Placing a Value on Public Library Services|
Glen E. Holt, Donald Elliott, and Amonia Moore
REFINEMENT OF THE CONSUMER SURPLUS ESTIMATES -- RESURVEY
If survey results had been ideal, the values derived from each of the four measurements using the three methodologies would have produced figures that were relatively close to each other. Such a clustering of estimates would have made the researchers more comfortable with the validity of their measurements. The initial survey did not produce this clustering of values. The researchers, therefore, decided to undertake a relatively inexpensive resurvey of those general users who had participated in the previous survey.
The research team was especially concerned about the very high values revealed in the consumer surplus methodology. The consumer surplus estimate should have been less than the WTA estimate. The SLPL team believed that, in spite of considerable effort, the original consumer-surplus questions had not been sufficiently clear.
To put the matter simply, research team members suspected that survey respondents had overstated the amount of services they would have purchased if the library did not exist. In analyzing the figures, it appeared that many of the respondents had included not only purchases to replace library services but also purchases they would have made in addition to library services. Straightening out this methodological problem was necessary before the team moved forward with the next planned phase of the services-valuation project.
To undertake the resurvey, Dr. Holt wrote a letter to all 322 completed-survey respondents. In the letter he asked if the respondents would accept a brief second call to help clarify a few answers they had given previously. In the letter SLPL enclosed a leather bookmark in appreciation in advance for their participation. Of the original 322 respondents, 235 participated in the resurvey.
There were few refusals, but participation dropped because of moving, unavailability during the time window of the resurvey or changes in phone listings that made it impossible for interviewers to reach them. These shifts tended to match the normal process of a high-mobility population with a high incidence of lapsed cards characteristic of the SLPL system through the previous decade.
As was the case with the previous full-scale survey, the sample was corrected for a low-response bias among low-income and African-American households. The sample responses were weighted to correct for the response rates, but again, reassuringly, the valuation estimates varied little from the unweighted results.
The resurvey results substantially corrected the inflated consumer-surplus estimate in the original survey. The resurvey allowed the elimination from the consumer-surplus estimates of respondents who used particular library services but would not have replaced them through market purchases. This result reflects the perceptions of public library professionals who have been told over and over by constituents that if the library did not exist, they would not have access to these services. The resurvey documented this observation statistically.
The resurvey also allowed a number of respondents to revise their estimates of library use. In most cases, these were downward revisions. Clarification of purchases to include only replacement purchases and exclude purchases complementary to library use also lowered estimates.