Stand up for the facts!
Our only agenda is to publish the truth so you can be an informed participant in democracy.
We need your help.
I would like to contribute
During the Democratic debate in Westerville, Ohio, Julian Castro took a shot at President Donald Trump’s stewardship of the economy — especially in three key Midwestern states Trump won in 2016.
"All of us are out there every single day talking about what we're going to do to make sure that more people cross a graduation stage, that more families have great health care, that more folks are put to work in places like Ohio, where Donald Trump has broken his promises," Castro said, "because Ohio, Michigan, and Pennsylvania actually in the latest jobs data have lost jobs, not gained them."
Castro’s campaign got back to us with an explanation, but it relies on an atypical reading of the numbers to achieve its effect.
Here’s what we found when we looked at the official Bureau of Labor Statistics numbers in the simplest, most traditional fashion.
The latest official release state-by-state employment trends is current through August 2019. Here’s what you get if you compare the August 2019 figures to the July 2019 figures:
Ohio: A gain of 3,700 jobs.
Michigan: A gain of 6,100 jobs
Pennsylvania: A gain of 16,500 jobs.
So how does Castro support his claim of job losses in those states?
His campaign told PolitiFact that he compared March figures with July figures, showing that employment dropped in Ohio by 7,300, in Michigan by 11,500, and in Pennsylvania by 4,500.
Note that Castro used July, rather than August, even though August numbers are available. He argued that July is a better measurement because the August numbers will be revised before being made official.
The August numbers will indeed be revised — here’s a primer on employment data revisions we put together recently — but it’s not a common practice to simply wave away the most recent figures because they will eventually be revised. All employment numbers are eventually revised, yet economists regularly study the most recent ones, politicians talk about them, and the media covers them.
More to the point, Castro didn’t provide any indication in his statement that he was using a non-standard comparison. What his figures amount to are a cherry-picked month-to-month quirk that paints a misleading picture of the job trends in each of these states.
For instance, here’s the change for the full-year period between August 2018 and August 2019:
Ohio: A gain of 24,800 jobs.
Michigan: A gain of 16,000 jobs
Pennsylvania: A gain of 38,400 jobs.
And here are the employment trends since Trump took office in 2016, using BLS data collected by the Federal Reserve Bank of St. Louis. Each state has seen its employment levels rise.
And here’s Pennsylvania:
Castro said, "Ohio, Michigan, and Pennsylvania actually in the latest jobs data have lost jobs."
Official government figures show that each state gained jobs between July 2019 and August 2019, the latest month-over-month period available. To arrive at his job-loss figure, Castro cherry-picked a time period that he didn’t mention in his remarks — a time period that ignores the existence of the most current data.
His assertion that employment levels have been crumbling in these states is belied by data showing that each state has seen job gains both in the past year and since Trump took office.
We rate the statement False.
Washington Post, debate transcript, Oct. 15, 2019
Bureau of Labor Statistics, "Table 3. Employees on nonfarm payrolls by state and selected industry sector, seasonally adjusted," accessed Oct. 15, 2019
Cincinnati.com, "Democratic debate: Here's what the candidates had to say about Ohio," Oct. 15, 2019
Federal Reserve Bank of St. Louis, Ohio employment level, accessed Oct. 15, 2019
Federal Reserve Bank of St. Louis, Michigan employment level, accessed Oct. 15, 2019
Federal Reserve Bank of St. Louis, Pennsylvania employment level, accessed Oct. 15, 2019
Read About Our Process
In a world of wild talk and fake news, help us stand up for the facts.