How much of Edmonton’s GDP is dependent on the energy industry? What is tourism’s contribution to our economy? What are the major international exports from Edmonton businesses? These are questions I get asked all the time. They’re important questions that seem pretty basic. The answer in each case tends to surprise people.
Well, maybe somebody knows. But the official answer lives on a Statistics Canada server somewhere, inaccessible to the public because of privacy concerns or data quality issues or the constraints of Statistics Canada’s budget. Or the answer exists, but the latest available data release is years old. It’s the biggest challenge of working with municipal data – there just isn’t very much of it.
In other cases, we have some available data, but the question we’re asking doesn’t map nicely to the data structure. For example, say we wanted to find out how many people are employed in Edmonton’s arts and culture sector. We have pretty detailed information on employment by industry in Edmonton (but, of course, the detailed information is only available every five years), so no problem right? Well, the industries are categorized according to a standard classification system called NAICS – and there’s no NAICS category named ‘arts and culture’.
So we need to come up with a list of NAICS categories that we’re going to call the arts and culture sector. Lucky for us, a quick Googling reveals that Statistics Canada has already done that for us. Great, are we finished? Nope. The classifications are filled with so many caveats and exceptions that even with the most detailed available data we’re going to end up with an estimate that drastically under-estimates the number of people who do work related to arts and culture.
You can see how even with a nicely defined question and the best available data, we quickly end up with an answer that’s not much better than ‘we don’t know’. Now repeat this for tourism, or the energy industry, or any number of other questions you might have. Either the data doesn’t exist, or it’s not structured in a way that’s very helpful.
The list of everything we don’t know about Edmonton’s economy is long – too long to write here. So here’s my wishlist of the top three things I’d like to know:
- The size and structure of Edmonton’s economic output. There are no official Statistics Canada estimates of economic output for Canadian cities. Every time you hear someone say that Edmonton’s GDP is x billions of dollars, that’s an unofficial estimate with a lot of assumptions built in. Official data on Edmonton’s output by industry would help us know how our economy is performing, which sectors are driving growth or lagging behind, and how productive our businesses are.
- An up-to-date time series of business creation and closure. Twice a year Statistics Canada releases a detailed count of business establishments by industry and size. But for various reasons these counts are not comparable over time. Even if they were, they only give us the net change – if the count increases by 1, we don’t know if 1 new business opened, or if 100 businesses closed and 101 opened. If we could see the market entries and exits over time, we could learn if efforts to increase entrepreneurship are working, or if new entrepreneurs in a certain industry need more support and training to help them succeed.
- Foreign trade and investment flows to and from Edmonton. If we knew what could goods and services were being exported from Edmonton, and where foreign direct investment was coming from, we would have a much better idea of where to point local businesses looking for new markets or new capital.
I’m not optimistic that we’ll ever get official answers on these topics. Even adding cities to GDP estimates would be a gargantuan undertaking for Statistics Canada, and their resources are more limited than ever. But there’s some hope. The rise of open data from government is a good step towards learning more about our city’s economy. If we can keep pushing for more, better quality open datasets, we can start to shrink the list of things we don’t know.