Ethics of organizational development — Organizational Development

OD is unusual among management disciplines in having an explicit value base. The founding tradition holds that organizations should be humane as well as effective, that people affected by a decision should have a voice in it, and that the practitioner's obligation runs to the whole system, not only to the person paying.

This creates real conflicts. The chapter exists to make them visible rather than to resolve them cleanly.

Evidence: Normative, not empirical. These are value commitments, and they should be argued rather than measured. Two of the specific claims here do have empirical support: surveying without acting lowers trust below the pre-survey baseline, and demographic cuts on small groups re-identify respondents. The rest is professional ethics.

Use when: at entry to every engagement, before collecting any data, and again whenever a finding implicates the person paying for the work. Also whenever you are asked to survey, to map a network, or to report anything at individual level.

Do not use when: never. These are standing obligations rather than a technique to be selected. The one thing to avoid is treating this file as a checklist to be completed once and filed; the conflicts it describes recur throughout an engagement.

Core values of the tradition

Humanism: people are ends, not merely means to productivity.
Participation: those affected help shape the change.
Transparency: data is fed back to those it came from.
Client self-sufficiency: the goal is a system that no longer needs you.
Whole-system view: an improvement that exports harm to another unit is not an improvement.

The standing conflicts

Who is the client? The sponsor pays; the workforce is affected. When findings implicate the sponsor, the contract you made at entry is the only thing protecting the work. See entry-and-contracting.
Consent under power asymmetry. An employee invited to a "voluntary" session by their own director has not freely consented. Design for genuine opt-out, and never report participation lists.
Surveys as extraction. Asking people about problems you have no intent or authority to fix is harmful. It reliably lowers trust below the pre-survey baseline. Do not run a survey you cannot act on.
Confidentiality vs. duty to disclose. Some findings cannot be held in confidence: patient safety risk, harassment, fraud, regulatory breach. State these exceptions before every interview, not after one occurs.
Anonymity that is not anonymous. Demographic cuts on small teams re-identify people. Set a minimum reporting threshold, commonly five to ten responses, and hold it even when an executive pushes.
OD as a euphemism. Being asked to run "culture work" that is preparation for redundancies, or "engagement" that is a legal defense, is common. You may choose to take such work; you may not pretend it is something else to the participants.
Manufactured consent. Consultation theater — running sessions after the decision is made — does more damage than not consulting, because it teaches people that participation is fake.

Data handling

Collect the minimum necessary.
Separate identifiers from responses as early as possible.
State retention periods and delete on schedule.
Know the regime that applies: GDPR and UK GDPR for personal data, HIPAA for US protected health information, plus local works council and union agreements which may require consultation before a survey runs.
Never feed employee free-text comments into an external tool without checking the data-processing terms.

Examples by audience

Health care: a staff survey that surfaces a named consultant's behavior cannot be aggregated away. There is a patient-safety duty. Plan the route for such disclosures before you collect anything.
Software: developer productivity telemetry becomes surveillance the moment it is reported per-person. Report at team level and above; say so publicly. See dora-metrics, space-framework.
Executive: the ethical test for a leadership team is whether it will hear a diagnosis that implicates itself. Contract for that outcome explicitly, and decide in advance what happens if the answer is no.

A practical test

Before acting, ask: would I be comfortable if every participant read the full contract, the raw method, and the report to the sponsor? If not, something needs to change — usually the contract.

Core values of the tradition

The standing conflicts

Data handling

Examples by audience

A practical test

See also