Methodology for prioritizing streets for multimodal treatments

By Robbie Webber
Building Complete Streets can assure that our transportation network is safe and comfortable for all road users: people biking, walking, and taking transit, as well as drivers of cars. But which streets are most appropriate for Complete Streets treatments—to encourage non-driving modes and maximize the safety of people using these modes—can be a tough question to answer. A recent article in Transportation Research Part A: Policy and Practice describes the prioritization methodology used in Quebec City for making these decisions. The technique may not work for all cities, and would be difficult to implement in larger regions, but the data and criteria used for decision making may be instructive to staff in other jurisdictions.
There are numerous design manuals that can be used to improve the multimodal potential of streets of all kinds, but less has been written about which streets should be given Complete Streets priority. As the article states, “At the current time, most of the Complete Streets policies and design guidelines pertain to the ‘how’ of street design rather than to the ‘where’ streets should be redesigned by priority.” Using a multicriteria decision aiding approach allowed city staff to decide which street segments held the greatest potential to become Complete Streets. The criteria used were:

  • Bicycle network. Data including where facilities existed, whether they were in good condition, where facilities were planned but had not been built, and the level of priority assigned in bike network plans.
  • Bus network. Various characteristics of ridership, types of buses, frequency, transit plans, and time of day for routes were used as proxies for the quality and ridership potential of the transit service on street segments.
  • Citizen input. Both a record of citizen complaints and petitions were considered as input, but only petitions were added to the model, because they involved more widespread indicators of needed changes.
  • Although many methods exist to measure connectivity, this model used Normalized Angular Choice, which measures the potential of a street to be used as the shortest path from every segment to every other segment within a given radius.
  • Human activity density. This represented the proximity of a segment to local and major trip attractors, and was measured by the numbers of residents, students, and workers in the smallest geostatic unit from the Canadian Census divided by the surface area of the unit in hectares.
  • Pedestrian flow. Origin and destination surveys are conducted by Quebec City every five years. The weighted pedestrian flow was calculated using these surveys instead of actual pedestrian counts, as the counts were deemed to be more variable and therefore less reliable.
  • This was calculated based on the number of collisions involving pedestrians as a function of the pedestrian flow.
  • Social and material deprivations. Based on socioeconomic indicators, street segments that had higher population density of people at risk were given more priority for Complete Streets.
  • Street width. Because narrow streets can be harder to retrofit, wider street segments were given priority in the model. Wider streets also allowed more opportunities for street trees.
  • Urban planning. Street segments that appear in more high-priority plans were given more weight in the model.
  • Urban street canopy. All other things being equal, street segments with a lower tree canopy coverage were given more weight.

Clearly, this methodology required data that may not be available everywhere. And many jurisdictions will not want to spend the staff time to collect and input all the data. But looking at the criteria, methodology, and reasoning for why the criteria were selected may help agencies prioritize which streets are most appropriate for conversion to Complete Streets. The full article, explanation of the criteria selection, and methodology can be found here.
Robbie Webber is a Senior Associate at SSTI.