Mean salary
Also called average salary · average base salary · mean base
Sum of reported bases ÷ number of graduates who reported a usable base. Base excludes bonuses, commissions, benefits, and perquisites, even if they look guaranteed. Hourly or contract pay is annualized with a typical workweek for the country of work. $0 (100% commission) salaries are not averaged. Unknown salaries are skipped. Company-sponsored and already-employed graduates are out of Table 4.A.
Mean salary on a CSEA full-time MBA report is the arithmetic average of reported base salaries for seekers who accepted new jobs by three months. It moves when a tail hire lands. The median often does not. Edition VII requires both numbers, plus low and high, whenever a cell clears the confidentiality floor.
How is the average built?
Sum of reported bases ÷ number of graduates who reported a usable base. Base excludes bonuses, commissions, benefits, and perquisites, even if they look guaranteed. Hourly or contract pay is annualized with a typical workweek for the country of work. $0 (100% commission) salaries are not averaged. Unknown salaries are skipped. Company-sponsored and already-employed graduates are out of Table 4.A.
Currency: the school may keep local currency or convert to US dollars with Financial Times rates on or near the collection cutoff. The school converts, not the student. Purchasing-power parity tables are optional extras and must not be the version sent to media.
Haas 2025, permanent work authorization: mean $177,988, median $190,000. That $12k gap is the left tail — public sector, early-stage, and smaller markets — not a data bug. Yale 2025’s $175k median with a $140k 25th percentile is the same shape from the other side.
When does the mean mislead?
| Distortion | What happened |
|---|---|
| Mean << median | A few low bases (nonprofit, founder-adjacent salary) |
| Mean >> median | A few PE / hedge-fund bases |
| Mean on n=4 | One person is 25% of the average |
| Mean without a response-rate footnote | Volunteer sample, often high-comp heavy |
Example: 20 reported bases, 18 of them $175–192k, two of them $90k. Mean ≈ $168k. Median $185k. A recruiter who quotes only the mean is describing a class that almost nobody in consulting actually signed.
The I/D rule still binds. Fewer than three reported salaries, or fewer than 1% of seekers, and the cell is I/D / N/A / Other. All those salaries still roll into the class-level mean.
Common errors
Calling mean “the average package.” Folding signing bonuses into the average. Including returning sponsor pay. Averaging $0 commission jobs. Comparing a USD mean to an unconverted rupee mean. Publishing a function mean that fails the three-or-1% test.
FAQ
Why is mean usually below median in MBA reports?
Consulting and investment banking cluster near the top of the list. A smaller set of corporate, nonprofit, and start-up bases sits lower. The middle of the sorted list is the cluster. The average walks toward the tail. That is a mix story, not a “the school is hiding something” story, unless the response rate is weak.
Do $0 salaries pull the mean down?
Edition VII says $0 (commission-only) salaries should not be calculated into average base. Blank / unknown salaries are also out. A 12twenty help article notes that some software includes zeros unless the office excludes them. Check the footnote, not the vendor default.
Can I add the mean signing bonus to the mean base?
Not as a CSEA “total compensation” figure. The required tables keep them on separate lines because one is a salary and one is a one-time check, and because different people report each. You can do the arithmetic on a napkin if you name the two response rates.
Is a regional mean comparable across schools?
Only if both used the same currency rule and the same three-month new-employment cohort. A PPP exhibit is a school-level extra. CSEA tells schools not to send PPP tables to rankings or media as the official file.
Related terms
Sources
- CSEA Standards for Reporting Full-time MBA Employment Statistics, Edition VII Table 4.A mean of base; $0 commission out; annualize hourly; exclude sponsors; I/D floor
- Berkeley Haas Full-time MBA Employment Report 2025 Permanent-auth mean $177,988 vs median $190,000 — left tail visible in the gap
