BC Stats Must Allow Verification of Population Numbers,
by William Warren Munroe, September 22, 2011
After a decade of changes to the way population numbers are created, BC Stats has finally made public the methods that will be used this year.
The fact that for the last 13 years, clients, taxpayers, and the public were not informed by BC Stats of the many changes in estimating population is something that must be addressed to ensure it not happen again. These numbers are used to justify opening and closing of public, and private, facilities; to set municipal insurance rates; and used by ElectionsBC, BC Hydro etc.
These numbers were referred to in October 2010, in a report (from a consultant hired by the QualicumSchool District)recommending thepermanently closure of the only high school in the town of QualicumBeach. The report by a consultant referred readers interested in the population estimates to the BC Stats methods paper. The BC Stats methods paper was dated 1998.
I contacted BC Stats to ask why the methods paper from 1998 had not been revised, pointing out that the methods had changed in many ways, particularly after 2001. I should mention that I have been asking that the real methods be made public since starting to work for BC Stats as the Population Analyst in 2002. My efforts were not addressed. Since being dismissed in 2006, I have written hundreds of letters and emails to the Ministry of Labour and Citizens’ Services, and other government officials explaining the real methods.
In August, the new Executive Director of BC Stats replied, informing me that the Generalized Estimation System (GES) methods paper from 1998 was going to be replaced with anew methods paper, now posted on the BC Stats website dated August 2011. The importance of the character of the management team for any Official Statistical Agency can not be underestimated. It means the difference between an Official Statistical Agency that uses scientific methods versus an Official Statistical Agency that allows outside influences on the numbers being provided.
The missing 13 years from 1998 to 2011 was a very interesting time period in BC and many decisions were made regarding the opening and closing of public and private facilities including closing of public schools(nearly 200 since 2001) referring to numbers created by BC Stats.
Many changes to the methods were made in a veil of secrecy under the previous management of Don McRae and Dave O’Neil who took over BC Stats and the Population Section directorship positions after 2001. Perhaps, the new BC Stats management will ensure users are kept up-to-date about the real methods and models used to estimate and forecast population in BC.
BC Statscan publicly announce changes to the way population numbers are createdin a timely manner.Prior to making changes, BC Stats canpresent proposed changes to it’s own analysts(I asked for - but never saw - the testing of Telus data),as well as outside analysts, universities, and other statistical agencies such as Statistics Canada, for comment. The testing of telephone landline hookups and other changes between 2001 and 2011 should still be published, and a review of the data used to justify closing public schools conducted, thus fulfilling the mandate of “Open Government”.
Also, in order to ensure that the numbers are not changed outside of the models without informing users, as well as it’s own analysts, verification can be allowed. This is easily done by making the inputs and the model available on the BC Stats website. These improvements will help reduce error and encourage education and understanding of the strengths and weaknesses of the real methods and models.
Letter to the new Executive Director of BC Stats, Angelo Cocco