The ability to understand the vulnerability of fishing communities is critical to understanding how regulatory change will be absorbed into multifaceted communities that exist within a larger coastal economy. Creating social indices of vulnerability for fishing communities provides a pragmatic approach toward standardizing data and analysis to assess some of the long-term effects of management actions. Over the past 3 years, social scientists working in NOAA Fisheries’ Regional Offices and Science Centers have been developing indices for evaluating aspects of fishing community vulnerability and resilience to be used in the assessment of the social impacts of proposed fishery management plans and actions. These indices are standardized across geographies, and quantify conditions which contribute to, or detract from, the ability of a community to react positively towards change.
The AFSC has developed indices for more than 300 communities in Alaska. We compiled socio-economic and fisheries data from a number of sources to conduct an analysis using the same methodology used by Colburn and Jepson (2012) and Jepson and Colburn (2013). To the extent feasible, the same sources of data are being used in order to allow comparability between regions. However, comparisons indicated that resource, structural, and infrastructural differences between Northeast and Southeast Alaska require modifications of each of the indices to make them strictly comparable. The analysis used for Alaska was modified to reflect these changes. The data are being analyzed using principal components analysis (PCA), which allows us to separate out the most important socio-economic and fisheries related factors associated with community vulnerability and resilience in Alaska within a statistical framework.
These indices are intended to improve the analytical rigor of fisheries Social Impact Assessments, through adherence to National Standard 8 of the Magnuson-Stevens Fishery Conservation and Management Reauthorization Act, and Executive Order 12898 on Environmental Justice in components of Environmental Impact Statements. Given the often short time frame in which such analyses are conducted, an advantage to the approach taken to date by the Principal Investigators is that the majority of the data used to construct these indices are readily accessible secondary data and can be compiled quickly to create measures of social vulnerability and to update community profiles.
Although the indices are useful in providing an inexpensive, quick, and reliable way of assessing potential vulnerabilities, they often lack external reliability. Establishing validity on a community level is required to ensure indices are grounded in reality and not merely products of the data used to create them. However, achieving this requires an unrealistic amount of ethnographic fieldwork once time and budget constraints are considered. To address this, a rapid and streamlined groundtruthing methodology was developed to confirm external validity from a set of 13 sample communities selected, which were based on shared characteristics and logistic feasibility. The goal of this research methodology is to confirm external validity of the well-being indices through measuring how well quantitative index constructs overlap with qualitative constructs developed from ethnographic fieldwork. Several inter-rater agreement tests, including a Cohen’s Kappa and Spearman’s rho, were used in assessing construct overlap by measuring how well ethnographic data is in agreement with the indices.
A K-means cluster analysis was used in determining community groupings based on similarities in the secondary data used in creating the indices. Once communities were grouped, 13 sample communities were selected based on the cluster characteristics and logistical constraints. An iterative, mixed-methods grounded approach was used in developing protocols for ethnographic fieldwork. Key-informant categories were identified based on the index-derived constructs, and interview protocols were developed to target specific themes thought relevant to those constructs. Interviews were open-ended to allow for emergent constructs to present themselves during the interview process. Finally, to supplement interview data, physical field assessments of community character, environment, and condition were conducted by researchers.
Once fieldwork was complete, summaries were drawn from researcher experiences and their interview interpretations, which will be used to create a qualitative ranking system. The next step for the groundtruthing exercise is to compare the qualitative fieldwork data to the quantitative indices. As a first step, a rapid assessment will be done in fall 2014. For each quantitative component, a ranking of “high,” “medium,” or “low” will be given according to the score created from the PCA. Members of the research team then will provide subjective rankings for each component based on ethnographic data, and the two ranking schemes will be tested for inter-rater agreement. Cohen’s Kappa will be used to test for perfect matches of rankings, which is the more conservative of two tests. The second test, Spearman’s rho, will provide a coefficient of “agreement,” and will not omit instances where there was not a perfect match. Together, these tests will provide a well-rounded picture of agreement between the qualitative and quantitative sets of ranks, and thus a general assessment of construct overlap. Reports documenting this phase of the project will be released in 2015.
Groundtruthing the results will facilitate use of the indices by the AFSC, NOAA’s Alaska Regional Office, and the North Pacific Fishery Management Council staff to analyze the comparative vulnerability of fishing communities across Alaska to proposed fisheries management regulations, in accordance with NS8. This research will provide policymakers with an objective and data-driven approach to support effective management of North Pacific fisheries.
By Amber Himes-Cornell and Conor Maguire and Stephen Kasperski
