I've already ranted a couple of times - here and here - about Statistics Canada's 'Attention Deficit Disorder': its habit of starting new time series using new methodologies without updating the historical data. As I put it in my first rant,
Statistics Canada must be the only statistical agency in the world where the average length of a data series gets shorter with the passage of time. Its habit of killing off time series, replacing them with new, 'improved' definitions and not revising the old numbers is a continual source of frustration to Canadian macroeconomists.
But of course ranting about the problem isn't the same thing as solving it. I kept thinking of this old Peanuts cartoon:
So I decided to light a candle.
I've started Project Link, whose goal is to take the fragments of data published by Statistics Canada and piece them together into a coherent whole. I've started with monthly data, focusing on the main concurrent indicators: GDP, employment (LFS and SEPH), unemployment rates, weekly earnings (SEPH), the CPI, manufacturing sales, exports, imports and retail trade. SEPH data go back to 1950 and 1951, and the LFS goes back to when the monthly survey started in 1953.
I've put together a few charts, and I might as well show them to you. Here are real weekly earnings going back to 1950:
We're all aware of pieces of this story - the post-war boom, the long period of stagnation during the 1970s, 1980s and 1990s and the recent resource boom - but this is the first time I've seen it all in one chart.
And here is the unemployment rate:
Anyone know the story of what was going on during the Diefenbaker years?
And there are other tidbits that one stumbles across while looking through microfiches of discontinued Statistics Canada publications. For example, I found this in the August 1962 issue of The Labour Force:
This surprised me. I thought that the surge in female labour force participation rates started in the 1960s and the 1970s, but it looks like it began in the early 1950s. My next priority is to break out the old LFS data by gender.
So there you are. I'll be updating the data from time to time, but since the series are all scaled to be consistent with the current series, you can update them yourself from Cansim.