For more information contact
Service BC Contact Centre

1996 Census:

Notes for Provincial Electoral Districts


This document is the latest in a series of census-based profiles of British Columbia Provincial Electoral Districts (PED) produced by BC Stats. Our print publication based on the 1991 Census was the first to feature both graphic profile pages and a series of tables showing specific characteristics across all PEDs. Single-page graphic profiles of current PEDs, based on the 1991 and 1996 Censuses, are posted on our web site.

The initial set of 1996 Census-based graphic profiles conform to the riding boundaries in effect for the general provincial election held on May 28, 1996 and for subsequent by-elections. This report provides a demographic profile of PEDs based on the boundaries proposed by the Electoral Boundaries Commission on June 3, 1999. These boundaries will be in effect in the first general provincial election following the date of their adoption.

Methods and Data quality

Readers should note that, for a number of reasons, the population counts published here do not exactly match that published by the Electoral Boundaries Commission. In addition, the sum of the ridings themselves does not total to the BC 1996 Census population published by Statistics Canada.

To calculate total populations for each PED, the Commission relied on population data from the 1996 Census at the smallest available level of geographic detail. This allowed the most accurate possible estimation of the effect of boundary changes on included population.

"Enumeration areas" (EAs) are the basic collection area for a census and for each EA Statistics Canada produces profile data. In densely populated urban areas some data is available on a block-by- block basis. However, that "block face" data does not reveal any characteristics beyond population and dwelling counts, whereas EA profiles contain a rich array of demographic information.

The best estimates of total population in a PED will be based on populations of EAs and, where available, block face populations. The Electoral Boundaries Commission used population data provided by BC Stats based on this approach. Table 1 contains BC Stats' estimates of 1996 Census populations for the proposed PEDs. This data adds up precisely to the provincial total as "exact" population and dwelling counts are published for all the EAs and block faces in the province.

The remainder of this publication relies solely on EA based data published by Statistics Canada. This loss of geographic definition in urban centres is in exchange for a rich array of demographic characteristics. Some idea of the difference is provided in Table 1. As we move to the right from the column with EA and block face data we first encounter unrounded EA populations. This is a measure of the loss of the block-face information. Moving again to the right we have the result of figures that have been randomly rounded for each EA and then added to produce an electoral district total. This is a measure of the loss incurred as a result of a deliberate obscuring of the data by Statistics Canada so that individual responses to a census cannot be identified.


In order to preserve confidentiality of individuals when publishing census results Statistics Canada employs rules of suppression and random rounding. The graphic profiles and data tables that follow are based on published data for non-overlapping groups of EAs that, when combined, most closely follow the PED boundaries. No data are published for EAs with a population of less than 40. For income tables, no data are published for EAs with a population of less than 250. All published data (except for one table with population and dwelling counts) are randomly rounded to multiples of 5 (numbers ending in 5 or 0). The charts and tables published in this document are based on figures that have first passed the suppression barrier and then been rounded at the EA level, before being aggregated to reflect riding totals.

Graphic Profiles

We have produced a graphic profile page for each of the 79 new PEDs. The profiles result from data assembled from the table section that follows. The actual source table(s) for each panel are shown in Figure 1.

The charts for ethnic origins and home language are sorted on the prevalence of these characteristics in the PED that is being profiled. We accomplished this by using detailed listings of the characteristics from the Profile series of Census products available from Statistics Canada. This is the first time we have had the ability to capture and organize the information in this way. We believe this presentation is much more useful and well worth the extra effort of production.

Official languages in Canada are English and French. The home language table above the chart identifies what portion of the population speaks a single non-official language at home. This is the group that is featured in the chart. There is no breakout when more than one language was identified (multiple response) and English and French are not candidates for the chart. In many areas of the province a very small portion of the population will be the subject of the chart. In other areas approximately half the population qualify.

We chose not to publish the detailed ethnic origin and language tables in this report as the size of the tables did not conform to that of the others. These tables are to be found on our Internet site in the 1996 Census area. Users may wish to download them and import them into a spreadsheet or database.


Each of the tables presented contains data summed to each of the 79 new PEDs. The "Report Total" is the sum of the PED data. The last line in each table provides a figure that Statistics Canada has calculated from unrounded EAs and then rounded only as a total. The difference between the "Report Total" and the "BC total from unrounded data" provides an indication of the quality of the tables.

Custom tables available

BC Stats regularly receives or purchases census data at the EA level and, based on translations of the EAs into different administrative boundaries, produces tables for other administrative regions within the province. The alternative- purchasing data for each of the required geographic systems so that region totals could be calculated from unsuppressed and unrounded components-would provide superior estimates but at a significant cost.

One benefit of our cost conscious approach is that we can produce similar output tables for components of any region. For instance, users may contract with BC Stats for the production of similar tables for neighbourhoods within ridings.


A glossary is provided to enhance the usefulness of this document. Not all census terms are common-place and even common terms have strictly defined meanings.