By themselves, subsections 3(e)(1) and 3(e)(2) do not appear to have changed agency practice markedly. Only when coupled with subsection 3(e)(3), the so-called "Privacy Act Statement" requirement, have they had a modest impact on agency information gathering.
In their 1975 annual reports, OMB specifically asked the agencies to evaluate the impact of subsection 3(e)(2) on the quality and quantity of information they gather directly from individuals. Of the agencies that commented, only the Department of Labor reported that responses to its survey questionnaires had improved2 but no agency reported that there had been a decline in the quality or quantity of information it was able to gather from individuals directly. This experience was reaffirmed in the 1976 annual reports of 43 agencies. Moreover, only the Department of Housing and Urban Development and a portion of the Department of Health, Education, and Welfare reported any difficulty in collecting information from individuals. The difficulty occurred only in surveys, and in the DHEW case was of minor consequence.3
During the Privacy Protection Study Commission's October 1976 staff workshop on research and statistics, the Act's effect on survey research was explored in some detail. The Department of Defense Manpower Data Center and the Agriculture Department reported that the Act had had no effect on response rates in their surveys. Veterans Administration participants, however, said that before the Privacy Act took effect, the VA's survey response rate had been 75-80 percent, but when Privacy Act Statements were incorporated into its questionnaires, the rate fell to around 60 percent .4 The Bureau of the Census also reported that since the Privacy Act took effect its participant refusal rate had increased 50 percent (from l.5 to 2.2 percent), but the Bureau's response rate is still about 97 percent. Census attributed the change to interviewers making less vigorous efforts to persuade hesitant individuals to participate.5