Strategic Thoughts.com
  •  About Me 
  •  FAQs 
  •  Mail Me 
  •  Links 
  •  Archives 
March 31, 2012

Fact-Check with CANSIM

The way the BC Liberals are doing themselves in, debate over policy issues may be irrelevant. Many voters have probably stopped listening to anything Premier Clark has to say.

When the Premier speaks, it's rarely about issues of concern to British Columbians and almost always about politics. Her message is that no matter how bad you think the Clark Liberals are, vote for them anyway or you will elect the NDP. That claim will not stop the loss of previous Liberal voters to both the NDP and the BC Conservatives. How long can it be before John Cummins asks Clark to stop splitting his party's vote?

In the event that discussion returns to policy issues, it is worthwhile to know that in February Statistics Canada made access to most of its data free, eliminating the previous charge of $3.00 plus HST for every series. For example, if you wanted employment data for BC and Canada from 1980 through 2012, it would have cost $6.72; if you also wanted full-time employment data, the bill would have increased. Complicated work involving interprovincial comparisons meant spending hundreds for data. The change makes it possible for anyone to fact-check many political claims. To demonstrate how easy it is to access and use this free data, a good start is to watch the five minute CANSIM tutorial.

LFS employment Jan 1991 - Feb 2012Let's say you are interested in checking claims about job creation. Clicking on the top level Statistics Canada website and scrolling down reveals a box with "latest indicators". Clicking on unemployment rate brings up the latest release of the Labour Force Survey (LFS). Scrolling down reveals a list of CANSIM tables for the LFS. It takes some digging, but going through the tables shows seasonally adjusted data in Table 282-0087. From there it is a matter of following the steps outlined in the CANSIM tutorial to select seasonally adjusted data for BC. It is important to select data in small bites because clicking on "select all" produces too much data to easily download and handle. The graph shown here was created by selecting just BC as the province, total employment, both sexes, 15 years and over, seasonally adjusted, January 1991 through February 2012 and time as rows. After clicking "apply" and "download" the data are easily imported into a spreadsheet for creating the graph. Notice it shows growth for most of the past 22 years, with the exception of the 2001 and 2008-2009 recessions. Most of what you hear about jobs in the 1990s compared to the last decade is political spin. You can also replicate the work of the BC Progress Board and see that relative to other provinces BC's job performance ranked fourth or fifth for most of the 90s as it did for the last decade, although BC's performance dropped to seventh in 2010.

The challenging part of using CANSIM to do your own fact-checking is it takes some familiarly with the data. Journalist Andrew Coyne illustrated that point recently when he Tweeted that the Statistics Canada website was difficult to use because he couldn't find quarterly GDP data for Ontario. Those familiar with the data were able to tell him provincial GDP data are only produced on an annual basis; one way of finding this data is to search CANSIM on the term "provincial economic accounts". Coyne was searching for something that has never existed. Keep that in mind when you hear claims about BC's economic performance. Many of the claims are impossible to verify because there are no data. Knowing that is important because it allows those who make claims to be challenged. Now that access to the 43 million data series on CANSIM is free, when claims are made, ask for the CANSIM table number that can be used to fact-check.