Tag Archives: data journalism

| My articles, Opinion, Public transit

Crunching the numbers in the STM’s budget

A week ago, the STM hiked its fares yet again, making it more expensive to take public transit in Montreal. The hikes represent about a 50% increase over a decade, well above the rate of inflation.

Politicians, social activists and regular transit users complained, as they do every year, that this is unacceptable, making it harder for those who don’t have money to get to and from work. The STM and its defenders point to the fact that public transit in Montreal is still much cheaper than Toronto and other comparable cities. The other side comes back with the argument that income and costs of living are lower in Montreal, so you can’t compare cities like that.

A few weeks ago, I started wondering whether the STM’s rate increases were truly reasonable, so I started going through its budgets and plugging numbers into a spreadsheet. I also got some numbers from Statistics Canada, such as the consumer price index in Montreal, the population of the city, the median family income.

The result is a story published in Thursday’s Gazette that tries to give a quantitative picture of the situation. The headline is that the increase in fares, at least for the years 2005-2013 (we don’t have figures for 2014 yet), corresponds to an increase in service in both métro and bus service, all going up about 30% over that period.

But is that the proper way to measure whether it’s reasonable? The increased service came with an increase in the number of users, which means more fare revenues. Should that be taken into account? What about our ability to pay?

With the help of the Gazette’s data guru Roberto Rocha, we put together an interactive chart with the story that allows you to make comparisons for yourself of how things have changed since 2005. Fare price versus total trips. Total salaries and benefits vs. total bus service. Revenue per trip versus total passes sold. Total revenue versus population.

But even that’s only a subset of the data analysis that can be done. So I invite you to do your own: Download this spreadsheet and compare numbers to write your own story.

STM fares in context (.xls, 42kb)

The figures don’t all look good for the STM. Salaries and benefits are going up higher than the amount of work done to justify them. And the amount of subsidies from the Quebec government has gone up more than 200%.

But before you blame the unions or some other invented bogeyman, consider that the cost per hour worked at the STM went gone up about the same as the median family income in Montreal.

Missing numbers

There were some comparisons I wanted to make that I couldn’t. Reading the comments below the Gazette piece, people point to executive salaries. I wanted to include that, but it’s hard to quantify because of the changes in executive pay. For example, I could put in the salary for the director-general of the STM. But that salary went down significantly when Carl Desrosiers took over the position, and in the latest numbers were still lower than his predecessor (though not much).

Another is the price of oil. Including its value could easily give the impression that the fare hikes are more than reasonable. But the price of crude has plummeted in recent months, which would not be reflected on these charts because they end in 2013. So I asked Jeanine Lee, our graphic artist, to take it out of one of the charts we were using.

And there are numbers that can’t be easily calculated, like the number of overall users. Or numbers that are subjective, like customer satisfaction.

What metrics would you use to judge the STM’s performance? And what do those numbers show?

| Montreal, Public transit, Video

An animated day in the life of Montreal’s bus network

It’s fun the kinds of things you can do with data.

Montreal’s transit agencies, including the STM, STL, RTL and AMT, have made their trip data public through a standard called General Transit Feed Specification (GTFS). This allows the data to be sucked into applications like Google Maps, making it easier for people to plan their trips. The time of every stop of every bus is a set data point.

In this video, published a few weeks ago on YouTube, someone has taken this data and created an animation of every bus trip during the average weekday in the Montreal area. STM, STL and RTL buses are represented by little dots that race along their routes.

It’s an interesting way to visualize the activity involved in public transit. The animation, which is presented as a 1:600 timelapse (every second represents 10 minutes), starts at 4am with just the night buses on the island of Montreal. After about 6am, it expands into the morning rush hour, and you can see a clear bias toward downtown from all directions. Some thoroughfares like Henri-Bourassa Blvd., Sauvé St., Parc Ave. and Côte des Neiges Rd. emerge as lines because they see so much bus traffic during this time. The traffic dies down a bit after the morning rush hour, though not as much as I expected. After about 3pm there appears to be a general bias away from downtown as the evening rush hour begins. After 7pm, it noticeably dies down, more so after 11pm and 12:30am, and after 2am it’s back to just the night buses.

Each of those dots is a bus with a driver in it. Some could have just a few passengers on board, while others could be so packed they’re not stopping to pick up more.

It’s an expensive system, and a complicated one. But without all those little dots, the city would grind to a halt.

If you’re interested in trying to figure out other cool ways of manipulating transit data, you can download the STM’s GTFS data yourself. Data from the RTL and STL and AMT are also available. (The AMT data includes commuter trains, its express buses and data from smaller transit agencies like the CIT du Sud Ouest and CIT La Presqu’île.)

Continue reading →