Democratic presidential candidate Andrew Yang calls himself "the Asian man who likes math."
He's pitched himself as an entrepreneur who understands technology and the economy, the kind of candidate most likely to bring a calculator to the Oval Office. It's a pitch that has impressed voters: He's sold 32,000 hats emblazoned with "MATH" and he ranked sixth in last month's NBC News/Wall Street Journal Democratic primary poll.
He's been campaigning on a promise to create a universal basic income — $1,000 a month — to save the nation from the many jobs that will be lost to robots. But the dire statistics he has used to make his pitch for the presidency don't always add up.
On the debate stage Tuesday, wearing a "math" pin on his lapel, Yang reeled off numbers and data points about automation, the digital revolution and a dying manufacturing industry. NBC News fact-checked several of those claims.
Manufacturing jobs lost to automation
In his closing remarks on the debate stage in Ohio Tuesday night, Yang referred to the loss of "4 million manufacturing workers here in Ohio and Michigan and Pennsylvania and Wisconsin and Iowa" thanks to the "fourth industrial revolution" — the onslaught of disruptive technologies like the internet, robotics and AI.
It's a data point he has used before on the trail, and it's way off the actual number of manufacturing jobs lost in those states, according to government data.
"I looked at the numbers,” he reportedly told a New York City crowd in June. "And Donald Trump is our president for one simple reason: We automated away 4 million manufacturing jobs in Michigan, Ohio, Pennsylvania, Wisconsin, Missouri, Iowa, all of the swing states that Donald Trump needed to win."
He said something similar in the June debate.
The five states Yang mentioned in the October debate lost approximately 1.09 million manufacturing jobs combined between the peak of national manufacturing employment, 2000, through October 2016, according to data from the Bureau of Labor Statistics. Add in Missouri, as Yang did in New York, and it's approximately 1.2 million jobs lost.
Pressed on why he was using bad data, Yang’s campaign repeatedly pointed to nine-year-old nationwide numbers, later insisting the candidate meant the entire country, not the five or six states he has listed again and again with this data point.
Let our news meet your inbox. The news and stories that matters, delivered weekday mornings.
Nationwide, the country lost 4.9 million — not 4 million — manufacturing jobs during the same period.
Additionally, Yang attributes these job losses to automation. Pressed for evidence, his campaign cited a Ball State University study that found that the majority of job losses were due to automation, not trade.
That single study, as Slate points out, has been widely criticized by other researchers for overstating automation’s impact.
Automation is not the only force affecting manufacturing jobs. International trade deals has made it cheaper to manufacture goods out of the U.S., resulting in outsourcing. Additionally, the 2007 recession slammed the manufacturing industry, too, according to government jobs data.
His campaign declined to comment on the use of inaccurate or misleading data. But on Friday, Yang spoke out about misinformation during a 10-hour question-and-answer livestream.
"It upsets me when people take aim at facts for a particular agenda," Yang said. "I mean, let's have a data-driven discussion. The intellectual laziness in certain quarters has been deeply disappointing."
Amazon's big business and low taxes
"Amazon alone is closing 30 percent of America's stores and malls and soaking up $20 billion in business while paying zero in taxes," Yang said during the debate. He made a version of this claim in the July debate, too.
Amazon isn't personally buying up stores and malls and closing them (they've actually opened some of their own brick-and-mortar shops). Online shopping has dramatically changed American commerce, of course, but Amazon isn’t the only name in the game. The online giant had a market share of 38 percent this year — down from 49 percent the year prior.
They are not “alone” responsible for closing stores and malls, as Yang said. His campaign later said they were simply "part" of the trend of retail stores shuttering.
The 30 percent number is itself one man’s estimate. One expert told Forbes he expected that roughly 30 percent of the nation’s malls will close or be repurposed over the next decade, though analysts in 2017 at Credit Suisse pegged the number of coming closures lower. Neither study blames Amazon alone.
Amazon’s tax bill — or namely, the lack of it — is a topic numerous Democratic candidates have zeroed in on, after one study concluded the company’s federal tax bill is zero.
But the study doesn't examine state and local taxes, and as The Wall Street Journal has reported in depth, the company’s tax returns are private and it’s impossible to exactly know the company’s payout. Asked by NBC News, the company declined to offer details of its tax returns, saying that they pay all legally required taxes.
Investor materials say the company paid $1.2 billion in income taxes around the world in 2018, including an estimated $129 million federal tax refund and a $322 million bill for state taxes, suggesting that while they may not pay a lot of taxes — they pay some.
It’s unclear exactly what Yang means when he says Amazon is soaking up $20 billion in business. The company raked in about $11 billion in profits last year, and did about $233 billion in revenues.
People's data is worth more than oil
"Right now, our data is worth more than oil," Yang said on the debate stage Tuesday. There’s no way of knowing if this is correct or not, and it's the subject of wide debate.
As of Friday, oil was sold for about $53 a barrel; it’s an energy source and a commodity, giving it global and continuous selling power. Data, on the other hand, is hard to value. It undoubtedly makes corporations money, but studies are inconclusive on what it’s actually worth, in part because it’s in fact limitless and of varying worth.
The data of a frequent online consumer in a wealthy country, for instance, might be more valuable than the data of an individual who rarely logs online, or perhaps doesn't have regular internet access. But at the end of the day, neither can sell their data.
Asked for more information on this claim, Yang's campaign sent NBC News an Economist opinion piece.
More opioid prescriptions than people in Ohio
On the opioid crisis, Yang said Tuesday, "There was a point when there were more opiate prescriptions in the state of Ohio than human beings in the state of Ohio."
This is true. The data comes from 2010, when there were 102.4 opioid prescriptions per 100 persons.