Summary

In response to a congressional request, GAO examined some of the characteristics of 65 demonstrations funded by the Department of Housing and Urban Development (HUD) that were conducted between 1974 and 1981. GAO studied their nature and extent and examined their design and evaluative aspects. GAO also performed case studies of six HUD-funded demonstrations to determine whether they had been selected, designed, and evaluated to achieve their general policy purposes.

A GAO profile of HUD demonstrations showed that 60 of the 65 programs included among their policy purposes either the production of new information to support program development or the promotion of a new approach to a program already developed. Other policy purposes included improving a process and providing training or services. Four of the case-study demonstrations provided reasonably clear results, although they were limited in various ways by design decisions. Policy purposes and program goals were reasonably clear and agreed on in all six cases. However, they differed in the degree to which their design and evaluation could provide information necessary to determine whether their policy purposes and goals were achieved. GAO listed several critical factors which should be met before initiating a demonstration and stated that successful implementation of a demonstration requires careful organization and administration and appropriate and aggressive encouragement of participation. GAO believes that data from both successful and failed demonstrations should be made available. Finally, assessing how users react to and use the results will help improve the demonstration as a policy tool.