Tuesday, February 14, 2012

Taking numbers and making pictures

One way to measure some amount of economic success is to compare the imports and exports of a place.  So let's do just that for all of Canada from last year and then break it down provincially.


Statistics Canada makes available the total international imports and exports of each province on a monthly basis.  With that, it is no great trick to get yearly totals and then calculate the surplus or deficit of trade.  Since it stands to reason that more populated provinces will have more in raw numbers, a comparison is most easily done on a per capital basis.  So, a quick check of the census numbers that have been just recently released and a comparison can be done. 


All that math breaks down to look like this: 

Region       Per Capita Trade Surplus or Deficit (2011)
National -$806.56
Newfoundland and Labrador $16,441.80
Prince Edward Island $4,773.76
Nova Scotia -$4,275.02
New Brunswick $1,590.72
Quebec -$1,550.74
Ontario -$7,745.24
Manitoba -$3,650.51
Saskatchewan $19,465.71
Alberta $18,809.29
British Columbia -$1,716.68
Territories $14,873.44


Well, hey, that's pretty neat, though the numbers just kinda float out there by themselves.  The quickest explanation for the disparity between provinces comes down to how much oil or fish a province has available for export.  If the answer is: "lots", that province will have a massive trade surplus. 

It isn't that simple, there's plenty of sustainability talk that probably needs to happen there.  Manitoba or Ontario could likely rocket to the top of the standings if every tree in the province was cut down and shipped elsewhere for pulp and paper.  But that's a really bad idea.  But as a birds-eye view, it is useful.

Tables are nice, but here's a neat thing that IBM let me do:

2011 Provincial Total Trade Surplus or Deficit Many Eyes
That should be clickable, so check it out. 

Now, here's the thing about the IBM Many Eyes tool: it's pretty awful.  For a couple of hours, I couldn't get the visualizations to work.  Turns out that I needed to update Java, but there was nothing on that site to tell me I needed to do this.  That's a really crappy thing to do to a user.

Next, it is pretty rigid.  As you'll note from my data set, I combined the territories and included a national average.  These are not things Many Eyes is capable of handling.  It's easy enough to exclude them, but it means some of the picture is missing.  Yes, I could have done the independently, I didn't because their populations are really small.

Finally, don't blame me for the colour choices.  If there is an option to change it from horrible brown and overused blue, I can't find it.  You take what you get with Many Eyes, which is better than nothing.  But it is hardly perfect and kinda skirts the territory of "not very good".

No comments:

Post a Comment