Why have economic development policies been so disappointing?

. . . and why do we keep doing the same failing policies over and over?

In 2016, I published a study on the failings of economic development policy. The focus was on Indiana, but the gist of the argument is essentially the same everywhere in the United States. There are a large number of interested actors (developers, economic development officials and some elected leaders) who want to keep the status quo. The study itself is here. Let me summarize the argument, which remains as fresh today as ever.

There is a large network of traditional economic developers, supported financially by municipal and county government and private firms (typically utilities). These officials usually act as a concierge to businesses considering relocating or opening a facility. This service often includes negotiating tax incentives, such as the ill fated Foxconn deal in Wisconsin.

The magnitude of incentives and the spending on these new companies are estunning. Indiana spends about $1.5 Billion a year or 12% of its annual state budget on incentives, subsidies and other inducements to relocate to the state. This is a relatively modest amount compared to surrounding states. So does this work?

The economic research on tax incentives is pretty lengthy, and mostly tucked in technical papers in academic journals. Fortunately, some of the best of it is published by Upjohn Institute's Tim Bartik, whose work largely argues for small, highly thoughtful incentives. I’m more dubious of the enterprise than he, but it is a matter of magnitude only. Little of the research offers a positive assessment of incentives as meeting even minimal cost-benefit criterion.

I make a somewhat different argument than much of the academic literature. I contend that the supply of ‘attractable’ businesses and jobs is now so small that attracting them can no longer be a reasonable public policy for state and local governments. Here’s why.

First, household consumption over the past century has shifted from 30% services to 70% services. I don’t show it, but business consumption is more service intensive as well (darned few web developers in 1974).

This is important because services are mostly produced locally, while goods can be produced anywhere. So, goods producing firms have ‘footloose’ jobs, while service producing firms have ‘non-footloose’ jobs. In the lingua franca of economics, these are often referred to as ‘exporting’ vs. ‘non-exporting’ jobs.

At the same time, the capital to labor ratio has risen enormously in the post-war years. In 1947, a million inflation-adjusted dollars of investment in manufacturing would also bring a community more than 22 jobs. By the mid 2000’s it had dropped to fewer than five. Often big investments actually destroy jobs through automation, so at the plant level, capital investment may not be a complement to labor. So, capital incentives just don’t attract the number of jobs they did in the past. This of course, is automation and technological change.

By the way, this is what it looks like in tomato canning. These are three photos of canning I scoured from the web. The first is in the 1920’s. Note, its a pole barn, benches, baskets, tables and knives. The second was thirty years later, a much streamlined process with conveyor belts. Finally, we have one from the 00’s. That was the hard one to find, because I wanted a human in the photo. Note the red jacketed worker.

Not surprisingly, this is accompanied by a big drop in the share of total income going workers in footloose businesses. Here I define footloose as manufacturing, movie and theater production, half of financial services, most logistics and a few other smaller sectors. This is the upper bound, since most of these aren’t really footloose.

So, how successful can a county or municipal government really be in attracting ‘footloose’ businesses and their workers? Not successful. This graph has some stale data in it, but since 1970, the U.S. has created some 100 million jobs in ‘non-footloose’ sectors and more than -8 million in ‘footloose’ sectors.

So, the policies that target firm relocation to a community will disappoint. Those that seek to attract a new ‘footloose’ firm, will disappoint. They are also costly, and often drain the community of resources better deployed to improve quality of place. But, that isn’t the incentive the economic development official faces.

The ticket to a new job for many developers is landing a big deal, especially a plant with 500 or more workers. Doing this almost ensures a local chamber of commerce employee the chance to be called up to the big leagues (a site selection firm) to make real money. But, what if that is now such a rarity that it is hardly worth pursuing? I mean, there are now so few of these big relocations that if they were randomly allocated to counties, you should expect one every 60 years or so.

It should be unsurprising that this work had the unintended consequence of causing many economic developers to write my university president and whine about my work. Anything I suppose to avoid dealing with the stark reality of figure 4, and the failed efforts to attract these jobs.