CSU Says Faculty Pay Is Competitive. The Public Still Cannot Audit the Model.

The new Segal report offers useful findings. Its missing work papers create the larger transparency question.

By Thomas J. Norman, Ph.D., Professor of Management, California State University, Dominguez Hills. Views are the author's own.

The California State University has published a compensation study that reaches a reassuring conclusion: most represented faculty and staff compensation is close to market or better once benefits are included.

That conclusion may be correct. The public still cannot fully audit how CSU reached it.

The August 2026 Segal report covers 29,116 represented instructional faculty and 16,518 represented staff. CSU reports that average base salaries are approximately 1 percent below the market median for both groups. Once benefits are included, total compensation rises to approximately 3 percent above market for faculty and 4 percent above market for staff.

The report also says 86 percent of faculty and 84 percent of staff are within or above a competitive range defined as 85 percent to 115 percent of market median. Including benefits, those figures rise to 95 percent and 97 percent. Seventy percent of faculty and 63 percent of staff are above the market median on total compensation.

Those are important findings. They should not be reduced to a single headline.

The report shows meaningful variation beneath the systemwide average. Staff earning less than $80,000 are less competitive on base salary. Sixty-six percent of employees below that threshold are below the market median, although only 17 percent fall below the competitive range.

Faculty results also vary by discipline. Business faculty are reported at 16 percent below market on base salary and 10 percent below market on total compensation. Engineering faculty are 15 percent below market on base salary and 9 percent below market on total compensation. Computer and information sciences are 13 percent below market on base salary and 7 percent below market on total compensation.

Other fields show the opposite pattern. English faculty are reported at 14 percent above market on base salary and 15 percent above market on total compensation. Foreign-language faculty are 14 percent above market on both measures.

Campus differences matter as well. The report identifies San Jose, East Bay and San Francisco as the least competitive campuses for instructional faculty. Dominguez Hills is reported at approximately 3 percent above market on total compensation. Systemwide averages can therefore conceal the experience of particular disciplines, campuses and salary levels.

The report contains another important qualification. It says the results are not fully comparable with earlier Mercer studies and should be treated as directionally informative. The Segal study includes approximately 75 percent of staff classifications, while the earlier Mercer study included approximately 50 percent. Segal also aged market data to July 1, 2026, using a 3.5 percent annual update factor. Staff comparisons rely primarily on higher-education data, while general-industry comparisons produce a different result.

These are not minor technical details. They are the model.

In an August 11 California Public Records Act request, I asked CSU for the underlying Segal deliverables, including appendices, comparator data, job matches, methodology, contract records, communications about market adjustments and the editable compensation model. CSU's September response stated that Segal retained proprietary records and that CSU had no further records available beyond the materials already shared.

That response leaves several questions unanswered.

Did CSU receive the final deliverables required by its contract? Did the consultant provide an editable model, underlying data files and documentation of assumptions? Which records were withheld because of a specific exemption? Which records do not exist? Which records are outside CSU's possession? Did CSU review segregable portions that could be released without disclosing proprietary information?

A consultant's proprietary claim does not answer those questions. A public agency may need to protect trade secrets, personal information or confidential material. It should still identify what it received, what it withheld and why.

California's public-records guidance recognizes a public right to inspect records maintained by government agencies. The state Department of Justice also explains that agencies should provide a written determination and reasons when records are withheld. Government Code section 7922.540 requires a written denial to identify the responsible officials and justify the withholding.

The transparency issue is therefore not whether every consultant worksheet must be posted online. The issue is whether the public can follow the audit trail from contract, to deliverable, to methodology, to conclusion.

At minimum, CSU should publish:

1. The Segal contract, amendments, scope of work and invoices.
2. A list of deliverables CSU received.
3. The comparator institutions and job-matching rules.
4. The geographic, aging and benefits assumptions.
5. A record-by-record disposition of the outstanding public-records request.
6. Any nonproprietary portion of the model or a usable explanation of how the calculations were performed.

The Board of Trustees should want this information for its own oversight. Faculty and staff should want it because compensation decisions affect recruitment, retention and morale. Students should want it because compensation is one of the system's largest recurring costs.

The CSU has given the public a conclusion. It should now provide the evidence that permits independent review.

A compensation study should not require public trust as a substitute for a reproducible method.

Disclosure. I teach compensation, am developing a compensation textbook, and serve on the executive board of the Dominguez Hills chapter of the California Faculty Association. I write here as an individual faculty member, not for the CSU, my campus, or the union.