Prepared by:
Rocky Mountain Social Science
P.O. Box 3462
Logan, Utah 84323
Prepared for:
Boyd County Monitoring Committee
P.O. Box 376
Butte, Nebraska 68722
July, 1992
Data used in developing the analysis of sociocultural conditions and potential social impacts of the proposed LLRW facility were collected over a seven-month period, beginning in November 1991. Because the use of multiple data collection strategies can substantially improve the scientific validity of research (see Webb et al., 1966), several complementary research procedures and approaches were utilized.
Field Data and Other Information Sources
To assure that the analysis was based on a well-established, first-hand understanding of local conditions and responses to the LLNW siting proposal, the initial phases of the research relied heavily on field research involving direct observation of local conditions and social processes as well as face-to-face interviews with a broad range of local area residents and officials. The two RMSS senior scientists traveled to Boyd County in December 1991 and again in February 1992 to conduct the field research. During these visits the investigators attended several public meetings pertaining to the LLNW siting issue. They also met with representatives of several local organizations to explain the purpose of the study and to solicit information about local conditions and perspectives regarding the proposed facility.
Key informant interviews. In addition, the investigators conducted a series of "key informant" interviews with a wide variety of local residents and officials. Starting with easily-
14
identifiable persons occupying organizational or governmental leadership positions, potential informants were identified through a "snowball sampling" procedure (see Babbie, 1975). Those initially contacted were asked to identify other persons who might be knowledgeable about community conditions or who might have important insights into local responses to the proposed LLNW facility. A total of over 50 individuals were interviewed during the two field research periods. The interviews were guided by a research protocol which specified topics important for discussion. However, the protocol specified neither the exact wording of questions nor the sequence of questioning. Rather, the flow and direction of the interviews were determined, in large part, by the informants
Those contacted included local business owners, operators and employees; elected and appointed community and county officials; religious leaders; public safety organization personnel; school system administrators and teachers; representatives and members of local organizations which both support and oppose the proposed LLNW facility;10 individual farm owners and operators; and a broad range of other county residents. Although the process for identifying and selecting such informants does not guarantee a statistically representative sample of the affected population, those interviewed did comprise a cross-section of the county population with respect to geographic location, occupational background, age, and significant numbers of supporters of both PFP and SBC were interviewed.
15
viewpoints regarding the proposed LLNW facility.
It should be noted that the greatest proportion of interview respondents were selected from the geographic areas closest to the proposed site. Thus, residents of Bristow, Gross, Lynch and Monowi were statistically underrepresented, while residents of Butte, Naper and Spencer were statistically overrepresented. The decision to focus the key informant interview efforts most heavily on those residents living closest to the proposed site was made for practical reasons rather than theoretical or political reasons. Funding restrictions made a full ethnographic description of the entire county impossible, so study efforts concentrated on areas nearest the proposed site. It was assumed that residents living near the proposed site would likely be more effected by the project and would therefore be more involved in and aware of the issues than those living further from the site. That is, self interest would motivate them to pay closer attention to the issues, and they could therefore provide more information per interview than those Boyd County residents dwelling further from the site.
Available data. Several sources of available data supplied information on the social and economic context of Boyd County and the communities within the county. These included various U.S. Census Bureau reports on population characteristics, employment, economic activity, and agricultural operations. Selected materials prepared by US Ecology and its contractors provided information regarding the characteristics of the proposed LLNW facility and potential project impacts. Several locally-assembled volumes
16
furnished information on the history of Boyd County and the towns of Spencer and Butte. Other historical materials dealt with early settlement and railroad development activities. Typed transcriptions and tape recordings of several earlier public meetings and hearings regarding the LLNW siting proposal were reviewed. Finally, selected articles from both the local weekly newspaper (Spencer Advocate) and several regional newspapers (particularly the Lincoln Star and the Omaha World-Herald) yielded additional background information regarding the project and the responses of various groups and organizations to it.
Survey Procedures
An additional major data collection activity involved the use of standardized self-completion survey questionnaires, which were administered to a representative sample of 300 adult Boyd County residents. The survey addressed a number of major issues, including respondents' sociodemographic characteristics, community satisfaction and value orientations, general environmental attitudes, and attitudes and perceptions regarding low-level nuclear waste storage issues (see Appendix A).
To ensure that a statistically representative sample of Boyd County residents was contacted, the selection of potential survey respondents required a multi-stage sampling methodology (cf. Mueller, Schuessler and Costner, 1977). A sampling frame which listed all housing units in the county was assembled. Part of the sampling frame was constructed from county maps of Boyd County's
17
eight incorporated villages.* A fieldworker then verified all of the houses and their locations in the villages with on-site visual enumeration. Those areas not represented on existing maps were also mapped by the fieldworker. The remaining part of the sampling frame included those houses not located within the villages, the locations of which were taken from the Farm and Home Directory (Farm and Home Publishers, Ltd.,1990)**
Each housing unit was assigned a number, and simple random sampling procedures were used to select a representative sample of households from throughout the county. The sampling procedure included selection of a "primary91 sample comprised of 300 housing units. In addition, the same random selection procedures were used to select a "secondary" or "replacement" sample of 150 additional housing units to provide for substitution when units within the primary sample were found to be unoccupied or abandoned, nonexistent, or nonresidential.
The second stage of the sample selection procedure occurred at the point of survey implementation and involved the selection of one adult from each sampled household for participation in the survey. Within each sampled household the person 18 years of age or older whose birthday had occurred most recently was selected to
*Anoka, Bristow, Butte, Gross, Lynch, Monowi, Naper and Spencer
**There were some overlaps between village maps and the Farm and Home Directory. inspection by the fieldworker eliminated any house listed twice. It should be noted that while some of houses from the non-village areas were vacant, that fact in no way detracts from the random qualities of the sample.
18
complete the questionnaire. This method results in an essentially random selection of adult household members without the complexities or intrusiveness of some more traditional respondent selection methods requiring detailed information on household composition prior to respondent selection.
Survey implementation commenced on April 6, 1992, with delivery and retrieval of questionnaires continuing through April 21. Questionnaires were personally delivered to members of the sample by two trained fieldworkers who gained the cooperation of the selected person before leaving the questionnaire. The fieldworkers returned to retrieve the completed questionnaires within a day or two after they were delivered. In order to obtain the highest degree of cooperation, an article describing the project and a photograph of the fieldworkers were published in the county newspaper immediately before the start of the survey and during the first week of survey administration. In addition, each fieldworker possessed letters from the leadership of both PFP and SBC supporting the research endeavor and encouraging their supporters to cooperate in filling out the questionnaires.
Of the 300 questionnaires that were delivered, two were returned only partially completed, and 36 individuals refused to participate in the survey. Usable responses were obtained from 262 individuals, representing an overall response rate of 87.33 percent. This unusually high level of cooperation for a selfcompleted survey (see Dillman, 1976) is indicative not only of effective survey implementation but also of the very high level of
19
local interest in the nuclear waste storage issue.
Data Limitations
The data generated by this study are vastly superior to information derived from public comments at meetings or through various other "public involvement" procedures associated with the site licensing process for the proposed LLNW facility. Data obtained through public hearings or comments do not allow for the level of analytic detail which is necessary for conducting a competent impact analysis, nor do they provide information which can be generalized to the affected population with any acceptable degree of confidence (see Heberlein, 1976). In contrast, the data generated by this study provide for both analytic detail and statistical inference to the county population, the latter guaranteed by the random nature of the sampling.
However, even with these overwhelming advantages there remain important data limitations which constrain the types of conclusions that can be drawn from this research. Most importantly, the researchers did not; have access to true baseline data on local social and cultural conditions prior to time when the LLNW facility became a public issue. Instead, the research was conducted more than three years after the possibility of siting the facility in Boyd County was first announced in 1988, the time when an ideal SIA should have been initiated.
During the intervening period between announcement of the proposal to site a radioactive waste facility in Boyd County and the time of the research, major shifts in the local social context
20
occurred. During that period there emerged new voluntary organizations, shifts in linkages between existing organizations and communities, altered social ties among area residents, new patterns of social and political involvement and participation, and changed attitudes and perceptions regarding waste storage, political processes, and a host of other related issues. In short, social structures in Boyd County and the local communities within it have undergone major changes since 1988. Although the research reported here can document many changes which have taken place, the absence of a true pre-project baseline limits the potential for establishing the magnitude of such changes or the relationships between apparent change, and the proposed LLNW facility.
Similarly, efforts to predict future impacts of the project are constrained by the need to rely on data collected only during a relatively brief research period prior to actual licensing, construction, or facility operation. Under the best circumstances, multiple over-time measures would have been made, allowing for more valid predictions of future behavior. Synchronic data are simply inadequate for trend analysis. Although such limitations are not uncommon in the impact assessment process, they are compounded in this case by the limited information currently available regarding some aspects of the proposed facility and its operation (e.g., the actual quantity and types of waste that would be accepted at the facility, waste transportation routes, transportation schedules, housing plans for construction and operations workers, etc.). Also, scientific estimates, e.g., probabilities, of the potential
21
for accidents or waste leakage from the facility are absent.
Despite these limitations, the research procedures outlined above provide a substantial data base that resolves many of the limitations resulting from less systematic approaches. The remainder of this report draws upon that data base to document social conditions and changes, local responses to the LLNW facility siting process, and potential impacts of constructing and operating that facility in Boyd County.
22