I have a bunch of work in progress with Eddy Malesky (and one paper with Mariana Medina and Ugur Ozdemir) on the politics of offering financial incentives to attract investment.  Our main argument is politicians use these incentives to claim credit (or minimize blame) for investment (or lack of investment).

This post offers a quick bit of descriptive data, mostly focusing on the question of the job creating elements of these incentives in the United States.  I’m drawing on some of the data from ICA Incentives (subscription required)

Over the last 2 years US states and localities (along with some federal programs) have provided investment incentives for over 4,000 investments. This week there were a total 79 new incentives announced.  A couple of these new incentives caught my eye.

There are a bunch of big incentive packages offered to firms, often by states desperate to create and maintain jobs.  How do we calculate the net benefits and if the incentive is worth it?  One way to measure jobs, but this is complicated.

For example, G&W electric was offered an incentive package to locate in Bolingbrooke, Il.  Only 30 news jobs will be created (at the cost of $250,000 per new job).  But the company claims that this incentive package will help the company retain 400 jobs in the Chicago area along with the 30 new jobs ($17,500 per job).  Similar complications arise if we examine Leanwee Stamping Corporation’s Tecumseh, Michigan investment ($10.40 million incentive for 46 new jobs and 425 jobs retained), Cablevision and Burlington Coat Factory’s decisions to remain in New Jersey have similar issues ($37.50 million incentive for 574 jobs retained and $41 million for 676 retained jobs plus 120 new jobs respectively).

This may sound like I’m cherry picking the extreme cases, but these are pretty representative of the major incentive deals.  The total database (including investments in Europe and a few other places) includes over 5,000 deals with an excess of $40 billion in incentives over the past two years.  The average per (new) job is $58,000.

Another interesting set of cases are the many controversial film tax credit programs that states offer.  This week a French movie production company, Why Not Productions, received a $1.96 million grant for 35 jobs from the State of Michigan.  This brings the total of incentives for this company to over $14 million for the creation of just over 200 full time jobs.  Not only is this a high dollar per job created ($68,000), but I’m also skeptical of the long run impact of impact.  After this production, how many jobs will remain?

I could go on and on, but this is just one week of incentives.  The point is that the cost-benefit analysis of these incentive are problematic for a couple of reasons.

1. How should a government weigh the benefits of creating new jobs and job retention?  For some firms (solar power), the job creation aspect may be secondary to other aspects of the investment.  What info should we use to make a cost-benefit analysis?
2. Firms can threaten to move in order to extract high levels of incentives.  Many of the large incentive packages are motivated for the purpose of job retention.  Would they actually move without the incentives?
3. These incentive programs make headlines for their short-run impact, but it isn’t clear how valuable these investments are in the long run.  Obviously film tax credits fit this pattern.  But many other projects, like wind farms, generate a bunch of construction jobs up front, but they retain very few jobs in the long-run.

Ok, thus far I’ve tried to point out that calculating the net benefits to a state or city seem very difficult.

While many of these incentive programs are based on formal incentive programs,  others required special legislation.  Some states actually have slush funds (State “Deal Closing Funds”) specifically for this purpose.  This leads to my final point.

4.  Politicians retain a great deal of discretion in their offering of incentives.

I am concerned about the lack of attention to these programs.  They are costly, difficult to evaluate their effectiveness, and politicians have a ton of discretion in their use.  At the very least, it seems like a topic looking at more closely.