Independence

Taken to signify a political stance for an evaluation that is not subject to the control of or that does not provide privileged access to any particular stakeholder group or constituency. Independence has been implied and advocated in a number of ways. For example, Scriven's goal-free evaluation seeks to uncouple evaluation from program prescriptions; Weiss' stakeholder evaluation seeks to honor the significance of all stakeholder agendas, not just those of the powerful; Stake's responsive evaluation was designed to broaden the data reach of evaluation beyond the limited concerns of the administrative system.

It is in MacDonald's democratic evaluation, however, that independence is explicitly discussed—in fact, well expressed in the maxim attributed to MacDonald: “An evaluation can be sponsored but not bought.” Independence is designed to strengthen the public credibility and undeniability of the evaluation by demonstrating that the evaluation has no preference for or obligation to any partial view and to ensure that it is a general resource for the citizenry. Independence does not imply freedom from contractual obligations or from [Page 199]the obligation to provide a service to the sponsor. It does, however, imply the need to negotiate an appropriate political role for the evaluation in which the power of the sponsor does not overwhelm the obligations of the evaluation to report to multiple audiences. A component of independence, and closely related to it, is impartiality: the formal disinterest of the evaluation in respect to program value positions and its outcomes.