Canada's economy and society are in a period of rapid and difficult change. Decision makers need timely and objective analysis of economic and social conditions to understand this transformation, to provide a basis for broad and informed debate on public policy, and to establish a foundation for intelligent policy formation. The need is particularly acute because Canada's social policy has not kept pace with the dramatic changes in its economic policy over the past two decades. Governments at all levels have acknowledged the importance of redesigning social policy so that it meets the needs of all Canadians, and leads us towards more civil and economically sustainable communities.

In one respect, Canada is well equipped to meet these needs. We now have a number of excellent and timely social surveys covering a variety of topics. Theoretical developments and advances in research design have led to the creation of longitudinal surveys that track individuals over extended periods of time. These new research tools provide information on the dynamics of poverty, the effectiveness of training programs, the consequences of job loss, the influence of childhood experiences and several other topics pertinent to redesigning social policy. Together, these form the basis for establishing a well-integrated system of "social statistics," a term that encompasses information describing a wide range of human activity and the social, economic, educational and cultural features that affect our daily lives.

However, Canada has very little capacity to conduct social policy research, evaluate social programs, or monitor progress towards achieving social aims. The federal government recently acknowledged the need to strengthen its research capacity, and established the Policy Research Initiative to recommend and oversee the implementation of an interdepartmental research agenda. Similarly, provincial governments, the private sector, and non-government organizations recognized this need and have attempted to revitalize the policy research community.

There are at least three significant barriers that need to be overcome if Canada is to develop its research capacity in social statistics:

• the lack of researchers trained in analytical methods that are appropriate for surveys with complex designs both within and outside government;
• restrictions on access to data; and
• the weak links between the work of social scientists and the potential users of the knowledge they generate.

Even though there is a tremendous appetite for social statistics about education, employment, health, literacy and other pertinent social issues, many of the important findings of social scientific research have not been adequately conveyed to the policy community, or to a wider public through the popular media.

As part of a response to the challenges that confront Canadian policy research, a joint task force assembled by the Social Sciences and Humanities Research Council (SSHRC) and Statistics Canada proposed the creation of a series of Research Data Centres (RDCs). The RDCs will be located throughout the country, so researchers will not obliged to travel to Ottawa to access Statistics Canada data. At the same time, the centres will be administered in accordance with all the confidentiality rules required under the Statistics Act. The Research Data Centres will meet, in a single location, both the need to facilitate access to data for crucial social research and the need to protect the confidentiality and security of Canadians' information.

A number of Statistics Canada's longitudinal surveys have already made important contributions to policy debate in this country. With increased and improved research, however, many of these surveys promise to deliver even more insight on issues surrounding children, youth, a dynamic economy, health and health care, and Aboriginal peoples-issues that are important to Canadians and reflect themes raised in the last speech from the throne.

The Research Data Centres will initially contain data sets from the following new longitudinal surveys:

• National Population Health Survey (NPHS)
• National Longitudinal Survey of Children and Youth (NLSCY)
• Survey of Labour and Income Dynamics (SLID)
• Workplace and Employee Survey (WES)
• Youth in Transition Survey (YITS)

In addition to the benefits to Canadians as a whole, the establishment of the RDCs will also directly benefit the Canadian research community. It is anticipated that the availability of the centres will allow the host universities and affiliated regional universities to attract top quality social scientists, and will assist in the development of strong programs in a number of disciplines. The RDCs could also support the training of Canadian graduate students (through work on research projects) in quantitative research in a number of disciplines, using the latest research techniques and data sources.

In this way, the Research Data Centres will provide opportunities to

• train a new generation of Canadian data specialists;
• build first-rate data and analysis expertise in this country;
• develop research centres in various regions across the country-in both larger and smaller population centres-that can take the data where researchers are, rather than concentrating materials in one or two major centres; and
• continue and expand the collaboration between Statistics Canada, SSHRC, universities and academic researchers that has already borne fruit under the Data Liberation Initiative.

Researchers wishing access to the confidential microdata in the Research Data Centres will submit proposals to a review committee operating under the auspices of SSHRC and Statistics Canada. The committee will carry out the adjudication of the program submissions based primarily on the research merit of the proposal, but also taking account of the need to access detailed data and the capacity of the centre. The review committee will consist largely of academic researchers, with representation from other research communities and Statistics Canada. Researchers who are conducting publicly available research on a not-for-profit basis at universities, research institutes, government agencies, and other public research organisations will be eligible to submit proposals.

The Research Data Centres will provide secure physical locations for confidential Statistics Canada data and all researchers whose project proposals are approved will be sworn in under the Statistics Act. Researchers will have access to Statistics Canada data sets that have been stripped of all personal identifying marks-for example, names, addresses and phone numbers.

The Statistics Act governs access to all Statistics Canada confidential data and its stipulations will be in effect in the Research Data Centres. Researchers whose projects are approved will be subject to a security check before being sworn in under the Statistics Act as 'deemed employees.' Deemed employees are subject to all the conditions and penalties of regular Statistics Canada's employees, including fines and/or imprisonment for breach of confidentiality. In addition, all data to be physically removed from secure areas will be carefully screened for confidential data, whether as direct listings or as possible residual disclosures.

Legally, the RDCs will be the equivalent of Statistics Canada offices with a comparable level of security, including stand-alone computers and physical access controls. Depending on the level of service required and the nature of the tasks to be performed, one or two Statistics Canada employees will staff each site. Their tasks will include vetting output for confidentiality, helping researchers to set up research files, providing advice on the contents of the complex data sets and dealing with access issues.

The results of the research performed at the RDCs will be presented in two stages.

The first stage will consist of research papers produced by the researchers. The papers will undergo the normal peer and institutional review processes. It is anticipated that most of the papers will be included in a Research Paper Series associated with the program. Statistics Canada will jointly operate this paper series with the Research Data Centres. The papers will contain quantitative analysis and results, but will not comment directly on policies or programs, as this is outside the mandate of Statistics Canada.

Having met these requirements, the researchers will then be free to publish a revised paper (or the original research paper) in an academic journal, or any other publication. The researcher would be free to comment on policy and to include any material they deem necessary. Statistics Canada retains the right to publish the original research paper and to scrutinize all publications for confidentiality.

Finally, researchers will be expected to provide feedback on ways to improve future Statistics Canada surveys and data sets.

The nine Research Data Centres are or will be open in the following locations:

• Statistics Canada Research Data Centre at McMaster University
• University of New Brunswick-Fredericton
• Université de Montréal
• Toronto Region-Statistics Canada Research Data Centre
• Dalhousie University
• Southwestern Ontario Research Data Centre
• Prairie Regional Research Data Centre
• RDC at the University of Alberta
• British Columbia Interuniversity Research Data Centre

The first RDC opened at McMaster University in 2000 and the remaining eight will open in 2001-2002, as facilities become ready.