November 9, 2007

Dear ANES Users,

We write to tell you about some important changes to our Restricted Data policy. Recent improvements in technology allow scholars to conduct a far greater range of data analyses than ever before. Instead of trudging through one dataset at a time, it is now possible to draw important inferences from complex analyses of multiple datasets. With these new analytic opportunities also come new challenges. A key challenge for us is one of respondent privacy.

Every time we conduct a survey, we interview and collect data on hundreds or thousands of people. We collect information that allows us to contact them (such as their address and telephone number) as well as information that helps us characterize them demographically. When we collect this information, we make a promise to safeguard their personal information. The integrity of the ANES depends on our ability to recruit respondents from all walks of life. To do this, we must maintain the public's trust. Therefore, keeping our promises and protecting respondent privacy is a critical priority for us.

One of the most important and effective ways in which we protect respondent privacy is by not including certain facts about them in the datasets that we provide to the public. In particular, we do not distribute information that someone else can use to identify respondents. We will never distribute some information (e.g., respondent names and addresses). Other information, such as respondents' verbatim responses to open-ended questions, their specific occupations, and some geographic information is provided on a very limited basis to researchers under our Restricted Data policy.

The opportunities and challenges of new technologies are prompting Institutional Review Boards (a.k.a., human subjects protection committees) and university lawyers across the country to make more thorough and precise demands on researchers. They are also inducing projects such as the ANES to update their Restricted Data policies to reflect the growing risks to respondent privacy and associated changes in legal liability. We are updating our procedures accordingly. In making these changes, we have reflected on our past experiences with Restricted Data requests and we have researched how a wide variety of comparable projects are dealing with the new realities of privacy protection. From this information, we have developed a set of practices that protect respondents while permitting innovative research.

As before, there will be fees associated with making a Restricted Data request. An application fee covers the staff time and materials required to manage and process Restricted Data requests. We have conducted an internal audit to determine the average cost to ANES of administrative work that must be done by project staff to process a Restricted Data request. The tasks involved include reviewing an application for completeness and accuracy; corresponding with requestors about application materials; preparing materials for review by the Board of Overseers; managing questions and discussions by such persons; acquiring and processing IRB documents from the requestors' universities; processing and registering agreements pertinent to the Restricted Data request; corresponding with relevant persons and entities about such documents including seeking approval from university lawyers, shipping the requested data; and following up with the requestor biannually. Materials required include folders to organize the process, paper to print correspondence and agreements, electronic storage media (CDs) to hold the datasets, and shipping charges for secure transfer of the files.

Our estimate of the average cost of these activities per Restricted Data request is $335. Since these efforts are made on behalf of individual investigators, and hence do not directly benefit the user community at large, they are not -- and have not been -- charged to ANES's NSF grant. These costs accrue to the requestor. Effective November 9, 2007, the administrative fee for a Restricted Data request will be $335. This amount does not include overhead and is less than the amount charged by most other comparable projects for which we have data.

As in the past, depending on the request, other fees may also apply. Requests for types of data that are readily available are offered at no additional charge. When a request requires extensive staff time for processing, our staff spends time working on compiling a Restricted Data product for a single researcher - making them unavailable for service to the broader user community. Following past practice, costs for this time are charged to the requestor and not to the NSF grants that support the main project.

The other major change to our procedures is that Restricted Data requestors must provide their home university's IRB approval for the research described. This requirement is already standard practice for researchers at many research universities. Our incorporation of this requirement also harmonizes our requirements with that of comparable studies.

In the upcoming years, ANES will continue to collect data in dynamic new ways. The range and amount of data that these efforts will provide are a strong new source from which to evaluate deep and important hypotheses about choice in elections. We look forward to working with you to adapt to the new challenges that we face and to make the most of the new opportunities before us.

Sincerely,

Jon A. Krosnick and Arthur Lupia
Principal Investigators
American National Election Studies