consequential manufacturing announcement in recent decades, Intel recently picked a county just outside Columbus, Ohio, for the site of what is to be the world’s largest semiconductor plant.
The choice is important for what it reveals about the business of attracting businesses — and what actually affects business-location decisions in 21st century.
To attract the kind of high-paying, advanced manufacturing jobs, cities and states need an abundant share of college graduates, a steady flow of new graduates and communities in which these workers will desire to live. That’s a sobering lesson for states like neighboring Indiana, which can compete – or even beat – Ohio on tax breaks, tax rates and regulatory environment, but falls short in quality of life and educational attainment.
The most public element of Intel’s
choice are the tax incentives offered by Ohio. These consist of an initial $1 billion infrastructure outlay for the site in Licking County, just east of Columbus. The second incentive is murkier. To land this $20 billion plant and the 3,000 expected workers, the Ohio legislature doubled the timeframe on incentives tied to its 15-year Commercial Activity Tax. This tax is assessed on a firm’s revenue, and the incentive now gives employers a 30-year income-tax break.
For Intel, the incentive, based on targeted employment and the potential revenue for this plant, would be an annual tax credit of close to $30 million per year. In present-value terms, the total incentive package is worth between $1.3 billion and $1.5 billion over 30 years. That translates into incentives of between $430,000 and $500,000 per job.
At first blush, this is a deal that compares to the debacle of the Foxconn incentive package in Wisconsin. But there are several key differences. First, unlike Foxconn, Intel has no track record of reneging on development deals with local governments. Second, the Intel jobs will pay an average of $125,000 in salary and benefits — more than twice that at Foxconn.
Finally, the Foxconn plant never, ever made economic sense. Onshoring the production of a product that could be made for the fraction of the cost overseas always branded Foxconn as a doubtful enterprise. In contrast, the Intel semiconductor plant can be economically justified for several reasons.
The COVID-19 supply-chain interruptions argues forcefully for North American manufacturing firms to diversify their semiconductor supplier locations. The demand for semiconductors will only grow, since anything that uses electricity and is more complex than a 1970s toaster or lamp requires them.
Ohio vs. Indiana
This factory is a 25-minute drive from the College of Engineering at Ohio State University and close to the fastest-growing parts of the Columbus metropolitan area. The entire metro area has absorbed some 130% of the state’s population growth since 2000 .
The salary levels also suggest that the workforce at this plant will be primarily comprised of college graduates. Ohio workers in the semiconductor industry earned $65,490 per year in the last 12 months before the COVID downturn. To be profitable, this factory will be much more than the clean-room production facilities of a traditional semiconductor factory. I suspect this site will involve considerable product development and testing.
This evidence points to the need for a large number of college graduates as a driving factor in Intel’s decision. Close to a dozen top engineering colleges are within a five-hour drive. These include Purdue University, the University of Michigan, Michigan State University, Carnegie Mellon University, the University of Kentucky and of course Ohio State.
The only other Midwest location that could boast the same geographic concentration would be Indianapolis. The fact that Indiana was not chosen in this case offers a harsh lesson for states that rely on incentives rather than an educated workforce as an economic development strategy. It is the same lesson the Amazon HQ deal provided state policymakers around the nation.
The Indianapolis and Columbus metro areas are similarly size and have absorbed all their state’s population growth in this century. Both were finalists in the Amazon headquarters competition and were wooed by Foxconn as well. Purdue has an objectively better-ranked engineering college, and taxes in Indiana are lower than those in Ohio. Indiana’s use of tax abatements and credits suggest it would have offered a similarly sized package.
So why is Intel going to Ohio and not Indiana?
The short story is the abundance of educated workers in Ohio. The Columbus metro area is already rich with college graduates, but it also has the local environment that can attract more.
A recent Brookings Metro study measured quality of life in each Midwestern county and then compared that measure to population growth.
Licking County ranks highly on the quality-of-life measure and has seen significant population growth this century. In Indiana, no county with a higher quality-of-life measure has a higher share of adults with college degrees than Licking County. Several Hoosier counties ranked well on quality of life, but none of these offered the number of potential college graduates as did central Ohio.
Statewide, Ohio just does much better than Indiana on educational attainment.
In 2020, 29.6% percent of adults in Ohio had a college degree; in Indiana, it was 26.9%. That may seem like a modest difference, but it places Indiana in the bottom 10 states in both college graduates and those holding an advanced degree. Ohio ranks in the middle third on both measures.
Most troubling, though, is that Indiana’s share of adults with a college degree has been in decline since 2018, a factor that would immediately remove it from the long list of applicants for an advanced semiconductor plant.
The problem isn’t brain drain. A recent U.S. Senate study places Indiana and Ohio among the lowest states losing college graduates to outmigration. The problem for Indiana is that it attracts too few college graduates from outside the state and it suffers a low share of students heading directly to college. Ohio has steadily boosted its share and was a full two percentage points higher than Indiana in 2018, the last year for which data is available. That again may seem like a modest difference, but it translates into 3,600 more students per year heading to college in Ohio than in Indiana. That matters when you are looking to fill 3,000 new jobs by 2025.
Though Ohio hardly spends a lavish amount on schools, it has allocated $3 billion more to education than Indiana over the past decade. Ohio continues to spend a larger share of its GDP on schooling of all types. Ohio spends almost 20% more per child on education, or roughly $1,500 per kid aged 0 through 24 than does Indiana. That extra spending spending just paid off.
Admittedly, we cannot know all the details among the locations Intel considered. But Indiana’s governor is a far better spokesman and advocate than Ohio’s, and a recent study gave Indiana’s state economic development officials much higher accolades than those in Ohio. Other than education, it is hard to conjure any real difference between the suitability of Indiana versus Ohio for this factory.
The only meaningful difference between these two states is the availability of well-educated workers. Fortunately for Ohio, that was almost certainly the driving force behind the most consequential industrial expansion in the country in this century.
Michael J. Hicks is the George and Frances Ball distinguished professor of economics and the director of the Center for Business and Economic Research at Ball State University in Muncie, Ind. Follow him on Twitter @HicksCBER.