The main lessons we’ve learned at Budgetpedia (so far)
This is my opinion, but two years in (starting July 2015), a few things are coming into focus. Quick review: our aim is to make the Toronto budget more accessible, mentally and for interventions. See budgetpedia.ca for what we’ve produced so far. We’re just getting started.
So here’s what we’ve learned:
- Humanize the numbers
- Taxonomies (classification schemes) are good
- We’re in the world of analytics, particularly comparisons and trends
- People want granular data — details, more than overviews
We’re in the second round of website development which is trying to combine these, so I thought it would be good to share. If you have thoughts on these thoughts, let us know (mail@budgetpedia.ca).
Humanize the numbers
This is the main, and hardest lesson. Numbers, particularly on a large scale such as Toronto’s budget (about $12.3B per year), are by their nature impersonal. Boring to most people. So the aim is to liven it up. Ways we’ve thought of so far:
- add a few relevant and enticing social and performance statistics
- add images
- add a brief narrative
- add contextual links, to allow people to explore further
- ultimately, add the ability for people to tell their personal stories in relation to the subject at hand.
Don’t get me wrong, clever graphics are cool! Here are some examples: New York Times, Arlington Virginia, Israel, Palo Alto (by OpenGov.com), Taipai, UsaFacts.org (Steve Ballmer’s new toy). But most people aren’t going to back to these. What are they going to do with them? Instead, they want more detail, real passion topics, and some context.
The best example we’ve found that incorporates these elements is datausa.io, which is a collaboration among Deloitte, MIT, and Datawheel. Looking at a random chart from that website here, note the sparse use of highlighted related data. Also note that each chart has (above it) contextual options, including View Data, Save Image, Share/Embed, and “Add to Cart” (a compilation option).
We’re working out our own approach to implementing these principles. It may take a few iterations to perfect.
Taxonomies (classification schemes) are good
Taxonomies are natural groupings according to carefully selected common, shared, highly relevant attributes. It’s hard to do well, but worth the trouble.
At Budgetpedia, we’ve developed a taxonomy for the operating budget that many people find is quite approachable. At the top level, the budget can be broken down into General Services that everyone uses (Transport, Utilities, and Public Commons), Support Services that can be characterized as constructive interventions (Human Services — income, housing, and child support, Emergency Services — police, fire, and paramedics, and Health Services — public health and long term care). See our blog The Toronto operating budget, simplified for a bit more detail.

These schemes can provide a level of comfort, natural completeness, and confidence that invites further navigation around the categories, and further research.
We’re in the world of analytics, particularly comparisons and trends
There’s something about our minds that makes differences and changes interesting. Enter analytics as a structured way to take advantage of this. At its simplest, analytics provide visualizations that communicate descriptions, comparisons, and trends to elicit insights. Much has been studied and written about this, and we’re trying to take advantage, without getting too complicated. Below is a chart that shows the change of Toronto’s labour costs over time, in relation to all other costs. The trend is clear (staffing costs have gone from about 41% to about 51% of all costs), and invites attention (what’s behind it still isn’t clear to us, but that’s another story).

People want granular data — details, more than overviews
Interactions we’ve had with people looking at our site, and responding to our presentations over the last several months have made one thing clear: people’s budget related passions are usually quite specific, related to their neighbourhoods or related to specialized concerns, or both. Things like employment issues, access for the physically challenged, youth issues. People want to see detailed data.
The City of Toronto maintains budget and accounting information in units called ‘cost centres’ of which there are some 13,ooo according to a list we obtained through a Freedom of Information requests. In terms of accounts, the City maintains ‘cost elements’ (like materials, wages, salaries & benefits). For staffing, the city maintains full-time-equivalent lists. It is the combination of these (cost centres, cost elements, detailed staffing complements) that we refer to as granular data. We believe that this is a practical goal for the delivery of the detailed data that people want.
Now all we have to do is get it. This looks like it will take some time.
Henrik Bechmann is the project lead of budgetpedia.ca. The opinions expressed here are his own.

