Putting the Con in Foxconn Wisconsin: How Commercial Studies Have Justified the Transfer of Wealth from Taxpayers to Corporations

For Governor Scott Walker and his boosters, the project is “transformative,” a “once in a lifetime” opportunity. For citizens? Not so much.

Scott Walker (Gage Skidmore)

This summer, the roar of bulldozers replaced the sound of tractors over thousands of agricultural acres outside of Racine in southeastern Wisconsin. Road, water, and electricity infrastructure construction was well under way for a new industrial complex for Foxconn, the Taiwan-based technology giant. In the project’s master plan, Foxconn will eventually own 4.5 square miles of once-prime Wisconsin farmland, an expanse about a fourth the size of the city limits of nearby Racine, population 77,000. The main 24-inch water main will have the capacity to deliver more than 20 million gallons a day of high-quality Great Lakes water (home to 21 percent of the world's remaining supply), although initial usage has been promised to be much less. Foxconn has pledged to be good environmental stewards, even as Wisconsin has waived many of its basic environmental regulations in order to expedite the project.

The company has promised to build a $9 billion to $10 billion sprawling industrial complex which will employ “up to” 13,000 workers in jobs the Wisconsin Economic Development Corporation (WEDC), the state’s semi-private economic development commission, has repeatedly described as “family-supporting.” Most of the attention has been on the proposed first North America–based LCD panel manufacturing plant, described by boosters as a state-of-the-art automated factory that will producing next-generation large-scale screens for TVs and other consumer applications. More recently, Foxconn has confirmed reports from the Asian press that revised plans call for a more modest factory making smaller screens. Foxconn’s initial construction project scheduled to begin this year is even less ambitious, designated as a massive low-tech assembly operation, putting together TVs and other consumer products using imported screens.

Industry experts put the employee count for even the original large-scale LCD manufacturing plant at 2,000, just 15 percent of Foxconn's promised workforce. Foxconn’s original plans estimated that three-fourths of its workforce in Wisconsin will be hourly, a pattern that fits the company’s core expertise: assembling iPhones and other electronic devices. For 15 years, Wisconsin taxpayers will be subsidizing Foxconn’s payroll for employees making at least $30,000 to the tune of 17 percent, while capital costs over the first seven years of construction will receive a 15 percent rebate. Should Foxconn hit its capital expenditure and employment targets, Wisconsin taxpayers will end up sending some $3 billion to the company. While state-level support is touted as tax relief, in fact Wisconsin has already waived almost all of the pertinent taxes for businesses. The taxpayer-funded $3 billion in incentives (the largest ever to a foreign company) will be paid largely in cash. The local municipalities are chipping in another $764 million via a tax increment financing (TIF) project so large that the state legislature had to pass a waiver, since this amount far surpasses the statutory prudence level of 12 percent of existing municipal property values. Additional taxpayer-supported costs have put Wisconsin's total outlay at well over $4 billion, more than $300,000 a job. How expensive is this?

As Jeffrey Dorfman, economics professor at the University of Georgia, reported for Forbes in 2017: “Realistically, the payback period for a $100,000 per job deal is not 20 years, not 42 years, but somewhere between hundreds of years and never. At $230,000 [or more] per job, there is no hope of recapturing the state funds spent from taxes on the company and its workers.” Tim Bartik, an economist at the nonpartisan, independent W.E. Upjohn Institute in Kalamazoo, Michigan, who specializes in economic development, concurs: “The per job cost is at least five or six times, maybe seven or eight times, the incentive average.”

For Governor Scott Walker and his boosters, the project is “transformative,” a “once in a lifetime” opportunity. Wisconsin Speaker of the House Republican Robin Vos called it “an American field of dreams.” Wisconsin Lieutenant Governor Rebecca Kleefisch went on the record touting the deal’s value because “We did not meet the company’s ask. Their number was bigger than ours.” While Foxconn has described their completed complex as a self-contained city, with housing and food services for workers, Walker has preferred to call it the first step in a regional high-tech manufacturing boom he calls “Wisconn Valley.” This vision of something wonderful blooming around the Foxconn facility helps explain the blank-check largess of the Walker administration. It’s also the sort of wishful thinking that flies in the face of the evidence.

Timothy Bresnahan, Alfonso Gambardella, and AnnaLee Saxenian spent two years studying the origin of technology clusters for their 2001 paper “‘Old Economy’ Inputs for ‘New Economy’ Outcomes: Cluster Formation in the New Silicon Valleys” in Industrial and Corporate Change. They wrote: “[O]ur case studies clearly show the foolishness of directive public-policy efforts to jump-start clusters or to make top-down or directive efforts to organize them. … Direct, top-down policies are most likely to fail. Particularly worrisome are policies that would direct at a level of detail such as picking the specific industries or technologies to be sponsored.”

Expert Wall Street investors have trouble valuing corporations over the next calendar year. It’s difficult to project the ten-year viability of even well-established manufacturing plants making slowly evolving products such as automobiles. How is it possible to value the 20- or 30-year prospects of a plant making a product, such as high-tech LCD TV screens, subject to rapid creative destruction? After all, 15 years ago we were all watching bulky rear-projection TVs.

In the face of all of these questions, to say nothing of the growing body of academic work shedding doubt on the wisdom of mega-incentive deals, Wisconsin officials dipped into a familiar well for politicians, municipalities, and state governments looking to invest public monies in large projects.

They brought in the consultants.

Economic Impact Studies

The economic impact–study business in the United States is well-entrenched and has a long history. The field was pioneered by Nobel laureate Wassily Leontief when he was serving as a consultant to the Office of Strategic Services during World War II. At the end of the war, he used the same sort of input/output model that is widely used today in commercial studies to examine the likely impact of the decommissioning of the weapons industry. In doing so, he not only established the power of these models, but their limitations, as he issued a warning that the United States faced a second Great Depression as a result of lost employment and production.

A small circle of firms routinely produce reports (all of them positive) examining the economic consequences of all kinds of major tax-dollar spending. Sports stadiums, convention centers, manufacturing and headquarters developments, and large convention hotels all are within this consulting industry’s purview. A typical description of the process is provided by one of the larger players in the field, KMPG (a firm not involved in the Foxconn studies): “Economic impact analyses measure the direct and indirect effects of a business, organization or event on the local, regional, state or national economy. The presence and contribution of a particular project or business generates total economic impact that is substantially larger than the project’s or business’s activities on its own.” The telling point being the assumption that the enterprise being studied not only will have a positive impact, but one “substantially larger” than its immediate business. This assumption is belied by experience. Imagine a long-term impact study focused on a nuclear power plant, taking into account its substantial construction and operational workforce plus the spinoff economic benefits of a new source of cheap, clean electrical power. Now suppose that plant is called Chernobyl.

The first economic impact report on the Wisconsin project was prepared by international accounting firm EY (formerly Ernst & Young) before Foxconn had been publicly identified as the company behind the initiative. It was commissioned by Foxconn and released in July 2017 under the code name “Flying Eagle.” The EY report is the foundational document justifying the massive payouts to Foxconn. It was in turn used by all the subsequent studies. Baker Tilly, an accounting firm headquartered in Chicago, used the EY report for an August 2017 follow-up study commissioned by the semi-private Wisconsin Economic Development Commission, Wisconsin's deal-making agency. A UW-Madison professor, Noah Williams, was commissioned to rush out a paper, also in August, called “An Evaluation of the Economic Impact of the Foxconn Proposal.” This paper also relied heavily on the EY report.

The Foxconn project was unveiled at a White House event on July 26, 2017. Joining President Trump on the podium were Paul Ryan (in whose district the factory would land), Wisconsin native and then–Chief of Staff Reince Priebus, and Governor Walker. Foxconn was represented by its Chairman, Terry Gou, one of the richest men in the world.

In the wake of this announcement came dual revelations. From the Foxconn corner came the warning that they were only interested in sealing the deal if it could be expedited. On the Wisconsin side, the magnitude of the deal so exceeded the budget and charter of the WEDC and the state's existing economic development authority that special legislation would be required. In the rush that followed, the economic impact studies were the key drivers of the Walker team’s campaign for public support and legislative momentum. The studies did not go without some initial critique. One of the first objections to the EY report was its failure to consider commuters from nearby Illinois who would be paying taxes to Illinois, undercutting the tax-based repayment calculation. This inspired the second commissioned study from accounting firm Baker Tilly. In the haste, signals must have gotten crossed because Baker Tilly’s initial report for the Foxconn-boosting WEDC put 40 percent to 50 percent of the construction workers and full-time Foxconn staff as out-of-staters. This conclusion blew up the entire economic case for Wisconsin’s investment. After what must have been a heated clarification meeting with the WEDC, Baker Tilly came back with revised figures of 100 percent Wisconsin residents.

Williams’s report was also clearly inspired by the out-of-state worker concerns, as he spent considerable energy juggling numbers to show how Illinois workers wouldn’t undermine Wisconsin’s payback. His argument was that more Wisconsinites now commute to Illinois than vice versa, so Foxconn would simply stem the flow. But neither of these reflexive and unconvincing studies got specific about how the drive time from the Foxconn site to the Illinois border is nine minutes, while it takes 25 minutes to drive from there to downtown Racine, the local area with the heaviest poverty and unemployment. Add to the reality of commuting convenience, demographics. The Interstate running past the Foxconn site leads to Lake County, Illinois, which has a population of 703,000, compared with Racine County’s 195,000. Residents of western Lake County currently put up with commute times that are the highest in the greater Chicago area. Next door, in McHenry County, Illinois, also bordering Wisconsin, half the population commutes a half-hour or more, mostly into Chicago metro traffic. As for the higher-paying Foxconn jobs requiring post-secondary education, Racine County and adjacent Kenosha County have college-graduate populations of 24 percent, compared with Lake County’s 44 percent.

In the wake of all these studies, the nonpartisan Wisconsin Legislative Fiscal Bureau prepared an important report for legislators who would be voting on the Foxconn deal. It was released on August 8, 2017, and used EY numbers to calculate the projected income and payback schedule. It was this report that produced a 25-year payback number that quickly gained currency in the legislative debate and in a wide range of media reports. Unfortunately, the overlay of urgency didn’t seem to allow for examination of all the underlying premises—for instance, the study’s choice to ignore the likelihood of out-of-state construction workers or full-time workers from nearby Illinois, to accept as granted the EY projection of $53,875 average salaries, and the Legislative Bureau’s decision to calculate the break-even point without use of discounted cash flow, the time value of money.

Finally, the December 2017 TIF deal, in which Racine County and the Village of Mt. Pleasant (where the plant will be located) committed to $764 million in bond-raised revenue for infrastructure in support of the Foxconn project, also used EY numbers to justify the expense and project the payback, this based on increased property values of the Foxconn development site, which in turn rely on Foxconn’s estimates of capital expenditure. As Mark Fralick, a Texas-based industry expert on automated production and a contributing editor to the most prominent U.S. trade magazine in Foxconn’s industry, Supply Chain Digest, put it: “Foxconn has a long history of overpromising and underdelivering. I just hope the government entities and municipalities involved don’t get screwed.” It’s not boding well. The TIF development agreement between Foxconn and Mt. Pleasant specifically calls for a “Generation 10.5 TFT-LCD Fabrication Facility,” with the implication for massive supply-chain requirements including the $1 billion-plus Corning glass plant predicted by boosters. Even before breaking ground, Foxconn has voided this promise in lieu of a more modest facility, while Corning has scoffed at the prospect for a huge plant in Wisconsin.

In the end, a pyramid of assumptions, report after report, and billions of dollars of taxpayer commitments are all built on the case presented by the initial EY “Flying Eagle” report. If this foundational document is flawed, then everything about the project is in doubt. So you would assume that the EY study, especially considering the possible prejudice associated with its Foxconn commission and Foxconn-sourced assumptions, was the subject of intense scrutiny. Perhaps you’d imagine that the subsequent studies would pick apart the original’s methods and conclusions.

You’d be wrong on both counts.

Projecting, Not Predicting

Economic impact studies rely on software that allows for the inputting of certain variables that in turn lead to a projected outcome. In an economic impact study like EY’s for Foxconn, the direct impact is the capital cost of construction and subsequent Foxconn hiring. Indirect impact comes from the additional workers employed by the construction contractors plus the supply-chain effect. Induced impact comes from the money all these new employees spend on food and rent and consumables. These indirect and induced effects are expressed as a multiplier. In this case, EY used established benchmarks to predict that for every ten jobs Foxconn creates, another 17 jobs will be created in Wisconsin.

Multipliers are the secret sauce in the consultant’s recipe. The number used, while based on seemingly hard evidence such as regional qualities and industry-specific metrics, is in actuality subject to considerable judgmental oversight and wiggle room. As Iowa State economist David Swenson, who has completed dozens of commissions using input/output models, explained: “You can’t get good output without precise knowledge of the inputs and I’m confident that Foxconn was not willing to reveal all the details of their business plan. You also have to have specifics about the supply chain—do these businesses exist in Wisconsin? I doubt they do. Without this information, you’re forced to rely on the default model, which may not match the realities of the project.”

Small adjustments in the multiplier can be leveraged into large shifts in economic benefits. The result is the predictable manner in which virtually every commercial economic impact study fulfills or exceeds the expectations of the project’s sponsor (typically, the entity writing the check to the consultant). And even if they seem somewhat magical, multipliers are real. The only caveat is that multipliers for an enterprise can end up being less than one or even negative (think of the former GM plants in Janesville, Wisconsin, or Flint, Michigan)—a result never occurring in these commercial studies. Peter Fisher, University of Iowa economist and research director at the Iowa Policy Project, has spent a career studying economic development efforts, concluding that much of the money spent makes little difference in location decisions. He is particularly leery of the long-term payback calculations in the Foxconn deal. “The average life expectancy of a new enterprise is under ten years.”

The Foxconn studies modeled the electronics industry to pick a multiplier. Yet most of the comparable companies don’t hire large numbers of low-paid workers to assemble things (more often, they offshore these operations), and neither do their enterprises include worker dormitories, on-site housing, a large cafeteria, or other amenities, as do Foxconn’s plans for what it is calling its “Smart City.” The giant Foxconn complex will be much more insular than a normal factory, both due to location and design. Many of its employees (think H1-B Chinese engineers and dormitory-dwelling immigrant labor) will not be seeking commercial housing in the neighboring communities, nor spending as freely there as the technicians and engineers at the kind of electronics companies that were used as models. These regional companies also are typically owned and headquartered locally, keeping profits in the community, rather than being exported out of the region to distant executives and shareholders.

University of Illinois at Chicago Associate Professor Joshua Drucker has studied the economic impact of the building and closing of major facilities, particularly military instillations. One of the lessons he’s learned is that the local and regional economy can change across time, often significantly and unpredictably. Not only do the commercial studies fail to try to encompass the broad range of possible outcomes, they apply a linear model, so that the impact of 1,000 jobs is exactly a thousand times greater than one job. “These assumptions simply don’t hold in real life,” Drucker explains. “For instance, in a new facility, one thing we see is a high percentage of jobs going to people from outside the immediate area.” The studies also entirely ignore the non-economic impact that a large project will have on a community in terms of quality-of-life issues, such as traffic and potential air and water pollution. What will be Foxconn’s short and long-term effect on what is now a sleepy rural area in terms of the housing market, social service agencies, and new stresses on infrastructure and local schools? Don’t look to commercial economic impact studies for the answers.

Despite all the peer-reviewed critiques from economists, politicians continue to enthusiastically commission reports and projects continue to be approved. While partisan lines seemed well-etched in most large corporate subsidy deals, with Republicans on the pro-business side, some of the most vocal critics of the Foxconn-type mega-deal comes from the far right. Roy Cordato is a lecturer at North Carolina State and senior economist at the John Locke Foundation, a conservative, free-market-oriented nonprofit.

“The missing ingredient in economic impact studies is economics,” Cordato states. He lambasts the shallowness of the software used by consultants and their typical dearth of the academic economics background needed to understand and explicate the inherent flaws. “All of these studies ignore basic principles of economics and, as a result, do not meaningfully measure what they claim to be measuring—economic impact.”

The philosophic underpinnings of Cordota’s bile is one that must have also struck a  nerve in many Wisconsin ideological conservatives. How can you support smaller government, free markets, and an America-first agenda while also boosting a multibillion-dollar government handout to a foreign corporation?

Cordota: “A lot of Republicans talk the talk about free enterprise but don’t walk the walk. In our battle against corporate welfare, we find ourselves aligned with advocates from the left with whom we might be normally in conflict.”

Disturbing Assumptions

All of the reports prepared to analyze the Foxconn deal have one common flaw: They were commissioned by interested parties. Even the report issued by UW Professor Noah Williams turns out to be tainted. His chair is funded by conservative sponsors (including the Koch brothers, big Walker supporters) and he personally sought an economics consulting position on the Walker presidential election team. This might help explain why he used a generous 2.7-times multiplier, producing a rosy prediction of economic bounty.

Here are four of the key flaws behind the economic impact reports and resulting contract with Foxconn:

Opportunity Costs. The most glaring and unsettling flaw in standard economic impact reports is their total lack of attention to alternative uses of the massive monies involved. This lack of vision is especially glaring in a state in which the governor has made major cuts to K-12 and university education and with serious infrastructure erosion, as witnessed by its roads being currently ranked among the worst in the nation.

As Cordato puts it, “Any economic impact study that does not attempt to assess these opportunity costs cannot legitimately be called economic analysis.” With job trends favoring the educated and trained, education spending is one of the key alternatives that should have been examined. In Wisconsin, for instance, the most robust job market in the state is in the Madison area, where entrepreneurs with University of Wisconsin ties have been responsible for numerous fast-growing startups with highly compensated workforces in medical software, biotechnology, and game development, among others. Cuts to education spending under Walker have already resulted in the loss of many highly considered UW professors and the degradation of the K-12 teaching profession, while in-state students at UW are asked to pay more than twice what they did in the 1980s in inflation-adjusted dollars. Cost-cutting at the university level continues. In March 2018, the University of Wisconsin, Stevens Point, which serves 8,600 students, proposed axing all of its humanities degrees in response to a $4.5 million deficit, an amount equal to 0.01 percent of the state’s financial commitment to Foxconn.

The beggaring of public education to raise money for corporate incentives has become something of a national scandal. In 2017, the Republican-led state government in neighboring Iowa decided the best way to balance its budget was to emulate Wisconsin’s deep cuts to public education. During the eight-year governorship of Republican Tea Partier Bobby Jindal, Louisiana’s state spending for its public universities dropped from 60 percent of their budgets to close to 25 percent. Seven states were spending 15 percent or less on K-12 education in 2015 than in 2008 in inflation-adjusted dollars: Arizona, Florida, Alabama, Idaho, Georgia, Texas, and Oklahoma. All seven have Republican governors and Republican majorities in both houses. Donald Hall, the CEO of Hallmark in Kansas City, has been a witness to many business incentive expenditures as Missouri and Kansas bid over corporate relocations. He told The New York Times that “If you’re looking at the competitiveness of a region, the most important thing a region can do is to focus on education. And the use of incentives is really transferring money from education to businesses.”

The study on tech clusters by Professors Bresnahan, Gambardella, and Saxenian took a different route to the same insight. They conclude the key factor in sparking the sort of synergistic development of successful enterprises that characterize employer hotspots like Silicon Valley or North Carolina’s Research Triangle is entrepreneurial talent and skilled labor. The best public policy? “Invest in education.”

Labor Dynamics. The primary benefit of the Foxconn deal touted by Walker and his boosters is the creation of good jobs. While the promise of thousands of $54,000-a-year jobs harkens back to all those lost, well-paid, union manufacturing jobs in the auto industry, the reality is likely to be otherwise. The rub is in the calculation of average wages. Foxconn lobbied hard to allow up to $400,000 of higher-paid jobs to factor in the average, allowing them to hit their average income target with more than 90 percent of jobs paying in the $30,000 range. Foxconn has already opened up a TV manufacturing facility in Mt. Pleasant, where it has been hiring workers through employment agencies. The “family-supporting” wage Foxconn is currently paying these assemblers? Fourteen dollars an hour, a figure that is not only not “family supporting,” but will potentially qualify families for food stamps, housing subsidies, and Badgercare, the state’s health-care program for low-income residents.

One of the inherent weaknesses in standard economic impact studies is the failure to calculate the difference between new job creation and job poaching. These reports assume all workers are new workers. This might be sensible if there were a large pool of qualified, unemployed workers. This is not the state of affairs in Wisconsin. In fact, quite the opposite. Not only is the state operating at near full employment, the demographics are getting worse, rather than better. In September 2017, the Wisconsin State Journal bannered a headline story on employment calling it a “looming workforce crisis.”

What happens to companies who lose skilled workers to Foxconn? They might fill these jobs. Then again, the new subsidized competition might price them out of the market, forcing them to downsize, relocate offices or jobs, or go out of business. For instance, the 2008 opening of a large Honda factor in Greensburg, Indiana, didn’t lead to the economic impact study’s projected boom of thousands of regional auto-parts supply-chain jobs. Instead, a 2015 study by California State University, East Bay’s Brian Adams (“The Employment Impact of Motor Vehicle Assembly Plant Openings,” Regional Science and Urban Economics) showed that after five years, the area had experienced a net loss of 500 supply-chain jobs as the limited labor force was cannibalized by the subsidized new factory. These negative possibilities are not a consideration in commercial economic impact studies.

The Wisconsin Fiscal Bureau projections show a major uptick in state tax revenue in 2019 and 2020, when they predict the 10,000 construction workers anticipated in the EY study will be deep into excavation, concrete work, and building fabrication. The assumption is that these 10,000 workers would otherwise be unemployed Wisconsinites. In 2012 a major study on a similar issue was commissioned by Metrolinx, the Toronto planning authority, to examine construction capacities prior to some major infrastructure projects. Unlike the studies on Foxconn, Metrolinx surveyed area construction firms and discovered major constraints among contractors likely to be involved, with skilled labor being the most important. The result was not unconstrained expansion of the labor market, but constraints on the number of simultaneous regional projects. One solution might be the migration of labor, but permanent relocation (with the anticipated Wisconsin tax revenue) is unlikely to be a major factor for specialized jobs that might only last a year. A more accurate scenario is Foxconn construction work forcing the delay of other construction projects.

The Payback. The nonpartisan Legislative Fiscal Bureau crunched the Foxconn and EY numbers to come up with a 25-year payback. Regardless of the assumptions involved (and there are deep flaws), the very exercise is problematic. Greg LeRoy, executive director of Good Jobs First in Washington, D.C., whose mission is promoting accountability in economic development, says “It’s preposterous to think anyone can predict what is going to be happening with the economy in 25 years, particularly in high-tech where product lifecycles are so volatile.” One example of the flawed assumptions: In its calculation of the 25-year payback for the Foxconn project, the bureau did not use the accepted standard of discounted cash flow. Instead, it calculated the break-even point as when Wisconsin gets its original money back—even though inflation alone across the 25 years mocks the entire concept. The Office of Management and Budget recommends that public projects use a 7 percent annualized rate of cash discount. Using such a number and retaining even the credibility-stretching bureau’s job assumption, the payback period quickly diminishes from the 25-year period the bureau produced to … never.

The False Security of Clawbacks. Over the years, as states and municipalities have been burned in corporate incentive deals, a sort of insurance has been added to these agreements. They give the sponsoring body authority to reclaim their upfront money if the corporate entity fails to fulfill its promises—typically measured by capital expenditure and job creation. Clawbacks are an important part of the contract between the semi-private Wisconsin Economic Development Corporation (formed soon after Walker's initial election in 2010, with Walker as chairman) and Foxconn.

As an April 25, 2016, story by Gabrielle Gurley in the Prospect explained, “Clawbacks have met with checkered success across the county.” Pro-business officials who are personally invested in a project find it hard to activate clawback provisions even when they are triggered due to fear it will sour their pro-business reputations and confirm their over-optimism. Another factor can be the leverage that a corporation has once it has accepted and used a large amount of public funds. The optimistic forecasts by Wisconsin’s fiscal bureau predict the state will be in the hole to Foxconn by at least $1 billion by the end of 2025. What would state officials say if Foxconn demands at this point a rewriting of the clawback schedule due to unforeseen industry or economic changes? Would they be willing to let Foxconn just walk away from the project? Could they successfully sue for clawback damages from a Taiwanese corporation or from the other guaranteeing entity, Chairman Gou’s personal corporation, based in the Cayman Islands?

A deep look at the economic impact study process is bound to end up in a place of a certain amount of cynicism. Academics of all stripes have been taking shots at the industry for years without making much of a dent. It’s a business model that is self-rewarding and self-perpetuating as long as politicians look for quick fixes, photo opportunities, and job-creation campaign fodder.

Michael LaFaive is another example of the strange bedfellows who are united in their distaste for taxpayer-funded megadeals driven by economic impact studies. He’s been keeping an eye on the escalating economic incentive game for years from his position as senior director of the Morey Fiscal Policy Initiative at the nonprofit Mackinac Institute in Midland, Michigan, a conservative think tank backed by the Koch and Bradley foundations, among others. He’s been lobbying for years for Michigan to get out of the economic-incentive bidding wars, without much success. In fact, Michigan was ready to pony up billions of its own to get Foxconn. “We won by losing,” LaFaive summarizes. “It’s not that people are blind to the liabilities of this endless bidding to get industry. I talk to state officials all the time and as leery as they might be, they believe they can’t unilaterally disarm.”

Some hopeful signs have emerged. Kansas and Missouri have signed what amounts to a détente agreement over not spending wildly to get corporate headquarters to move a few miles over their state lines. Florida has adopted an Australian-developed software model that allows its leaders to compare opportunity costs among several expenditures, so that decision-makers can look at the relative cost/benefit among disparate spending options such as a corporate incentive, beach improvements, or tax relief.

LaFaive sees two potential end games as corporations become bolder with their demands (think Amazon’s new headquarters) and states and cities show greater creativity and willingness to raise the requisite funds. “Either the whole thing collapses on itself, possibly via scandal, or people come to their senses. All it will take is a few governors calling their colleagues together to declare a truce.”

Timothy Bartik, the W.E. Upjohn Institute economist, has kept a jaundiced eye on economic development spending for many years. He is particularly critical of the conceits behind many of the impact studies and subsequent state decisions that assume incentives are the primary factor in location decisions. “I’m also flabbergasted at how the studies behind the Foxconn project looked at it from a purely statewide fiscal perspective, and ignored the substantial costs to the local communities in terms of added educational demands, police and firefighting, roads and social services.”

The Foxconn project has turned into a major issue in this November's gubernatorial election. Walker, who has been cited in numerous national roundups as a vulnerable incumbent, initially seemed surprised by the backlash against the project. In the months after the announcement, he was jubilant over landing the deal, no doubt still deep in the thrill of winning the multistate bidding war. He was pleased to have ammunition to fight his critics who pointed to how he had fallen far short of his first-term job development projections. He was also sensitive to his vulnerability over the state’s economic competitiveness exposed by his presidential primary opponents in 2015, particularly Donald Trump. But when late 2017 polling showed the unpopularity of the multibillion-dollar commitment, he temporarily eliminated the topic from his stump speeches.  But there were never any signs of doubt about the deal from the Walker camp. For a career politician better characterized by his ambition than his vision, with the prospect of fighting a blue-wave election cycle, Walker felt the need for the kind of dramatic and short-term boost provided by the prospect of thousands of imminent Foxconn jobs.

Since then, Walker, with Foxconn’s assistance, has been assiduously working on promoting the project’s statewide benefits, a tough sell considering the project’s geographic location in the far southeast corner of the state. In July, Foxconn announced plans to develop R&D “innovation centers” in far-off Eau Claire and Green Bay, which have been noted as odd choices, given the educational and commercial landscape, which puts liberal-leaning Madison and its University of Wisconsin at the center of engineering and technological talent.

Critiquing the Foxconn deal was a common thread among the Democrats vying to run against Walker this November. In their first televised debate, all eight candidates pitched in. Madison Mayor Paul Soglin took this shot at Walker: “I’d like to know what he was smoking when he negotiated this deal. In China they’re laughing at us.” Another candidate promised to sue to break the contract. Eventual September primary-winner Tony Evers, the state’s superintendent of public schools, has also been anti-Foxconn, although less stridently, stating, “Obviously, Walker sold the farm on that one.” Among his first ads was one criticizing Foxconn, in which he concludes, “Just think if we invested that money in our schools instead.”

Would the Foxconn deal have been approved it had been put to a statewide referendum, rather than pushed through a compliant, Republican-controlled statehouse? It seems highly doubtful, since polls show significant skepticism over the benefits of the deal even among the state’s Republicans. Business leaders such as Kevin Conroy, CEO of fast-growing Madison biotech Exact Sciences, maker of Cologuard, has severely critiqued the deal, lamenting the lost opportunity cost of using even a fraction of the investment as incubation and venture capital, with the potential to create the kind of high-paid workforce seen in his company and other emerging companies in the state.

As Wisconsin superintendent since 2009, Evers has seen the dramatic effects that Walker’s education spending cuts and attacks on public unions have had on the quality of K-12 and university education. He’s on the side of evidence-backed, peer-reviewed research that demonstrates that taking money from education and using it for corporate subsidies can be a disastrous path for any state. Historically, the Wisconsin electorate has proven to be finely balanced, most recently electing both Tammy Baldwin and Ron Johnson, among the most liberal and most conservative senators in Washington. The soft-spoken Evers may believe he can convince Wisconsin voters that Walker has misspent billions of taxpayer dollars to bring Foxconn to the state. But as of yet, he has not articulated exactly how he would go about modifying the signed contracts and the promised subsidies, nor is it clear that an onslaught of Walker ads touting the thousands of wonderful Foxconn jobs and concomitant statewide economic bounty will fail to sway. After all, which is easier to parse: 13,000 “family-supporting jobs” or “just thinking” about what a few billion dollars could do if spent on education.

Long after deal-maker Scott Walker and the current legislators who voted for Foxconn incentives are out of office, Wisconsin and Racine County taxpayers, as well as their children, will still be paying the bills.

Don't let big tech control what news you see. Get more stories like this in your inbox, every day.