All Employees Are Prohibited From Engaging in Discrimination and Blank Including Against Members of the Public When Providing Governmental Services or Administering a Government Program Unless Federal or State Law Mandates Otherwise

All employees are prohibited from engaging in discrimination and harassment, including against members of the public when providing governmental services or administering a government program, unless federal or state law mandates otherwise. The blank is "harassment," and the paired ban matters because discrimination and harassment are distinct legal theories that reach different conduct, trigger different proof standards, and demand different employer responses. This explainer unpacks how the two prohibitions interact inside public-agency operations, where the employee's conduct is also the government's conduct.

The sentence is written for a public employer, not a private one. It covers the clerk at the counter, the inspector in the field, and the contractor's staff who administer a benefit program under color of state authority. It also signals an important asymmetry: the "unless federal or state law mandates otherwise" clause is a narrow carve-out, not a general permission slip.

Mechanism: Why the Two Prohibitions Are Distinct

Discrimination and harassment are often used interchangeably in ordinary speech, but in employment and civil-rights law they describe different conduct. Discrimination is a decision or policy that treats a protected class differently, such as denying a permit application because of religion or routing a housing complaint differently by national origin. Harassment is conduct that creates a hostile, intimidating, or offensive environment, or that is severe or pervasive enough to alter the terms of service.

The mechanisms differ in three ways that matter operationally. First, discrimination typically requires proof of a comparator: how was a similarly situated person treated. Harassment requires proof of the conduct's character and its effect on the recipient, which is why documentation of the pattern, not just the final decision, controls the outcome. Second, discrimination is often a single act with a clear decision-maker; harassment is often cumulative and diffuse, spread across coworkers, shifts, and channels.

Third, the government-services context adds a layer that private employment does not have. When an employee discriminates against a member of the public while administering a program, the agency may face both employment-type liability (as an employer of the offending worker) and program-access liability (as the entity delivering the service). That dual exposure is why the sentence pairs the two terms rather than listing only discrimination.

A plain-language version: discrimination is usually about a decision, harassment is usually about a climate, and in government services both can become the agency's problem at once.

Quantified Context: How Often These Claims Arise

The scale is larger than most thin explainers suggest. The U.S. Equal Employment Opportunity Commission receives tens of thousands of discrimination charges each year, and a substantial share include harassment as an alleged basis. Harassment charges have historically accounted for roughly a third of the agency's annual charge receipts, which is why the commission treats harassment prevention as a standing enforcement priority rather than a periodic one.

On the public-services side, federal agencies that fund state-administered programs impose civil-rights compliance conditions on recipients. A state agency that runs a federally funded benefit program is typically bound by both its own anti-discrimination policy and the funding agency's nondiscrimination rules, which can add complaint procedures and data-reporting duties that private employers never face.

The numbers that matter most for a supervisor are internal, not national. Complaint volume, time-to-resolution, and the share of complaints that allege conduct by a single repeat actor are the metrics that predict liability. An agency that tracks only the number of formal complaints will undercount harassment, because harassment is the category most likely to go unreported.

Edge Cases: Where the Rule Gets Hard

Several scenarios test the sentence's boundaries.

The first is the "unless federal or state law mandates otherwise" clause. This is not a broad exception. It covers situations where a law requires differential treatment, such as certain age-based program eligibility rules or lawful set-asides. It does not authorize an employee to treat a member of the public differently based on personal preference, and it does not insulate a practice that a court later finds unlawful.

The second is the contractor and volunteer problem. Many government programs are administered by contractors, subrecipients, and volunteers. The sentence says "all employees," but program-access rules often reach further. An agency that assumes its policy stops at payroll will miss the staff who actually interact with the public.

The third is the "members of the public" trigger. Employment policies govern how workers treat each other. This sentence extends the duty to people outside the workforce, which changes the investigation: the complainant may have no employment relationship with the agency, no access to internal grievance channels, and no knowledge of the agency's process.

The fourth is the intersection of speech and conduct. A single offensive remark may not rise to actionable harassment, but a pattern of remarks directed at service recipients can. The legal line is context-dependent, which is why the counterintuitive detail worth flagging is this: the most legally dangerous conduct is often not the loudest incident, but the quiet, repeated pattern that no one documented.

Red flag: a supervisor dismisses a complaint because "the customer was difficult." Red flag: a pattern of complaints about one employee is treated as separate, unrelated incidents. Red flag: contractor staff are excluded from the agency's complaint-tracking system. Red flag: the "law mandates otherwise" clause is cited without identifying the specific statute. Red flag: no written record exists of what the complainant was told and when.

Misconceptions: What the Sentence Does Not Say

The most common misconception is that the sentence bans only intentional discrimination. In practice, policies and practices with a disparate impact can violate civil-rights rules even without discriminatory intent, depending on the governing statute. The sentence's wording does not resolve that question, so agencies should not read it as an intent-only standard.

A second misconception is that "harassment" means only sexual harassment. The term covers harassment based on race, color, religion, sex, national origin, age, disability, and other protected characteristics, and in the government-services context it can include conduct by one member of the public against another that the agency fails to address.

A third misconception is that the "unless law mandates otherwise" clause is a defense to any policy the agency prefers. It is a narrow carve-out tied to a specific legal requirement. If an agency cannot name the statute, regulation, or court order that requires the differential treatment, the clause does not apply.

A fourth misconception is that compliance is satisfied by posting a policy. Posting is necessary but not sufficient. The operative controls are training frequency, complaint intake, investigative timeliness, and corrective action.

Timeline: How a Complaint Should Move

The time course matters because delay is itself a liability. The table below reflects a defensible internal sequence; specific deadlines are set by the governing statute or funding agreement, not by this article.

The reconciliation point is this: the "prompt" standard in harassment guidance and the fixed deadlines in a funding agreement are different thresholds measuring different things. Promptness measures the agency's response speed; a funding deadline measures compliance paperwork. An agency can meet one and fail the other, so track both.

Practical Actions: What to Do First

Concrete steps beat general commitments. The thresholds below are operational rules of thumb, not legal mandates.

Name the specific statute when invoking the "law mandates otherwise" clause, and keep that citation in the file. Extend the policy in writing to contractors, subrecipients, and volunteers who administer programs. Train public-facing staff annually on both discrimination and harassment, with separate modules for each. Log every complaint the same day, including informal ones, and record what the complainant was told. Assign a single intake owner so patterns across shifts and offices become visible. Review corrective actions within 30 days to confirm the conduct stopped.

The most useful discipline is pattern review. Pull complaint data quarterly and ask whether one employee, one office, or one program appears repeatedly. A repeat-actor pattern is the strongest early warning an agency can have.

Limitations and Scope

This article explains the stable legal architecture behind the sentence and does not resolve any specific agency's obligations. The controlling rules vary by jurisdiction, by funding source, and by the statute under which a program operates. Federal and state requirements can differ, and a funding agency's conditions can add duties beyond either.

For any actual dispute, policy drafting, or investigation, consult a qualified local attorney or civil-rights compliance officer. This article is general educational information, not a legal opinion, and it does not promise any outcome or result.