From Jobs to Podcasts, It Pays To Be Skeptical About the Biases and Blunders of Bean Counters

A suggestion that the Bureau of Labor Statistics might stop issuing jobs numbers is outraging those who hold the statistics sacred. But when everything from test scores to temperatures are shouted with mathematical precision only to be revised with a whisper, the uproar is a healthy reminder to question the bias of bean counting.The wrangling over numbers began when President Trump fired President Biden’s BLS commissioner, Erika McEntarfer, and his replacement, E.J. Antoni, floated ending monthly employment reports. The White House later said its “plan” and “hope” is to keep preaching the BLS gospel, eliciting sighs of relief from the faithful. The White House laid out its case in a statement documenting “inaccuracies and incompetence” at the BLS that had “completely eroded public trust.” The BLS had “revised down … payroll growth for the year ending in March 2024 by 818,000 jobs — the second-largest benchmark revision on record.” Partisans were left wondering if Ms. McEntafer had put her finger on the scale to help Vice President Kamala Harris. In 2002, a CNN commentator, Robert Novak, reported “that the Clinton administration” did just that. It “grossly overestimated the strength of the economy leading up to the 2000 election” to show “the economy on an upswing through most of the election year while in fact it was declining.”Novak wrote that there was “a political motive for Democratic cooking of the government’s books,” and that the estimate “in 2000 was conducted by career public servants who” were kept in their jobs by the Republican, President George W. Bush.Mr. Trump chose to shake up the BLS rather than risk that their cooks would keep serving up flawed numbers. “Garbage in, garbage out,” an IBM programmer, George Fuechsel, is credited as saying, a warning that it’s best to follow when digesting all flavors of data. Take polling averages that mix surveys using different sample sizes, demographics, or methodologies to aggregate apples and oranges. There are also “technical errors” like the ones the White House cited at BLS, or the CDC “coding error” that invented 72,000 Covid-19 fatalities in 2022. On Wednesday, the Hollywood Reporter reported that “podcast metrics can’t be trusted,” either. Hollywood studios are notorious for keeping the production budgets secret to make films seem profitable. X, meanwhile, fights an ongoing battle against fake accounts and bots that boost follower counts.The heat index is just as unreliable, based on subjective “feels like” metrics. In 2019, Slate.com called the wind chill factor “a meaningless number,” too, and “the weatherman’s favorite alarmist statistic.” It’s based on explorers in 1945 who observed plastic water bottles freezing at different rates in the Antarctic winds.Even numbers themselves can be used to play with our minds. According to a 2013 Columbia University Business School study, odd numbers appear more specific and convey credibility. Marketers have long applied this knowledge, pricing items at $9.99, say, rather than $10 even. On Wednesday, the Illinois State Board of Education reduced the benchmarks for proficiency, which are used for standardized tests. Changes like this explain why grades rise but admission scores fall. “ACT scores are down,” as Reason put it in October, “grades are up. Something’s fishy.” Fake figures help explain why Afghan security forces, touted as ensuring its future, didn’t show up after Mr. Biden’s military withdrawal and resulted in the Taliban surging back into power unopposed.A former finance minister of Afghanistan, Khalid Payenda, told the BBC in 2021 that government officials had created “ghost soldiers” to show Washington progress and keep its money flowing. “Most of the 300,000 troops and police on the government’s books,” the story said, “did not exist.”Data give politicians, investors, and voters information to weigh when making decisions. Yet the numbers are only as trustworthy as the people and methodology that deliver them. A healthy skepticism, well short of blanket denials, is the best way to avoid being duped when what’s gospel today turns out to be garbage tomorrow.

From Jobs to Podcasts, It Pays To Be Skeptical About the Biases and Blunders of Bean Counters

From Jobs to Podcasts, It Pays To Be Skeptical About the Biases and Blunders of Bean Counters

More from the New York Sun

Andy Warhol — Before His First 15 Minutes of Fame

Anthropic Bets Its Smartest Model Will Set Claude Duet Apart

‘Wanted’: How the West Was Overwrought

Alexander the Great Conquered the World — and Left It in Ruins

Vance May Pass on 2028 Presidential Bid After Midterm Loss

Justice Department May Take On New Role After Democrat Oversight Chief Warns Administration To ‘Lawyer Up’

Blanche Scoffs at Allegations Wiles Stymied MAHA at Behest of Food, Drug Companies

UFC Boss as White House Press Secretary? Trump Could Be Asking for a Fight

More from the New York Sun

Andy Warhol — Before His First 15 Minutes of Fame

Anthropic Bets Its Smartest Model Will Set Claude Duet Apart

‘Wanted’: How the West Was Overwrought

Alexander the Great Conquered the World — and Left It in Ruins

Vance May Pass on 2028 Presidential Bid After Midterm Loss

Justice Department May Take On New Role After Democrat Oversight Chief Warns Administration To ‘Lawyer Up’

Blanche Scoffs at Allegations Wiles Stymied MAHA at Behest of Food, Drug Companies

UFC Boss as White House Press Secretary? Trump Could Be Asking for a Fight