Our latest project on economic development incentives is up for comments.

http://equitablegrowth.org/electoral-institutions-and-electoral-cycles-in-foreign-direct-investment/

This project is a field experiment with Mike Findley and Dan Nielson on how election timing shapes the use of local economic development dollars.  Our pre-analysis plan was registered at EGAP.

The Washington Center for Equitable Growth is circulating this as part of their working paper series. (They did not fund the research).

Comments welcome!

I’ve started yet another project on the costs of offering subsidies to individual firms. These economic development incentives are especially common in the United States where states and cities often compete with each other over firms.

One part of the project is compiling incentives data contained in a database called Incentive Monitor, which is one of the most comprehensive databases of incentives. There are other incentive databases, including Good Jobs First Subsidy Tracker.

Using the Incentive Monitor database we can get a snapshot of the use of incentives in the United States. In the year 2014 alone I can identify over 2,000 individual economic development incentives at an aggregate cost of $10 billion, although this largely ignores many of the local economic development incentives such as tax incremental financing. The incentives in this data base are credited with creating a total of just under 309,000 jobs, or just over $32,000 per job.

These aggregate numbers can be misleading, where the over 2,000 economic development programs are each associated with fiscal costs and job creation numbers. In some cases these incentives are associated with projects that create zero jobs, either these are associated with other goals, such as promoting renewable energy, or in “safeguarding” jobs from being poached by other cities, states, or countries.

By removing the incentives that have no state job creation numbers, the average amount of incentives per project remains at over $31,500. Whether you think this is a big or small number is up to you. What I found interesting is the dramatic differences in the cost of incentives per job by state.

Some states provide either provide no incentives or the number of observations are so small we don’t want to draw inferences. But for the states that offered at least ten incentives in 2014, there is a clear pattern.

First, a number of states provide incentives, but at a substantially lower cost per job than the national average. A large cluster of states offer incentives at cost per job in the teens. What do states like Florida ($10,000 per job), Indiana ($12,000), Maryland ($18.521) and Mississippi ($19,136) have in common?

The Pew Charitable Trusts has been active in helping states reform their economic development programs. They have identified states that have reformed their programs from 2012 to 2014, and five of these states are in our data set. Which states reformed their incentive programs? Florida, Indiana, Maryland, and Mississippi along with Louisiana. We’ll get to Louisiana in soon.

These states that have enacted reforms aren’t the only states with less costly incentive programs, and this description doesn’t make a causal claim that these reforms have reduced the costs of these programs. But there is clearly a cluster of reformer states that have much less costly incentive programs.

Second there are a number of states such as Connecticut that were consistently above the average. Investors in Connecticut were offer 87 incentives at an average cost per job of $59,000. This average is driven up by two massive incentives to United Technologies and Praxair. But of the 87 incentives, 52 were above the national average in terms of costs per job.

Louisiana, one of the states that enacted reforms of their incentive programs, was very active in offering incentives (125 incentives in 2014) and at a cost of $58,083 per job. Iowa performed better but was still above the national average (58 incentives at a cost of $42,000 per job). Illinois provided 31 incentives at a cost of $63,000 per job.

Making relative comparison of these programs is difficult given the different goals of the programs and the additional incentive programs in states that may not be captured in this data. For example, the many tax increment financing districts in Illinois are not included in this data, which could potentially increase or decrease the cost per job averages.

The point is that there are three clusters. Low cost per job, high cost per job and New Jersey.

New Jersey offered 86 incentives at an average cost of over $238,500 per job. That can’t be correct. Let me double check.

It is correct. New Jersey has clearly offered some massive incentives that created very few jobs. For example, the highly criticized $82 million incentive to attract the Philadelphia 76ers practice facility is credited with 250 jobs ($328,000 per job). This helps us get to the $238,500 per job estimate.

But offering incentives at a whopping per job cost isn’t an outlier in New Jersey. It seems to be their economic development strategy.

If we rank all 2,000 plus incentives on a cost per job basis, six of the most expensive (per job) incentives are in New Jersey. This includes the high profile Subaru incentive to attract the company to Camden, New Jersey. Of the top 50 most expensive incentives per job deals, here is the state distribution.

Nineteen of the fifty most expensive incentives were New Jersey incentives.

What is happening in New Jersey?

Interestingly, economic development incentives are under fire in New Jersey, but probably not for the reason that you have expected. The state has under funded their economic development budget and the state owns companies their incentive checks. $785 million in checks.

The main reason is that these incentives, unlike many of the tax credits in other states, require a budget allocation. For example, in most other states an economic development agency can be authorized by the legislature to allocate tax credits, lowering the firm’s future tax bill.

In New Jersey, firms pay their taxes, but are issued a rebate from the state coffers. What is the difference? The difference is that the state needs to have the money in an economic development fund to provide the rebate to firms. But with the budget crisis in New Jersey these programs have been underfunded and thus the rebate checks haven’t been sent out.

I wish I had some clever conclusions on what is happening in New Jersey. Something very, very different is going on in New Jersey. And that isn’t a compliment.

I have a blogged about a some of my recent research projects examining the use of financial incentives to attract firms. My co-authors and I have written that politicians use these incentives to claim credit for investment, but these incentives have very little impact on job creation.

With the support for the Ewing Marion Kauffman Foundation, I have been examining incentive policies in the Kanas City region. As part of this project I surveyed all of the recipients of the Promoting Employment Across Kansas (PEAK) program, the flagship Kansas economic development program.  I have used observational data to compare incentive recipients to non-incentive recipients.

I also conducted a survey of recipients on how incentives affected their employment plans. Using data from a Freedom of Information request I identified 105 PEAK incentive applicants. For all 105 firms, the managers received a recruitment email and a link to an online Qualtrics survey asking a battery of questions about their company’s involvement with the PEAK program. I received a total of 25 responses for a response rate of 23%.

What are the findings?  Incentives are popular with companies and they are ineffective in creating employment.

You’re still reading?

When asked about the program, the vast majority of respondents had a positive experience with the PEAK program and would recommend this program to other firms. A striking 71% of respondents indicated that the program was either “very efficient” or “somewhat efficient” and 92% of respondents would “definitely recommend” or “probably recommend” this program to other firms.

What was the purpose of these incentives? The majority of firms (58%) indicated they were applying for PEAK incentives for expansion, while 42% and 25% indicated this was for relocation and retention.[1]

For the firms that received competing offers from other locations (63% of the firms), 67% of respondents indicated that the PEAK incentives where more generous than the competitor offers. Only 13% of respondents indicated that the PEAK program was less generous.

The program is efficient and generous.  What is not to like?

But are the incentives effective in helping create Kansas jobs? The key set of questions are the counterfactural on what would the firms had done without the PEAK incentive. I asked two direct questions to managers. First U asked if the firm would have left the state of Kansas without a PEAK incentive, providing three possible answers.

1) Without the PEAK incentive would your company have left the state of Kansas?

Next, I asked about the firm’s expected employment without a PEAK incentive:

2) Without the PEAK incentive would you have company hired less employees or the same number of employees?

In response to question 1, only 22% of respondents indicated that they would have left the State of Kansas without the PEAK incentive program while 48% indicated they would have not have left the state. 30% of respondents indicated they were unsure. This question on relocation is a difficult one, since some of the firms were applying for PEAK incentives for expansion.

The second question is a more comparable measure of the impact of PEAK incentives. 67% of firms claimed that they would have hired the same number of workers without the PEAK incentives, while 29% of firms indicated that they would have hired fewer workers.[2]

Thus for most firms, employment decisions were independent of the incentives.

The final question asked about a specific policy. The Missouri State Legislature introduced a bill that proposed limiting incentive competition in the Kanas City area. I asked the respondents if they had heard of the bill (50% had) and if they supported this legislation. The largest percentage of respondents “neither supported nor opposed” the legislation (42%) with a smattering of responses in the supported or opposed followed by 21% of respondents indicating “don’t know”.[3] Thus there is only limited support, and limited opposition, for ending a very specific type of incentive competition in the region.

In summary, firms tend to like this program. Taxpayers probably should not.

[1] Note that these categories aren’t mutually exclusive. Respondents had the option to check more than one.

[2] 4% of respondents were unsure.

[3] The distribution is as follows: Strongly support (0%), Support (8%), Neither support nor oppose (42%), Oppose (17%), Strongly oppose (13%), and Don’t know (21%).

As part of a book project I’m reviewing information on the performance of investment incentive programs in the United States.  These programs are designed to help attract new investment, encourage expansion, and retain existing firms.  I’ve been collecting audits of these programs by state legislatures.

The general consensus from these audits are that the programs lack oversight and that it difficult to verify the performance of these programs.  That is putting it nicely.  Most of these audits range from critical to scathing.  Here are some news stories on the most recent audits.

• A very critical audit of the Wisconsin Economic Development Corporation (the organization and the incentive programs).
• The Texas Enterprise Fund critics include legislators that signed the legislation. Other Texas programs received similar reports including incentives from school districts.
• Claims of lack of transparency and favoritism in the Utah incentive programs.
• I’ve written about the The Promoting Employment Across Kansas (PEAK) program in the past.  The third part of their legislative audit should be out in Dec.  The first two parts were very critical of the program and lamented the lack of systematic data collection on the program
• PEW has a nice overview of the past Minnesota, Louisiana, and Massachusetts audits.

More to come.

Blogging has been light for a variety of reasons.  A few updates.

First, my International Studies Quarterly article on how voters respond to incentives just came out.  See here (gated).

Second, I am presenting work at the Kansas Policy Institute conference on Wichita’s proposed 1% sales tax hike.  What do I know about this?  I know there is a proposal to create a jobs program that includes incentives.  Here is the program.

I think there will be some fireworks at this conference.  More to come on this very, very soon!

I have been working on a bunch of projects on the use of financial incentives to attract companies.  One part of this project examined the “border war” between Kansas and Missouri. Google it.  Companies jumping back and forth between Kansas City, KS and Kansas City, MO purely to maximize incentives.

I presented a short policy paper on how the main Kansas incentive program affects job creation.  The short answer is that I find that the incentive program doesn’t create any jobs (when we compare job creation with a control group of firms).  Details on my paper and the conference can be found here.

I have been invited to Wichita to present my work to local officials as they debate an increase in sales taxes to fund a number of projects, including a job creation package.  The details of this package have just come out.  Here is the full information that I have on the program.

There are some obvious issues that anyone could point out on this program.  A regressive tax (sale tax) is used to fund a jobs program that will be targeted to select companies.  This should at least be cause for some concern.

But the bigger issues on these job programs are: 1) how will this program help create jobs, 2) how will firms be targeted, 3) how will the program be evaluated, 4) what market failure is this program going to address (why doesn’t the private sector already do this).

1)  How will this program create jobs?  By doing everything.  Infrastructure, workplace training, harness the power of clusters while diversifying the economy.  Basically they will create jobs by doing everything.  Which generally means they are doing nothing.  Hard to take this proposal seriously.

2)  How will firms be targeted?  I’m not sure.  This is the only clear language on this:

The City of Wichita would allocate 20% of a 1-cent sales tax to an independent commission. This new commission would be appointed by the City Council and led by private sector business people. Decisions about who receives funding, the number of jobs, and the impact on community would be made in public meetings and tracked through a website.

3)   How will the program be evaluated?  From the proposal:

“Several metrics will be used to track the impact of Jobs Fund expenditures. These would include:

Total Wichita employment increase over five years (target is 7% – 20,000 more jobs).

Average wages increase over five years (target is 5% over cumulative CPI).

Property tax growth over five years (target is 15% or $469 million).

What? The plan for the evaluation of a $80 million program is to eyeball regional economic growth?

4)  What is the motivation for this project?  From the final paragraph:

“Why can’t the free market determine job and economic growth?
The answer is this is a market-driven approach. Businesses have their markets and so do communities. Wichita is competing against 14,000 other entities trying to steal companies and jobs. Oklahoma City is just one of the cities that has a person that spends majority of time in Wichita talking to our companies about moving or locating new work there. If a company needs a rail spur and Wichita can’t provide it, other communities in the market will and Wichita loses those jobs.”

This competition rhetoric is something that drives me crazy.  Paul Krugman’s somewhat dated book, Pop Internationalism, has a nice chapter on this.  I always thought that government officials were being opportunistic in painting their preferred policies as being driven by “competition” and not some other motivation.

And then I started meeting local government officials in Kansas and Missouri.  I think they truly believe that they must make these policy changes and that many of the local economic woes are caused by others.  Not China.  Their neighboring states.

I think they are wrong, but I can see that there is nothing I can say to convince them otherwise. My goal in Wichita is to present my research on how ineffective targeted incentives are for job creation, and talk more broadly about the failures of many local economic development programs.

But if they insist on moving forward on this program, I hope to at least provide some advice on how to limit the corruption, waste, and sometime unintended consequences of these programs.  I think this is where my IPE perspective could help.  Many of these US local development programs look like some of the failed programs in the 1970s and 1980s around the world.

Any advice or suggestions are very much appreciated.