In a world where data and narrative often collide, Bloomberg’s recent analysis claiming that 94% of jobs created in 2021 went to racial minorities serves as a striking example of what happens when statistical interpretation goes off the rails. At face value, such a number seems almost impossibly skewed—one that raises eyebrows not just among analysts but anyone with a grasp of basic arithmetic. Yet, despite glaring flaws in methodology and interpretation, Bloomberg stood firmly by its report.
At the core of this statistical misstep is a fundamental misunderstanding—or perhaps willful disregard—of how workforce turnover and job creation actually work. Bloomberg analyzed employment data from S&P 100 companies, focusing on Equal Employment Opportunity Commission (EEOC) filings. These reports provide annual racial breakdowns of a company’s workforce but lack granular details about who was hired into new roles versus who replaced existing roles.
The error begins with Bloomberg assuming that every demographic shift in the overall workforce came from newly created jobs rather than the far larger pool of replacement hires. They saw a net increase of 320,000 jobs across the companies in question and simply divided the demographic changes by that number—ignoring the reality that workforce turnover far exceeds net growth.
To put it in perspective: turnover rates across large companies typically hover around 18%. That means roughly 1.6 million employees likely left their jobs in 2021 across the S&P 100, in addition to the 320,000 newly created positions. So instead of analyzing demographic shifts across 320,000 jobs, Bloomberg should have been looking at nearly 2 million hires.
This oversight isn’t just a rounding error—it’s a fundamental flaw that skewed every conclusion Bloomberg drew.
When corrected for workforce turnover, the numbers paint a much more balanced picture. About 46% of new hires in 2021 were white, a figure slightly below the 54% white composition of the existing workforce. Hispanics saw a modest rise from 17% to 20% of new hires, Asians increased from 10% to 12%, and Black representation crept up slightly from 17% to 18%.
Far from being a sweeping revolution in corporate diversity or a dystopian display of systemic discrimination against white workers, the numbers reflect a gradual demographic shift consistent with broader generational trends. Younger, incoming workers are naturally more diverse than older, retiring ones.
The hypothetical example provided in this analysis drives the point home: imagine two companies, each with 100 employees, all of them white. One replaces 20 retiring employees with Black hires, while the other creates 10 new positions, all filled by white hires. If Bloomberg’s flawed methodology were applied, they’d claim that 200% of new jobs went to minorities. It’s an absurd conclusion, but it mirrors the kind of statistical gymnastics Bloomberg performed on a grand scale.
Robert VerBruggen of the Manhattan Institute summed it up succinctly: “If whites are disproportionately retiring and non-whites are disproportionately getting jobs, that will skew the numbers, giving the appearance of a serious commitment to equity—or mass violations of civil-rights law, depending on one’s perspective.”
Pollster Patrick Ruffini echoed this, calling Bloomberg’s analysis “statistically illiterate”. If retirees are predominantly white and younger hires reflect the more diverse composition of America’s younger generations, demographic shifts in hiring are not only expected but mathematically inevitable.
This isn’t just a debate about data interpretation—it’s about the power of narrative. Bloomberg didn’t just get the math wrong; it broadcast a fundamentally misleading story under the banner of rigorous analysis. And notably, none of the companies highlighted in the report stepped in to correct the record. Why? Because the error flattered them. The idea that corporate America had delivered on its post-2020 diversity pledges made for a tidy, positive headline.
In an era where trust in media is already precariously low, such missteps carry heavy consequences. Data journalism—especially on complex, sensitive topics like race and employment—requires precision and intellectual honesty. When those standards slip, the resulting misinformation doesn’t just muddy the waters; it erodes trust in legitimate analysis and feeds skepticism from all sides.
Bloomberg’s insistence that it “stands by its reporting and analysis” despite clear and irrefutable errors speaks to a deeper issue. Journalism, especially at outlets as prestigious as Bloomberg, should serve as a clarifying force, not a distorting one.