Guest Post: State and Local Subsidies to Business More Out of Control than Ever
guest post by Kenneth Thomas from his Middle Class Political Economist blog
I’ve just completed a new paper (not yet published, so I can't present it all here) showing the effectiveness of the European Union's rules to control investment incentives. Comparing U.S. bidding wars for investment with what happens under the EU's state of the art rules (see below) helps show just how much money is wasted by state and local governments here. As I have posted here before, the annual subsidies given could hire all laid-off state and local government workers. In this post, we examine incentives over $100 million as well as the top 25 incentives since 2000 in both the EU and U.S.
Since the beginning of 2010, there have been at least 20 $100 million incentive packages given in the U.S., compared to just four in the EU. This includes a $1 billion package (present value) given by the state of Michigan to Chrysler in 2010. By contrast, the largest package in the EU in this time was about $285 million. Overall, nine of the top 25 investment subsidies given since 2000 have been given in 2010 and 2011. This is twice as many as you would expect randomly (25*2/11=4.5), which suggests to me that things are more out of control than ever.
An important metric for comparing the size of incentives is what the EU calls “aid intensity,” which is the subsidy divided by the investment. This lets you compare incentives for projects of different sizes. Under the EU's current rules for large investments, which came into effect in 2002, the largest subsidy by aid intensity was 23.19%, a $161 million package that went to Ford Craiova in Romania in 2008. Of the top 25 packages in the U.S. since 2000, only three had a lower aid intensity than Ford Craiova, one was about equal, and the rest were higher, including four over 100%, with one as high as 385%, almost four times the cost of the investment! Thus, the highest aid intensity in the EU was virtually the lowest aid intensity for large projects in the U.S. And EU rules limit the highest subsidies to the poorest regions; the higher the GDP per capita, the lower the maximum allowable incentive, with the richest regions not allowed to give investment incentives at all.
What the EU originally called the Multisectoral Framework on Regional Aid to Large Investment Projects came into effect in 1998, and in 2002 the rules were tightened to sharply reduce the maximum subsidy the European Commission would allow* for investment projects over € 50 million. This can be clearly seen in a list of the top 25 incentives in the EU (you'll have to wait for the paper, or see Table 6.2 in Investment Incentives and the Global Competition for Capital as the top five have not changed since the book was published), where four of the five largest were given before the 2002 reform. Similarly, companies that received incentives under both the original rules and the reformed rules received much lower aid intensity under the new rules. For example, Advanced Micro Devices received a subsidy equal to 22.67% of its investment to locate in Dresden, Germany, in 2004 under the old rules, but only 11.9% in Dresden under the new rules in 2007, and 10.83% when its joint venture, Global Foundries, set up shop in Dresden in 2011. The rule change clearly worked to ratchet down incentives.
The European Union rules show that there is an alternative to giving large incentives to attract investment, that there is no reason to give away free factories to rich companies. But even in rich areas of the U.S., government officials do not want to give up their subsidy powers, so it will take constant political pressure to obtain what is ultimately a federal solution. The only way to make this politically feasible is through constantly reminding people of the high costs, what we have to give up to pay them, and pointing out feasible alternatives.
* Yes, you read that right. In the EU, the 27 independent Member States can only give a subsidy to a business if the European Commission authorizes them to do so.