Economic impact studies are common. They usually are invoked by government or special interest groups to demonstrate how a particular spending or tax incentive policy will affect variables such as employment and economic growth in a particular geographic area. The John Locke Foundation has argued repeatedly that these studies are an example of bad economics. This should come as no surprise. For the most part, these studies are done by consultants rather than credentialed economists. These consultants are experts at manipulating the commercial models used to generate the impact numbers. Two notable examples are IMPLAN (Impact Analysis for Planning) and REMI (Regional Economic Modeling, Inc.).

For several years now, the John Locke Foundation has been pointing out the fundamental flaw in all of these studies, namely that they are designed to give only one kind of result – positive. In other words, they will always show GDP and employment gains from the government program they are assessing. The question for the studies is typically not about whether the program will enhance job and GDP growth but by how much they will do so. This is the case because they ignore what economists call opportunity costs: the idea that all resources are scarce and therefore have alternative uses. The value of these alternative uses, in terms of GDP and job gains, have to be considered before one can reach any conclusion about whether the program will, on net, be beneficial. Instead of making these necessary calculations, the authors of these studies have what they believe is a way to legitimize ignoring them. They either explicitly or implicitly invoke what I refer to as the “manna from heaven” argument. It goes something like this: if the money that funds the program or initially stimulates a pattern of spending that comes from outside the geographical area for which the economic impact is being assessed, opportunity costs can be ignored. In fact, they can be assumed not to exist.

This is because, as the argument goes, if the money for the project under consideration is coming from Washington, D.C. or a state government (if you are assessing the impact on a particular region within a state) or an out-of-region business being subsidized with spending or tax breaks, it can be treated as “manna from heaven.” It is “new spending” that is flowing into the district under consideration and would have no opportunity costs because it is not being diverted from other uses.

For example, this is how the authors of the much-publicized study arguing that Medicaid expansion, funded mostly by the federal government, would create over 37,000 jobs in North Carolina, justify ignoring opportunity costs. They claim that:

[O]ur methodology addresses this [opportunity cost] problem by being based entirely on the net federal funds that will flow into the state solely due to Medicaid expansion; we exclude the use of state funds which might be used for other purposes. The additional federal matching funds derive from external sources and would not flow into North Carolina if there was no Medicaid expansion.

This kind of assumption, although not made as explicitly, underlies the majority of economic impact studies. For example, a recent study of the impact of state subsidies to UNC Asheville on the Asheville region implicitly made a similar assumption because the subsidies flowed into the region from taxpayers statewide. It is money that would not otherwise have been spent in the region, i.e., manna from heaven, or in this case from Raleigh.

The problem is that this assumption does not, in fact, absolve these studies from their obligation to consider opportunity costs or qualify as legitimate economic analysis. This is because real opportunity costs are not about revenues or where the money comes from but about resource use. And while the money being spent on the project may come from “somewhere else,” i.e., be manna from heaven, the actual resources – the workers, the land, and the capital – are not. The spending, wherever it comes from, actually represents an added claim on limited resources, including workers, that are likely employed by someone else. When a job is “created” by government, which necessitates new spending, a worker is being bid away from other businesses and potential investments in the area. The same is true for all the resources that are being purchased with the new and additional spending that it generates. Unless an economic impact study takes account of these alternative resource uses in terms of lost jobs and GDP, they should not be considered useful.

The idea commonly adopted by those who produce economic impact studies, namely that so long as money for a government spending project comes from “somewhere else” there will be no negative consequences for jobs or economic growth that have to be considered, indicates a distinct lack of basic economic knowledge. And it is not just a minor error. In making this error, the analysts are essentially rendering their conclusions useless.