Data Confidentiality
Introduction
Fisheries data from commercial, recreational and subsistence fishing operations should be presented for scientific and management purposes in a manner which balances the rights and operational confidences of our sources of information with our scientific and management obligations to make information accessible.
We have a responsibility to our data sources such as the fishing vessel captains and owners to be prudent in the use of such data. From a self-interest point of view, we want to insure that data are recorded accurately by the fishing community in logbooks and other data collection forms. Therefore we want to inspire their confidence by our appropriate use of confidential data.
Of course, confidential data may be used by researchers for analytical purposes subject to abiding by the non-disclosure standards. The real issue is in the publication of information or information summaries which might be viewed as revealing the operational practices or other sensitive information about individuals or companies.
Where the following standards cannot be met while meeting scientific and management requirements, the Laboratory director may make an exception. The following are standards for public dissemination of data and data summaries in this web site as well as in scientific papers and fishery management reports published by the PIFSC. The standards apply to data collected through the PIFSC. Standards for data from other sources depend on the conditions placed upon receipt of such data.
Vessel Identification
Individual vessel identifiers cannot be attached to any individual data items which are made public. Vessel identifiers include vessel name or permit number and individual data items include fishing location, catch, and effort.
Vessels can be differentiated by random identifiers for some purposes, as long as such an identifier did not reveal actual identities.
Any fishery-wide aggregations of data shall include information from three or more individual vessels. Effort information, including just the presence of fishing, can be just as sensitive as the actual catch itself. All data analysis programs should include a procedure for calculating the number of vessels within the aggregate. Wherever possible, aggregations should be large enough to include more, rather than fewer, vessels.
If fisheries operations data are aggregated into classes of vessels such as large, medium and small-sized vessels or types of operations such as vessels targeting tuna, swordfish, or mixed species, then there should be three or more individual vessels in each aggregate. Wherever possible, classes should be large enough to include more, rather than fewer, vessels.
In many cases, confidentiality can be assured by increasing the time or area range of strata such as using multiple months, multiple years, or multiple locations.
Fishing Location
Fishing location information is most sensitive because it is essentially a "trade secret" for the vessels involved.
Latitude and longitude information should be reported or plotted primarily on 5 degree squares, which is the de facto international standard for fisheries data exchange.
However, as long as there are at least three vessels included in the overall scope of the data, then data may be reported or plotted for fewer than 3 vessels per 5 degree square. This would not be the case for smaller area strata, such as for 1 degree squares.
Where data are reported for a standard fishing area, such as the Northwestern Hawaiian Islands fishing bank, the same aggregation rules apply but the 5 degree rule need not apply.
Where finer displays of data are required, such as in examining the 50-25 mile longline area closures, smaller area aggregations may be used. In these cases, alternative ways of displaying data may be required such as contours or bands of fishing activity rather than lat/long squares or individual data points. Care should be taken on the temporal scale. Monthly plots of 1 degree square information are much more sensitive than annual plots.
Other Considerations
Where a fishery has expanded over the years to cover previously sensitive time-area strata, then those strata should be amenable to reporting.
Finally, presenters of information based on confidential data should put themselves in the shoes of the suppliers of such data and consider how they would view the use of their information.