First-ever binding end to a border war: Missouri-Kansas UPDATED

Kansas Governor Laura Kelly has just signed an executive order that prohibits state subsidies being used to move existing Missouri firms in four Missouri counties to three Kansas counties, which together make up the Kansas City metropolitan area. Unlike previous voluntary no-raiding deals, such as NY-NJ-CT, Council of Great Lakes Governors, and even Australia's Interstate Investment Cooperation Agreement, this is not a voluntary agreement, but one with the force of law, a first in the country's history.

The binding nature of the agreement comes from the fact that both states have legally bound themselves. In June (h/t New York Times), the Missouri legislature passed and the Governor signed into law restrictions on using state subsidy funds (such as Missouri Works) from being used to attract companies operating in three counties in Kansas in the Kansas City metro area. The law stipulated, however, that this would not go into effect unless Kansas passed similar binding regulations on its state subsidies (Promoting Employment Across Kansas, or PEAK) being used to attract firms on the Missouri side of the metro area. Today's action by Kansas' Kelly meets the stipulations of the Missouri law.

Kansas and Missouri have thus joined the European Union as the only areas with legally binding no-raiding rules, anywhere in the world. Of course, the EU's provisions, based in the Guidelines for Regional Aid, cover the entire territory, whereas the Kansas-Missouri ban only applies in the Kansas City metropolitan area. However, since that's where almost all the job piracy between the two states takes place, the two states' action is truly historic.

The two states made a previous attempt at a binding agreement in 2014-16: Missouri passed a similar law to this year's in 2014, but Kansas' last-minute law just before a 2016 deadline in the Missouri law precipitated a stalemate, and no new deal was made until this year.

It should be remembered that an agreement of this type has been advocated for by a number of companies in the region, notably Hallmark and the Hall Family Foundation. Their work uncovered the hundreds of millions of dollars that had gone into the border war and led to pressure being put on both state governments to end the madness.

New article on tax increment financing in Missouri shows impact of KS/MO border war

After several years of work, my colleague Susan G. Mason (Boise State University) and I have published a new article on TIF in Missouri, specifically in the St. Louis and Kansas City metropolitan areas. "Exploring Patterns of Tax Increment Financing Use and Structural Explanations in Missouri's Major Metropolitan Regions" appeared in the July 2018 edition of the HUD journal Cityscape, downloadable for free here. We omitted the two cities from our earlier statistical analysis (in the paywalled Economic Development Quarterly, May 2010) because they are much larger than any other Missouri city and use TIF far more than any of them, making them statistical outliers.

In our earlier article, “Tax Increment Financing in Missouri: An Analysis of Determinants, Competitive Dynamics, Equity and Path Dependence,” we found that early adopters of TIF tended to be heavier users of TIF far past the first TIF adopted, that TIF as used in Missouri exacerbated inter-jurisdictional inequity (cities with higher poverty rates were less likely to use it than cities with lower poverty rates), and we found strong evidence for competitive dynamics in the use of TIF: Cities that were adjacent to a TIF-using city were two and a half times as likely as average to use TIF themselves, and implement more TIF projects.

The new article is an exploratory study, as it is impossible to generalize from two cases. But one thing we established clearly, based on complete data from 1988 to 2013 for St. Louis, and from 1988 to 2012 for Kansas City, is that Kansas City's tax increment financing projects are marked by much higher aid intensity (the EU term that equals subsidy/investment) than those of St. Louis. Indeed, even excluding 2009 in St. Louis, which was marked by several multi-billion projects with low aid intensity (and at least in the case of Northside Regeneration, had substantial state funding not reflected in Exhibit 4 of the article), the overall average aid intensity, ex-2009, is 17%.

By contrast, in Kansas City, the average aid intensity of the city's TIF projects comes to a whopping 36%,* more than twice as much. Everyone we interviewed on the question considered that there is much greater competition for investment with Kansas than with Illinois in the two metro regions. The difference in aid intensities is consistent with this thesis.

These points take on renewed importance as Missouri's Governor Mike Parson is convening the Governor's Conference on Economic Development in Kansas City this week. As you may recall, Missouri and Kansas came close to a truce in the border war in 2016, but it collapsed when Governor Sam Brownback proposed legislation that did not match what Missouri passed in 2014. Since Missouri has a new governor now and Brownback will be replaced in 2019, there is renewed optimism that a binding halt can be brought to the use of state subsidies in the border war.

The Governor's Conference will include a panel Friday on the border war in which I'll take part. We will discuss the background, data on the cost of the border war, and a potential solution. I will also point out that TIF is impacted by the border war, too. It will be great to try to persuade lawmakers and the governor.

Unfortunately, the cost to attend is prohibitively expensive, but I will report back next week.


* The mandatory typo occurs in Exhibit 3, Kansas City, in the line summarizing 1988-2006. First, the correct amount for total TIF reimbursement is $3,527,420,262, excising the stray "1" in the millions area. The correct aid intensity for that period is 36%, not the 21% listed and probably copied from the previous line.

Show Me the (Subsidy) Spreadsheets!

Good Jobs First's new report, Show Us the Subsidized Jobs, is its third assessment of state subsidy transparency, following up on reports published in 2007 and 2010. The good news is that transparency continues to spread: From 23 states in 2007 to 37 in 2010 to 47 plus the District of Columbia in 2014.* The bad news is that for most states, online transparency still has a long way to go.

Why is transparency important? As the reports says, without it, it

makes it impossible for the public to get at
even the most basic return on investment,
accountability or equity questions. Which
companies received subsidies (and what
kinds of companies)? Are they delivering
on job creation? How good are the new
jobs? Where will the jobs be located?
Reasonable people cannot have an
informed debate and policymakers cannot
watch the store without good job-subsidy
data.

Moreover, the cost of state (and local, much more poorly reported on) subsidies comes to some $70 billion a  year, according to my estimates. Moreover, there is a tremendous opportunity cost associated with this: While of course some jobs cost more and some cost less, at $50,000 per year in salary and benefits, that is enough money to hire 1.4 million state and local workers, more than have been laid off since the start of the recession.

Good Jobs First toughened its scoring system this year, making the raw numbers not directly comparable with previous reports. In particular, there is more emphasis on publishing outcome data such as jobs actually created, wage levels, etc. It is also important to note that the report only counts online transparency: If a state agency will give you some information on request, that's nice but won't improve a state's score. We live in a wired world, obviously, so transparency performance should conform to that.

On its 0-100 scale, the best performing states are:

1) Illinois 65
2) Michigan 58
3) North Carolina 48
4) Wisconsin 46
5) Vermont 43

Even these have a long way to go, especially on job outcomes. But they are doing far better than the 14 states with scores below 10 (including Arkansas' corrected score of 8). For example, Georgia, at 4 points, gets demerits because two of its major programs take down the reports a mere 30 days after they are posted.

Taking down data is a pet peeve of mine. Missouri, which ties for 14th for its reporting of state programs, also has quite good data online for local tax increment financing (TIF) subsidies, because the cities are required to file annual reports with the Missouri Department of Economic Development on a project-by-project basis. This requirement was given new teeth a few years ago, so the reports are now close to complete for recent TIF projects. Moreover, they have always been required to give both projected and actual job creation data (though not wage data). The reports provide the amount of the subsidy and the amount of the investment, allowing for the calculation of what the European Union calls "aid intensity," which is the ratio of the subsidy to the investment. This is an important measure, because one obvious reform that could be made for any subsidy program is to cap its aid intensity. Why should Electrolux get a 99% free new factory in Memphis?

However, Missouri TIF reporting has long been hampered by projects that are listed in one year's report disappearing in later reports. Moreover, all the old reports have been taken down by the state, making historical research (trends in aid intensity, anyone?) next to impossible. What I personally have by way of historical TIF data I obtained by hand transcribing information from the pdf format reports onto a spreadsheet. But obviously the state keeps these files somewhere, and Department of Economic Development officials in all probability have a master spreadsheet encompassing all project reports received. The ideal for research and democratic accountability purposes is simple: Show me the spreadsheets!



* The report says 46 states, but after its publication Arkansas belatedly identified two programs with online disclosure. This leaves only Delaware, Idaho, and Kansas as the only states with no online disclosure of their subsidy programs.

Cross-posted at Angry Bear.

The Folly of Subsidizing Retail

Next to giving subsidies for a company to relocate, or to prevent it from locating, the lease defensible common use of investment incentives is for retail. Why should this be? Let me count the ways.

Most importantly, retail is a derivative economic activity, as David Cay Johnston says. A location's population and income determine how much retail it can support. For this reason, the apparent job creation of retail subsidies is completely phantom, as sales and jobs are simply transferred from older stores to newer locations. The best proof of this is contained in a groundbreaking study by the East-West Gateway Council of Governments, the regional planning agency of the St. Louis metropolitan area.

East-West Gateway's study found that from 1990 to 2007 (i.e., before the financial crisis), the over 100 local governments of the St. Louis metro area had collectively provided over $2 billion in subsidies for malls and other retail facilities. Most of this was in the form of tax increment financing, a popular local subsidy tool in both Missouri (to the tune of $339 million annual average from 2004 to 2006; see p. 7 in the source) and Illinois. Yet, by the end of this 17 year period, there were only 5400 more retail jobs in the metro area than at the beginning. This would total $370,370 per job if the jobs were created by the subsidies; however, it is more likely that they are simply due to income growth in the region. (Note to reporters: This would be a great study to replicate in your area.)

Second, retail jobs are not all that good. Here is a custom graph from FRED showing the nominal wage trend for all production and non-supervisory workers (blue) and for production and nonsupervisory workers in retail (red). As you can see, the wage gap has been increasing for 40 years. As of September, the exact figures were $13.86/hr. for retail vs. $19.81/hr. for all private industries. Moreover, given that the overall total as shown in the graph actually represents a decline in real wages, the decline in retail is much more pronounced.

In addition to the low pay, retail workers rarely get benefits. According to a new report, only 29% of retail workers get health care benefits, even though half of retail employees have college degrees and 70% are over age 24.

Third, retail does not generate much secondary employment, the way manufacturing does. It does require warehouse jobs, but those generally get subsidized, too, as in the case of Wal-Mart.

The bottom line, then, is simple. Retail is a derivative economic activity that generates almost no new spinoff activity, and local governments (except perhaps in poor areas with food deserts) should not subsidize it. As we have seen in St. Louis, local governments have proven perfectly capable of wasting billions of dollars for temporary gains in sales tax revenues. It's time to stop the madness.

Are Incentives Ever Good Policy?

Most of my work on investment incentives is quite critical of their use. As a result, I often get questions about whether there are any good uses of incentives, or whether states have no choice but to offer them because other states or nations are using them. The research of Tim Bartik sheds important light on this topic, in terms of identifying types of incentives that can have positive national benefits. I consider also some circumstances in which using investment subsidies can be politically justified, as well as the question of whether governments are trapped in the incentive game.

According to Bartik, two types of intervention can have positive national benefits, as opposed to benefits for the local jurisdiction.His work shows that typical "smokestack chasing" incentives, while they have a positive local benefit, have negative consequences for other states, that exceeds the benefit to the state that offers the incentive: hence, a negative net national impact. This is consistent with my work at the theoretical level that governments face a collective action problem in their use of subsidies. It also is consistent with empirical finding in international work that when other countries in your region use incentives, it decreases the amount of investment your country receives.

As I said , Bartik identifies two types of cost-effective incentives: customized training for companies on the demand side, and early childhood education programs and generalized job training programs on the supply side. Neither of these is smokestack chasing, and they make up a small portion of the incentives most states offer. For example, in North Carolina, in fiscal year 2008-9, the state's Economic Development Inventory reports that total economic development spending came to $1.2 billion. Of this, less than $11 million (1%) was spent on all types of training programs. North Carolina has a well-known early childhood education program called Smart Start, which is not considered part of the economic development budget. After a "one-time reduction" (quotes because there was another reduction the following year), the program had $194 million in FY 2008-9, about 16% of what was spent on economic development subsidies.

There is one time when I consider smokestack chasing to be politically justifiable, when it is restricted to locations that meet objective criteria of economic deprivation. This is what the European Union does, and I show in Investment Incentives and the Global Competition for Capital that EU policy has reduced what individual governments have spent on incentives, and only allowed them in the poorer regions of the EU. Some U.S. states have crafted their incentive laws to favor poorer counties, but that targeting usually is weakened over time as big companies say they'll only come if they get the maximum subsidy in a richer county. With no centralized (federal) law to enforce targeting, states have repeatedly weakened it when they do have it. This is one reason I favor federal regulation of incentives.

The bigger reason is that the states don't take the impact of their incentives on other states into account when deciding their policies. Kansas' government is not elected to care if the jobs it "creates" are merely lured across the state line from Kansas City, Missouri, let alone whether Kansas' job creation indirectly reduces jobs in other states. Therefore, the states have to be restrained by federal action to get rid of the scores of billions spent on subsidies annually.

But what should states do until then? Can states unilaterally disarm? I think the answer is probably not, although the province of Alberta in Canada has done so (oil revenues are probably the reason it can do this). But you won't get credit from me for "responsible" policy if you aren't actively working for a federal solution to the problem. And as we saw as recently as 2006 when the U.S. Supreme Court ruled on Cuno v. Daimler-Chrysler, states were actively lobbying Congress to overturn the ruling should it have followed the Appellate Court's ruling in favor of plaintiffs. In other words, states don't want to give up their economic development tools, because they still fail to appreciate how they are bound in a collective action problem that is ultimately only solvable at the federal level.

In the meantime, critics need to highlight the enormous cost in cash and lost government jobs. Local successes do occur, such as the defeat this year of the $360 million Aerotropolis subsidy in St. Louis opposed by the liberal Missouri Budget Project and the libertarian Show-Me Institute. But there is still much work to do in ending the economic war between the states.