Sunday, July 5, 2020
By:
Units are notoriously important in physics, and the Mars Climate Orbiter (MCO) is the quintessential example demonstrating this truth. In 1999, NASA lost contact with the MCO after it was likely destroyed in the Martian atmosphere or escaped back into heliocentric space. The issue? Contractors had used customary units, but NASA engineers used metric units leading to the catastrophic failure that cost NASA the MCO.
This week, I was remined that units matter in the political sphere as well. I was writing a memo on a bill that was widely citied by officials as creating 100,000 jobs directly and through ripple effects in the broader economy. Notice the unit of measurement here, jobs. I looked into this statistic to verify that is was accurate which ultimately led me to a report from the Interior Department. The report estimated that the bill would create roughly 100,000 jobs-years.
What is a job-year you might ask? Well, according to the report, “job-years measure the total number of annualized full and part-time jobs accumulated over the 5-year duration of [bill]-related expenditures. Job-years is a measure of the quantity of employment supported by project expenditures and is not a measure of the number of workers. For example, if a construction project employs a worker for 18 months, this worker would be counted as 1.5 job-years (18 months/12 months in a year).”
The example at the end of the explanation really helps contextualize the metric, and it also highlights the discrepancy between the units used in the analysis and the unit used to advocate for the bill. Maybe it’s just me, but when I hear something is creating 100,000 jobs, I imagine 100,000 long-term jobs being created for 100,000 people. In reality, it could be 20,000 jobs existing for five years. Using a different unit of measure, jobs as opposed to job-units, isn’t technically wrong, but if you think about jobs reports in the way I do, it could be quite misleading.
I haven’t a single clue how widespread the job-years metric is in economic analyses; I imagine it’s more common in construction-type projects though. Either way, coming across the measure this week was an important reminder to mind the units and to understand how statistics are derived and employed in debate. It’s easy to throw out a bunch of numbers, but only additional research into the nitty-gritty will reveal the truth behind those numbers. The devil’s in the details after all.
Stay well,
Kyle Blasinsky