A case study in how media and researchers can exaggerate economic impact findings
A company in 2009 with just 75 employees was projected to grow so large that the plant it wanted to open in Michigan would create 27,142 jobs by 2024, or roughly the size of the entire population of the city of Garden City.
Sound too good to be true?
Not if you buy into the math used by advocates of the renewable energy ballot proposal and some of the mainstream media.
That 2009 company is A123 Systems and the Michigan Economic Development Corp. (MEDC) projected the battery manufacturer would create 2,217 jobs by 2024, not 27,142.
The 27,142 jobs would actually be "job years," a term used in a study done under a contract between Michigan State University and the Michigan Environmental Council, that uses job years to describe the economic impact a renewable energy ballot proposal would have if it were passed in November.
The MSU study clearly states that job years are based on jobs calculated for the life of a plant that lasts 20 to 30 years. So one job would equal 20 to 30 job years.
However, proponents of the renewable energy ballot proposal and some others say job years and jobs are interchangeable. If passed, those groups are reporting the renewable energy ballot proposal would create about 74,000 jobs even though the MSU study clearly states it's 74,000 job years.
Douglas Jester, a principal of 5 Lakes Energy LLC, a consultant who supports the energy ballot proposal, said the job years designation is used in economic projections in part so that jobs that took three months to complete weren't described as a year-long job.
However, when an economic impact analysis projects job growth for up to 30 years like the MSU study did, the difference between job years and jobs can be extensive.
For example, in 2007, the MEDC used an economic model to project that Google would create 2,033 total jobs by opening an office in Ann Arbor. But if the MEDC had used job years, the number would have jumped to 4,471, according to the projections in the MEDC’s internal memos. And if you look at the full 20-year analysis, the Ann Arbor office would have been responsible for 2,245 jobs by 2026, which would translate to 37,057 job years over that 20-year period.
Obviously in an analysis that spans 20 years or more, jobs and job years are not interchangeable.
Yet, that is what MSU researcher Charles McKeown claims, according to a Detroit News reporter who said McKeown explained to her the two terms were interchangeable. An Aug. 10 article in The News said the energy ballot proposal would create 74,000 jobs.
Crain's Detroit Business also originally reported the number as being jobs, but the business publication has since corrected its story online to report the numbers as job years instead of jobs.
McKeown didn’t respond to an email seeking confirmation of what he told The Detroit News.
The email sent from The News reporter read:
According to Charles McKeown, one of the three Michigan State University researchers who completed this study, the terms "jobs" and "job years" are used interchangeably in terms of economic modeling.
As you mentioned, the researchers defined a job year as "full employment for one person for 2080 hours in a 12 month span."
"Jobs and job years are used interchangeably," McKeown reportedly told The News. "The rest is trying to make a controversy out of semantics."
However, as the A123 Systems jobs forecast shows, there is a difference between projecting 2,217 jobs and 27,142 job years.
James Hohman, the Mackinac Center for Public Policy fiscal analyst who first caught the error being made by proponents, said it was "misleading the public" to say jobs were the same as job years in a report that looked at 20 years of jobs impact.