Case Study

Personalized Routes and Bus Lottery

Background

In 2018, the City of Durham partnered with the Center for Advanced Hindsight to compete as one of 35 Champion Cities in the Bloomberg Philanthropies Mayors Challenge. An initial seed grant of $100,000 was used to test ideas aimed at reducing single-occupancy vehicles (SOV) or drive alone commuting into downtown Durham.

Problem

Driving to work alone remains the most common way employees commute in the U.S., at 76.4% of mode split in 2017. The City of Durham employs around 2,800 people with a SOV commute rate of 85%. The goal of this study was to reduce the SOV commute rate of city employees and to maintain this reduction throughout the five-week study.

Exploration and Prototyping

We gathered input from 1,100 unique people through 10 feedback approaches and decided to move forward with two of the four proposed ideas: personalized routes and a bus lottery.

We prototyped an efficient mobility program that included a digital planning tool and a weekly bus lottery.  The tool uses live transportation data to auto-generate emails that recommend optimal, personalized routes based on an employee’s location, work hours, and unique circumstances. It also highlights personal benefits from calories burned, money saved, and free time gained.

Figure 1. Iterations of the personalized route prototype.

 

Experimental Design

We ran a randomized controlled trial (RCT) to test two main hypotheses:

  1. Providing people with personalized alternative transportation routes reduces people’s habit of driving alone
  2. Offering people a chance to win a weekly $163 cash prize for riding the bus increases the attractiveness of transit and usage of the personalized route planning tool

To test these hypotheses, 1,570 City of Durham employees were randomized into three conditions:

  1. Control Group – received only surveys about their commuting behavior
  2. Personalized Route – received personalized route emails and surveys
  3. Personalized Route and Bus Lottery – received personalized route emails, an email about the bus lottery, and surveys

Personalized route emails displayed routes by bike, bus, walking, carpool, and park and ride from individuals’ home to work addresses. They also included trip time comparisons and listed the potential health, time, and monetary benefits of commuting via this new alternative mode in comparison to driving alone. The bus lottery email informed employees about their eligibility to enter into a bus lottery, which offered a weekly chance to win a cash prize for riding the bus.

Results

We measured the effectiveness of the intervention through three main strategies: pulse surveys (self-report of commuting behaviors for each commute on a weekly basis), evaluative final survey (reflection of commuting behavior and time use over the whole study period), and bus data (tracking of individual GoPass numbers for the first two weeks of the study).

Figure 2.  Average commute behavior as a percentage of total commutes throughout the study.

Relative to the control group, the average proportion of drive alone trips was 8.2 percentage points lower for those who received the personalized route and 9.3 percentage points lower for people who received the personalized route and bus lottery. Over the five-week study period, we saw a stable difference between conditions with no sign of decline.

Figure 3. Weekly average commute behavior as a percentage of total commutes during the study period.

We saw, with reasonable consistency across data collection methods and over time, that the personalized route condition performs most effectively and leads to at least a moderate decrease in drive alone trips.