The Mean-Median Gap

In a perfectly symmetric wage distribution — imagine a world where exactly half the workers earn above a number and half below, with the same spread on each side — the mean and median are identical. Report either one and you've described the center of gravity accurately. But virtually no occupation in the American economy is symmetric. In 760 of 797 occupations with valid data, the mean exceeds the median. Only 37 occupations — less than 5% — show the reverse. The American labor market is, with rare exceptions, a right-skewed machine. The top pulls the average above the typical experience.

The mean-median gap measures this skewness in practical terms. If the median wage in an occupation is $60,000 and the mean is $72,000, the gap is $12,000 or 20%. That 20% tells you that a cluster of high earners — the rainmakers, the senior partners, the workers in high-cost markets — are inflating the average by a fifth. Anyone who reads a headline saying “the average X earns $72,000” and assumes that's the typical experience is off by $12,000, or roughly one month of take-home pay.

Across all occupations, the average gap is 9.1% — meaning the mean exceeds the median by about a tenth, on average. The median gap (the gap for the typical occupation) is 7.3%, pulled lower because a few extreme outliers inflate the average gap itself. Even the measure of skewness is skewed. The distribution of the gap percentages ranges from –10.9% (subway and streetcar operators, where the mean is below the median) to +316.5% (athletes and sports competitors, the single most right-skewed occupation in the economy). Between those poles lies a map of where top earners matter most — and where they barely register.

To understand the stakes, multiply the gap by headcount. If the mean-median difference for a given occupation is $10,000 per worker and the occupation employs 500,000 people, the aggregate “extra payroll” above what the median-based estimate would suggest is $5 billion per year. This isn't imaginary money. It's real wages paid to real workers — but concentrated in the upper portion of the distribution. Across the entire economy, total payroll computed at mean wages is $10.02 trillion. Computed at median wages, it's $9.13 trillion. The difference — $890 billion — is the aggregate payroll that exists because wage distributions are right-skewed rather than symmetric. It's the total price tag of having top earners in every occupation.

The occupation with the most extreme mean-median gap isn't even close to the second. Athletes and Sports Competitors have a median wage of $62,360 and a mean of $259,750 — a gap of $197,390, or 316.5%. The mean is more than four times the median. How is this possible? Consider what the BLS is counting. This occupation includes roughly 14,370 workers: minor league baseball players earning $35,000, semi-professional soccer players earning $50,000, Olympic-caliber athletes on modest stipends — and then a relatively small number of NBA, NFL, MLB, and NHL players earning millions. LeBron James and a minor leaguer in Tulsa share a BLS code. The handful of athletes earning $20 million or more per year yank the mean into the stratosphere while the median — reflecting the 50th percentile athlete, probably a journeyman in a lower-tier league — sits at $62,360. No other occupation produces this kind of dispersion.

The BLS doesn't even report a 90th percentile for athletes, because the right tail is so extreme that the survey can't reliably estimate it. The P90 field is blank — a rare occurrence that speaks to the radical concentration of earnings in professional sports. The mean of $259,750 is itself a dramatic undercount of what the top earners actually make, because BLS wage estimates are capped at certain thresholds and because many top athletes' compensation arrives through endorsements, not payroll. The real mean — including all forms of compensation — would be dramatically higher.

Beyond athletes, the most skewed occupations cluster into recognizable categories. Media and entertainment dominate: news analysts and reporters (75.9%), radio and TV announcers (75.4%), producers and directors (36.9%), music directors (32.3%), and photographers (30.9%). These are all winner-take-all fields where a small number of stars earn orders of magnitude above the median practitioner. Anderson Cooper and a local news reporter in Omaha share an occupation code; the former earns $20 million, the latter $45,000. The math produces a 75.9% gap.

Commission-based finance accounts for the next cluster: personal financial advisors (56.9%), securities sales agents (41.3%), and insurance sales agents (35.0%). In these occupations, the gap is structurally embedded in the compensation model. A financial advisor managing $500 million in assets generates fee income that is orders of magnitude above an advisor with $5 million under management. Both earn commissions; one's commission base is 100 times larger. The mean reflects the advisor with the big book; the median reflects the advisor grinding cold calls.

Management is the third skewness factory: general and operations managers (29.3%), chief executives (27.4%), property managers (24.0%), and health services managers (16.8%). The management category is inherently skewed because it aggregates across employer size and industry. A general manager of a 10-person landscaping company earns very differently from a general manager overseeing a $2 billion division of a Fortune 500 firm. But both carry the same BLS code (11-1021), and the small number of high-tier managers pulls the mean far above the median.

The most important statistic in the mean-median gap story isn't the highest percentage — it's the highest dollar volume. When you multiply each occupation's per-worker gap by its employment count, you get the total payroll attributable to skewness — the extra dollars that flow to the upper portion of the distribution above what a median-based estimate would predict. The results are astonishing.

General and Operations Managers are in a category of their own. With 3.58 million workers and a per-person gap of $30,170, this single occupation accounts for $108.1 billion in skewness — more than the next four occupations combined. To put $108 billion in perspective: it exceeds the entire GDP of Ecuador. It is roughly $875 for every American household. And it flows to a relatively small subset of the 3.58 million people who carry this job title — specifically, the senior operations leaders at large corporations, the divisional heads, and the managers of high-revenue establishments whose compensation pulls the mean almost $30,000 above the median.

Why is this one occupation so dominant? Three reasons. First, it is enormous — the sixth-largest occupation in the country. Second, it is wildly heterogeneous. The BLS code 11-1021 captures everyone from the general manager of a local Subway franchise to the COO of a regional hospital system to the operations lead of a logistics firm with 2,000 employees. Third, its compensation is tied to the scale of the operation managed, which varies by orders of magnitude. Managing a $5 million restaurant and managing a $500 million manufacturing division are both “general management,” but the latter commands 5–10x the pay. The top of the distribution isn't slightly above the median; it's in a different economic atmosphere.

The second entry, lawyers ($23.6 billion in total gap), illustrates a different version of the same dynamic. BigLaw associates in New York, Chicago, and San Francisco start at $235,000 and reach partner-level incomes well above $500,000 — but those firms employ a small minority of all lawyers. Most attorneys are solo practitioners, public defenders, prosecutors, or in-house counsel at mid-size companies earning $80,000 to $130,000. The median lawyer earns $151,160 — already high — but the mean is $182,760, pulled upward by a right tail of partners, named partners, and rainmakers at elite firms. The gap is $31,600 per lawyer, and across 748,000 lawyers, that sums to $23.6 billion.

Perhaps the most surprising entry is registered nurses, ranked seventh with $15.9 billion in total gap. Nurses have a small percentage gap (5.2%), which makes them appear relatively egalitarian. But they employ 3.28 million people — the third-largest occupation in America. Even a $4,830 per-nurse gap, multiplied by 3.28 million, produces a substantial sum. The skewness in nursing comes from travel nurses and specialized practitioners (CRNA-adjacent roles that share the RN code in some classifications) who earn well above the median $93,600, combined with the geographic premium that sends San Francisco nurses above $130,000 while rural Alabama nurses hover near $66,000. A 5% gap sounds modest until you realize it operates at the scale of an army.

The sector map of skewness mirrors, but does not perfectly replicate, the inequality map from Episode 5. Arts, Design, Entertainment & Media leads again with an average gap of 29.5% — nearly triple the economy-wide average of 9.1%. This sector is the skewness capital of the American labor market. The typical arts occupation has a mean that's nearly a third higher than its median, reflecting the winner-take-all dynamics that concentrate enormous compensation at the top of every creative field. A handful of A-list actors, directors, and designers in each category distort the average for entire professions.

Sales (15.8%) and Business & Financial Operations (14.7%) follow, both reflecting commission-based or performance-linked compensation structures that systematically produce right-skewed distributions. Management (14.0%) and Legal (14.0%) round out the five most skewed sectors. Together, these five major groups — arts, sales, business, management, and legal — employ approximately 37.4 million workers, or about a quarter of the workforce. But they account for a disproportionate share of the total $890 billion payroll gap, because their high mean-median percentages operate on substantial wage bases.

At the other end, Food Preparation & Serving (4.7%), Production (4.8%), and Healthcare Support (4.9%) have the smallest gaps. In these sectors, the mean barely exceeds the median — the distribution is close to symmetric. Why? Because there is essentially no mechanism for an individual food preparation worker or production assembler to earn dramatically more than their peers. The wage structure is flat: governed by hours, shifts, and seniority rather than individual output. There are no star dishwashers pulling down $200,000 and distorting the average. The ceiling is low, the floor is firm, and the distribution stays tight.

Notice the pattern: the sectors with the smallest gaps are the ones we identified in Episode 5 as having the lowest P90/P10 ratios. Low within-occupation inequality and low mean-median gaps go hand in hand. When there's no long right tail, there's nothing to pull the mean above the median. But the relationship isn't perfectly linear. Healthcare practitioners have a modest gap (6.1%) despite including some very high-earning specialties, because the massive base of moderately-paid nurses and therapists anchors the median close to the mean. Education has a higher gap (12.7%) than you might expect, driven largely by postsecondary teachers — where adjunct-to-full-professor spreads create dramatic skewness — and by school administrators who command six-figure salaries above the teaching median.

Only 37 occupations in the entire OEWS data show a negative mean-median gap — where the average worker earns less than what the mean reports. These are left-skewed distributions: the lower end of the pay range pulls the average down below the typical experience. They are rare enough to be individually interesting.

The left-skewed occupations come in two flavors. The first is regulated, union, or government-scale positions where a pay ceiling caps the right tail. Subway and streetcar operators (−10.9%) are the most extreme example: their wages are set by transit authority contracts, with a ceiling defined by the top of the union scale. Most operators cluster near the top of the scale (the median is $84,830), but a cohort of newer hires at the bottom drags the mean down to $75,620. The distribution isn't bell-shaped — it's piled up against the ceiling with a left tail of recent hires. Judges (−8.4%) work the same way: most sit on state courts earning near the median of $156,210, but a subset of part-time or small-jurisdiction judges at the bottom pulls the average below the typical salary.

Power plant operators (−3.7%), nuclear technicians (−3.4%), electrical repairers (−3.4%), power-line installers (−2.6%), air traffic controllers (−1.3%), and elevator installers (−1.6%) all share a common trait: they are unionized or licensed occupations in capital-intensive industries where pay is high, structured, and ceiling-bounded. There are no star power plant operators earning $500,000. The scale tops out, and the only variation comes from the bottom — newer workers who haven't reached the top step yet. This inverts the usual pattern: instead of a small group of top earners pulling the mean up, a small group of new or part-time earners pulls the mean down.

The second flavor is flat low-wage occupations where even the bottom is near the floor. Dishwashers (−1.3%) have a median of $33,670 and a mean of $33,220 — nearly identical, with the tiny gap reflecting a handful of below-minimum-wage situations (likely tipped or training positions) that slightly depress the average. School bus drivers (−0.8%) show the same pattern: a part-time workforce where most earn near the center of a narrow band, and the left tail (drivers working fewer hours or in lower-wage districts) barely nudges the mean below.

These 37 occupations collectively employ about 1.7 million workers — just 1.1% of the workforce. Their existence is a useful reminder that left-skewed wage distributions are theoretically possible, but practically rare. The American labor market is overwhelmingly structured to produce right skewness: there's almost always someone at the top earning dramatically more, and almost never a structural ceiling that prevents them from doing so. Unions, government pay scales, and regulated industries are the only consistent producers of the reverse pattern.

The mean-median gap is not merely an academic curiosity. It has practical implications for anyone evaluating a career, negotiating a salary, or trying to understand labor market statistics.

When someone reports an “average salary” for an occupation, ask whether they mean the mean or the median. In 760 of 797 occupations, the mean is higher — often substantially. If you see that “the average financial advisor earns $160,210,” know that the median is $102,140, and most advisors earn well below that reported average. Recruiting materials and career guides overwhelmingly cite the mean, because it's the higher, more attractive number. The BLS publishes both; most secondary sources cherry-pick.

A high mean-median gap signals an occupation where upside is real but unevenly distributed. Personal financial advisors (56.9% gap) offer enormous potential earnings — but only to those who build large client portfolios. Securities sales agents (41.3%) can earn $215,000 at the 90th percentile — but only if they land institutional accounts. Insurance agents (35.0%) have uncapped commission potential — but the median reality is $60,370. If you're evaluating one of these careers, the mean tells you what's possible; the median tells you what's probable. The gap between them measures the career lottery you're entering.

A low gap signals an occupation where what you see is what you get. If the mean-median gap for dental hygienists is 3.8% ($86,570 median, $89,850 mean), you can trust the average as a reasonable forecast of your likely earnings. There's no hidden upside of superstars pulling the average away from reality. The distribution is compact, the career path is predictable, and the reported average is honest. For risk-averse workers, low-gap occupations are more reliable income commitments than high-gap ones, even if the reported averages look similar.

The gap widens with seniority in many occupations. This is because tenure effects compound: the general manager who started 20 years ago and now oversees a large division is the one pulling the mean above the median. Early-career workers cluster near or below the median; mid-career workers approximate it; late-career stars explode above it. If you're entering a high-gap occupation, the mean-median split is a preview of the divergence your cohort will experience over time. Some of your classmates will earn at the 90th percentile; most will not.

For the biggest occupations in the country, the gap varies dramatically. At one extreme, general and operations managers show a 29.3% gap — by far the highest among giant occupations. This means the commonly cited “average manager salary” of $133,120 overstates the typical manager's experience by nearly $30,000. At the other extreme, food preparation workers show a −0.8% gap — their mean ($33,940) is actually below their median ($34,220). For 889,000 food prep workers, the average and the typical experience are essentially identical.

Among the very largest occupations (those employing over a million workers), wholesale sales representatives (22.0%), service sales representatives (22.6%), and accountants (14.5%) show the widest gaps. These are all occupations with meaningful variation in employer type and performance-based compensation. Meanwhile, fast food workers (2.9%), stockers (4.9%), and nursing assistants (4.4%) have near-zero gaps. Their wage structures are so flat that the mean and median converge. When you hear that “the average fast food worker earns $31,350,” you can take that number almost at face value. When you hear that “the average sales representative earns $81,260,” you should mentally subtract 20% to get closer to the typical experience.

The most interesting case is software developers: a relatively modest 8.6% gap despite high wages. The median is $133,080 and the mean is $144,570 — a difference of $11,490. This is smaller than you might expect for a field known for enormous compensation at top-tier firms. The reason is that software development has a genuinely broad middle class. Yes, FAANG engineers at the top earn $250,000 or more, pulling the mean upward. But the vast mass of developers at mid-size companies, agencies, and non-tech enterprises earn between $100,000 and $170,000 — close enough to the median that they don't create extreme skewness. Software development is an unusually well-paid field with less skewness than you'd expect — a career where the typical experience is not far from the reported average.