Inventory of Community Data Sources
Summary and Next Steps
This inventory was compiled to provide municipalities and their communities with more information about what social data are available and how they may be accessed. There are two main issues to consider in making data more accessible at the local level: first, the organization and dissemination of available data, and second, the development of new data sources.
Dissemination of Existing Data
The national Census remains the most comprehensive source of municipal-level social data in Canada. Data are available for geographies as localized as a single city block, and these can be aggregated to provide the data for user-defined neighbourhoods and other small areas. The main limitations of Census data are that they cover only a limited number of issues − there is nothing about population health, for example − and the fact that they are collected only once every five years. Moreover, there is a considerable time lag between when the data are collected and when they are published. That delay is even longer for data released at lower levels of geography − i.e., for Census Tracts and below, neighbourhood planning districts, municipal wards, etc.
There have been important advancements in increasing the accessibility of Census products by making data down to the CSD level available free on Statistics Canada’s website. Other major improvements include publication of the 2006 Census Tract Profiles, the ability to locate specific CTs using postal codes, and the new Census Trends product which tracks key indicators over time for CMAs, CDs, and CSDs. However, users can only access data for one area at a time. More importantly, few people know about these new products, so much more could be done to raise their visibility. For example, given the proven success of the basic community profiles, Statistics Canada could combine CT profiles with a time series to create a single, comprehensive ‘Community Profile’ product.
Survey data are available more frequently than the Census, and the addition of questions to the monthly Labour Force Survey regarding immigrant and Aboriginal status will provide a new and more frequent source of data for these target groups. Still, sample sizes are limited. As a result, data are available only for the larger geographic areas, such as CMAs, but even there the sample may still be too small to provide useful information about any single municipality. Unlike the Census, which can provide data down to the block face, surveys provide no information at the sub-metropolitan level.
Statistics Canada should be asked to routinely tabulate national survey results for CMAs, where sample sizes warrant, as it does for provincial, territorial, and national results. For example, the General Social Survey (GSS) could provide data for many CMAs. In addition to designing survey samples that are large enough to capture data for all CMAs over a certain minimum size, Statistics Canada should ensure that, for each of those CMAs, the same information is collected and disseminated.
The Summary Tables − formerly called Canadian Statistics − provide an overview of statistical information on Canada’s people, economy and governments. Individual tablescan be selected by subject, province or territory, or metropolitan area. The Tables constitute a first step in consolidating and aggregating existing CMA-level data from non-Census sources. However, it is not clear from the Statistics Canada website whether the Summary Tables include data at the CMA level. Moreover, the Tables do not have the same information for all CMAs and only appear to include consistently (social) data from the Labour Force Survey, Building Permits and the Census. It is unclear why some CMAs are included and others are not. For example, while Barrie and Kelowna had larger populations than Abbotsford, Thunder Bay and Trois Rivières in 2001 and 2006, they are not included in the list.
Whenever surveys produce data for CMAs − such as in the General Social Survey cycle on victimization − those data should be included in the Summary Tables. And as survey samples are redesigned to provide data for all CMAs of approximately the same size, as suggested above, those data should also be included in the Summary Tables.
New Data Sources
Administrative data hold the most potential for further data development. Examples include the Longitudinal Administrative Databank (LAD) and much of the educational data collected by Statistics Canada's Centre for Education Statistics (CES). LAD is an annual subset of data on families collected from information provided to Revenue Canada in personal income tax returns. The CES, in collaboration with the Council of Ministers of Education, publishes Education Indicators in Canada, a comprehensive, pan-Canadian collection of national, provincial and territorial statistics on enrolment, the awarding of diplomas and degrees, tuition fees, and the costs for elementary, secondary and post-secondary education, among others.
The LAD and Education Indicators in Canada are good examples of how administrative data can provide a wealth of information at relatively low cost and with minimal response burden. They are also available more frequently than the Census and most national surveys. However, there are some disadvantages. The issues on which data are collected may not correspond to research or policy needs, they may cover only a limited population, and there can be problems with data quality. The level of geography reported may vary from one data supplier to another, as is the case, for example, for crime data and some health statistics. Even where data can be mapped by postal code, those data may not aggregate to the expected geography. Still, where postal codes are reported reliably, there is a potential to create user-defined areas such as neighbourhoods.
A separate initiative is clearly needed to explore how administrative data could be more fully exploited to provide annual information at the community or neighbourhood level. The Health Indicators product produced jointly by the Canadian Institute for Health Information and Statistics Canada (see p. 41 in this document) demonstrates that it is possible to provide community data on a regular basis. And the Canadian Council on Learning has extended existing survey data to the community level in developing their Composite Learning Index (see p.40 in this document). A similar organizational model should be considered for developing and housing administrative data for use by municipalities. Both CIHI and CCL are potential partners in improving the quality and availability of community-level data.
There are also important gaps in key content areas of social data − such as data on housing and homelessness. For example, while an increased number of shelters now contribute to HRSDC’s Homeless Individuals and Families Information System (HIFIS), coverage in western Canada remains incomplete (See p. 8 of HIFIS Annual Report 2006-2007, accessed January 29, 2008.) Implementation of HIFIS should be monitored, so data sharing protocols can be established as coverage improves. (More information on HIFIS available here).
At a practical level, income tax records represent a good starting point in making more social data available at the neighbourhood level. Infrastructure for the collection and dissemination of these data is already in place, and links between postal codes and Census geographies have been created. A number of indicators could be determined, such as median income, the proportion of families in low income, the proportion of the population receiving employment insurance, social assistance, CPP, and Child Tax Credit, and the proportion making charitable donations. While such data are currently available on a cost-recovery basis, Statistics Canada should be encouraged to make them freely available − at least down to the CSD level − as is now the case for the Census.
Other countries have succeeded in creating successful dissemination models for small geographic areas by combining data from various sources. In particular, the website for U.K. neighbourhood statistics is especially impressive, with over 250 data sets. In the United States, the United Kingdom, and Australia, users can simply type in a postal code and obtain a wide range of information for that area. Within Canada, Newfoundland and Labrador, Nova Scotia, Québec, and B.C. have also created successful data dissemination models.
All these examples demonstrate that it is possible to make community-level data readily accessible. The challenge now is to do so for all Canadian communities.
| <-- Back | Next --> |