Flagging the data

You may have noticed the increasing presence of buckets of bright orange flags placed on either end of crosswalks throughout Halifax. Since council passed its Pedestrian Safety Action Plan in 2014, community members have been allowed to set up crosswalk flags “with input from HRM staff to help ensure programs are carried out appropriately and safely.” There are now, according to city staff, “approximately 154” crosswalks with flags in HRM, a small fraction of our total number of crosswalks, but a growing and significant number.

Last week council considered passing new regulations to create a formal approval process for crosswalk flags. The new rules would have seen flags removed at about 56 current locations. After a couple hours of discussion, council rejected staff’s proposal, opting to leave current flag locations alone, but to suspend further installations indefinitely, while asking staff to “begin a pilot of various treatments to enhance the visibility and safety, with relevant high-quality data collected pre-treatment and post-treatment, including at least one treatment of crosswalk flags,” with a progress update due in December.

I both love and hate crosswalk flags.

I love that they are a cheap, simple, grassroots answer to a classic problem facing pedestrians: drivers who just won’t yield. But I hate that they are essentially one more way to shift responsibility for the behaviour of said drivers onto pedestrians.

In some places, crosswalk flags are not so much a safety measure as a necessity for ever being able to cross the street. Take some of the T intersections along Waverley Road, where traffic might never yield to a pedestrian without being actively flagged down. A bucket of orange flags in those cases is more about establishing basic access than increasing safety. In low-pedestrian count places where the city won’t put up lights or even in some cases mark the crosswalk, they provide a way for the infrequent pedestrian to get across the street.

But there’s resistance to crosswalk flags from Halifax’s traffic services staff, dating back to when former Traffic Authority Ken Reashor simply refused to follow a council recommendation to implement them based on his legal jurisdiction over the matter.

It’s this resistance that I don’t quite understand, especially considering where it comes from. The crosswalk flag system is essentially perfectly consistent with the predominant messages that come through pedestrian safety campaigns in Halifax: wear bright clothes, be seen, make eye contact. My kid learned, presumably at school, to stick his arm straight out to indicate when he’s crossing.

Well, crosswalk flags directly enable all of this. With a crosswalk flag in hand, even the most funereal dresser is suddenly donning hazard orange. And the simple act of picking up the flag means you are actively engaging in “getting noticed” as Halifax’s Heads Up campaign advises. And as Councillor Stephen Adams charmingly pointed out at last week’s council meeting, “If someone has a flag in their hand, at least they’re not texting.”

And that brings me to what I don’t like about crosswalk flags: ultimately, they are one more method for putting the onus on the pedestrian to “get noticed.”

If flags were to become ubiquitous in Halifax, there’s a distinct possibility that in a few years some provincial government will even decide to make them mandatory, just like they did with pedestrian-activated lights at crosswalks. (You can actually be ticketed if you fail to press the button at a crosswalk. If Nova Scotia ever requires me to carry a flag to get across the street legally, well, then my Nova Scotia days will be numbered.)

All these concerns aside, what councillors really wanted to know last week was, do they work? Do crosswalk flags make pedestrians safer? Because we don’t really know. Most of the research out there focusses on marked versus unmarked crosswalks, and even when it is expanded to include different types of markings, I couldn’t find much that addressed flags directly. And flags are a distinctly different kind of marking, designed to move along with pedestrians.

Our own city reports to date have included limited data collected by city staff during “data collection season” from May to October. But the numbers are small. Staff observed 131 pedestrian crossings at 12 different crosswalks over the course of months, to come up with a flag usage rate of just under five per cent. And then 413 cars passing through 24 different crosswalks over the course of months, to determine average yielding rates from 89 to 95 per cent.

But as councillors Cleary and Austin pointed out, the data doesn’t reflect who the five per cent are (i.e., are they seniors and/or kids?) nor does it tell us the differences in behaviour at a crosswalk before and after flag installation. In short, the data we have is not much to go on.

Which is why it’s great that council is asking for more. But it’s a tall order, and all that data collection and analysis isn’t free. I know councillors would like to get back some solid evidence in December on which to base a yea or nay on crosswalk flags, but the reality is, they probably won’t. Read between the lines of last week’s motion for a “pilot of various treatments to enhance visibility and safety” and you find a much longer term, more comprehensive research project that could inform all of our street design and traffic tool decisions.

And why shouldn’t we go for that? If the urge to make informed, evidenced-based decisions is strong enough at council, then perhaps what we need is something like what New York, Seattle, and New Orleans have done recently with non-profit think tank Datakind. As described in a recent Citylab article by Laura Bliss, New York is two years in to a massive data project “to build a tool capable of projecting the impact of any given engineering intervention, on any given New York City street.”

In Halifax, we have a natural partner for such a project: the Dalhousie Transportation Collaboratory (DalTRAC), with whom we’ve just signed the contract on the first ever Halifax mobility and travel survey (a full year after council approved it, but better late than never I guess). It would, of course, take years to build such a model. New York City’s DataKind project is still years off from being able to recommend specific treatments for specific intersections, but in the meantime as the model grows it is still useful to engineers along the way.

Halifax council should harness its enthusiasm for evidence-based decision making, and instead of simply requesting an internal staff report, actually make the proper investment to either hire the personnel, or partner with DalTRAC (or DataKind, or anyone with the capacity) to build a truly useful tool to inform our future safety decisions.