With another dubious assumption, Harvard study authors tilt the scales against mining

Recently, I wrote about a paper by two economists from Harvard University, James H. Stock and Jacob T. Bradt, which argued that “Over time, the economic benefits of mining would be outweighed by the negative impact of mining on the recreational industry and on in-migration.”

There was no real rationale offered by Stock and Bradt for their central assumption, namely that if the proposed Twin Metals operation goes ahead, “recreational employment” will contract at either 1.2% or 2.4% annually. Indeed, a look at the data on employment for Marquette County in Michigan, home to the Eagle Mine since 2014, finds that Leisure and hospitality employment has actually risen since the mine opened.

But, as my colleague Isaac Orr explained in the Star Tribune this weekend, there are further concerns about Stock and Bradt’s paper. In it, they note that they “omit induced employment effects”. Economic modelers IMPLAN define ‘induced effects‘ as

The response by an economy to an initial change (direct effect) that occurs through re-spending of income received by a component of value added. 

In other words, when a mine opens the jobs created in the mine are direct effects, the jobs created in the mine’s suppliers are indirect effects, and the jobs created by the spending of the workers employed in the direct and indirect effects are the induced effects.

Stock and Bradt exclude these numbers from their analysis. They do so for two reasons.

First, there is considerable uncertainty associated with estimating induced effects in a region because those estimates depend on the availability of unemployed or underemployed resources locally and on the amount of income that is retained locally rather than spent elsewhere in the state or out of state; this uncertainty calls into question the value of computing induced effects.

There is, of course, uncertainty in all economic projections, Stock and Bradt’s included. Nevertheless, induced effects are a standard part of economic impact analysis.

Second, there is in any event no reason to think induced effects would differ depending on the income source so they would be proportional to direct plus indirect income changes for both the mining and hospitality industry.

This is dreadfully wide of the mark. It is true that “there is…no reason to think induced effects would differ depending on the income source”. If you created jobs for ten hockey players on $100,000 each and ten dancers on $100,000 each, the induced effects would be the same (assuming the indirect effects were also).

But that is simply not the case with jobs in mining and jobs in leisure and hospitality. Figure 1 shows average annual pay in all private establishment sizes in 2017 in St. Louis County for Metal ore mining and Leisure and hospitality. For mining the figure was $88,885, for leisure and hospitality it was just $16,542. As mine workers have more money to spend than leisure and hospitality workers, it follows that their will be more induced jobs resulting from mining than from leisure and hospitality.

Figure 1: Average Annual Pay for All establishment sizes in St. Louis County

Source: Bureau of Labor Statistics 

And that is what the IMPLAN modeling software shows. Table 1 shows the results for Minnesota generated by IMPLAN of 100 new jobs in both copper mining and full service restaurants holding all else equal. Because of the higher wages in the direct effects and the larger scale of the supply chains in the indirect effect, 100 copper mining jobs generate 128 induced jobs while 100 restaurant jobs generate just 20.

Table 1: Economic impacts of 100 new jobs in Minnesota

Source: IMPLAN

Induced jobs are 15% of the total effect for restaurant employment but 46% of the total effect for copper mining employment. Excluding the induced effects will lower the expected benefits of both, but that the reduction will be greater for mining.

In short, by excluding induced jobs based on such a bizarre assumption, Stock and Bradt give a distorted picture of the wider impacts of employment in mining and leisure. As with their assumption of jobs losses in the leisure industry, this serves to make mining look relatively less beneficial than leisure.

John Phelan is an economist at the Center of the American Experiment.