crime control programs,
Smallest Geographic Unit:
- 2004--2005 (fall 2004 and fall 2005)
Date of Collection:
- 2004--2005 (fall 2004 and fall 2005)
Unit of Observation:
All farms in nine California counties during 2004 and 2005.
Data Collection Notes:
The Agricultural Census and Census Bureau data collected during this project, as well as data from the site visits and interviews, are not available as part of this collection.
Agricultural crime is a serious problem in the United States, yet few prevention or reduction interventions have been developed or evaluated to combat this problem. Accordingly, the Urban Institute and Florida State University multidisciplinary research team employed a multimethod approach to evaluate the Agricultural Crime, Technology, Information, and Operations Network (ACTION) project, located along the sourthern coast and central valley of California and supported by the Bureau of Justice Assistance. The goal of the research was to provide policymakers, practitioners, program developers, and funders with empirically-based information about whether ACTION works. The objectives included assessing the effectiveness of the ACTION project in reducing agricultural crime, examining the costs and benefits and cost-benefit analysis (CBA) issues related to evaluating the program, and identifying how programs like ACTION can be effectively adopted by other jurisdictions.
Two paper-and-pencil, self-administered surveys -- one in fall 2004 and the second in fall 2005 -- were sent to samples of farmers in the nine Agricultural Crime, Technology, Information, and Operations Network (ACTION) counties in California. The researchers identified farms using lists provided by Agricultural Commissioners in each county. Specifically, an ACTION staff member contacted each county's Agricultural Commissioner office and obtained their most accurate pesticide permit holder lists in June and July of 2004. This list was used for the 2005 survey as well. Before sampling for the second wave of survey mailings in 2005, the researchers confirmed with Agricultural Commissioners' offices that their lists of pesticide permit holders do not change more than 10 percent from year to year as permits are valid for three years at a time.
The lists provided relatively complete enumerations of farm operations that use pesticides -- which typically includes all but the smallest operations and livestock operations that do not grow their own feed -- because these farms must register with the Agricultural Commissioners.
Conversations with Commissioners' offices and representatives from the National Agriculture Statistics Service indicated that the sampling approach was likely to capture the majority of pesticide users, particularly medium and large operations. Nonetheless, any results using these data may not fully generalize to all farms, such as those focused primarily on livestock or that do not use pesticides, or to farms beyond the study region. The survey instruments, developed with the assistance of experts who consulted on the project and with farmers, and drawing on instruments used in previous studies (e.g., Bean and Lawrence 1978; Farmer and Voth 1989; Cleland 1990), were administered in the fall (2004 and 2005, respectively) and asked farmers about experiences with agricultural crime victimization during the 12 months prior to the survey. It also asked questions about characteristics of their farm operations and the activities that they take to prevent agricultural crime.
Advance notice of the study was given to farmers through the use of postcards, then surveys were sent to farmers in three waves at one-month intervals, with the second and third waves targeting nonrespondents. Cover letters were included, as was a letter of support from the California Farm Bureau Federation and a self-addressed, stamped return envelope. For the first survey, concern about the relatively low response rate prompted the researchers to reconsider incentive options for the follow-up mailing. The research team decided to provide a one or two dollar incentive in each survey mailed in the third follow-up. Because prior research provided no basis for determining whether two dollars would garner more responses than would one dollar, half the final follow-up received two dollar
incentives and the other half received one dollar incentives. Subsequently, the researchers found little
difference in the response rates of farmers who received one dollar versus two dollars. However, the incentives did appear to boost the response rate. Given the improvement gained from using the incentives, and the lack of difference in using one dollar versus two dollars, the second-year survey included one dollar incentives in all three waves of mailings. In addition, a raffle incentive was introduced in the final follow-up; respondents who returned the survey within three weeks were eligible to win $100. The Fall 2004 Agricultural Crime Survey (Part 1) contains data on 823 respondents (farms) and the Fall 2005 Agricultural Crime Survey (Part 2) contains data on 818 respondents (farms).
The researchers sent surveys to samples of farmers in the nine Agricultural Crime, Technology, Information, and Operations Network (ACTION) counties. For the Fall 2004 Agricultural Crime Survey (Part 1), a total of 2,286 surveys were initially mailed, but 64 were returned due to bad addresses. Of the 2,222 mailed surveys with good addresses, collectively constituting the eligible survey pool, 960 were returned. Of these, 134 were not farmers (having so indicated on the instrument) and 3 were not completed. Thus, the final sample was 823 respondents (farms). For the Fall 2005 Agricultural Crime Survey (Part 2), a total of 2,286 surveys were initially mailed, but 30 were returned due to bad addresses. Of the 2,256 mailed surveys with good addresses, collectively constituting the eligible survey pool, 1,004 were returned. Of these, 157 were not farmers (having so indicated on the instrument) and 29 were not completed. Thus, the final sample was 818 respondents (farms).
Mode of Data Collection:
Data were obtained from mail surveys.
Description of Variables:
The survey asked farmers about experiences with agricultural crime victimization during the 12 months prior to the survey. It also asked questions about characteristics of their farm operations and the activities that they take to prevent agricultural crime. More specifically, variables in the Fall 2004 Agricultural Crime Survey (Part 1) included questions about crime on the survey respondent's farm or ranch, the reporting of crime on the respondent's farm or ranch, the protection of farm or ranch assets, law enforcement in the respondent's county, and general characteristics of the respondent's farm or ranch operation. Similarly, the Fall 2005 Agricultural Crime Survey (Part 2) included variables relating to crime on the respondent's farm and in his or her county, law enforcement in the respondent's county, information about agriculture crime, safety and security on the respondent's property, as well as information about the respondent and his or her farm or ranch operation.
Of the 2,222 mailed surveys with good addresses for the Fall 2004 Agricultural Crime Survey (Part 1), 960 were returned which resulted in a response rate of 43.2 percent. Of the 2,256 mailed surveys with good addresses for the Fall 2005 Agricultural Crime Survey (Part 2), 1,004 were returned which resulted in a response rate of 44.5 percent.
Presence of Common Scales:
Several Likert-type scales were used.
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of
disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major
statistical software formats as well as standard codebooks to accompany the data. In addition to
these procedures, ICPSR performed the following processing steps for this data collection:
Created variable labels and/or value labels.
Standardized missing values.
Checked for undocumented or out-of-range codes.