Parklet & Bicycle Lane Impact Analyses

Published: March 2014

by Ben Weinryb Grohsgal and Karina Qian / Harvard Kennedy School of Government

Graduate Scholars Ben Weinryb Grohsgal and Karina Quian from the Harvard Kennedy School tested methods for analyzing how San Francisco’s parklets and bike lanes impact their neighborhoods. The policy analyses by Grohsgal and Qian using sales tax and crime data found that parklets are associated with dropping crime rates and higher sales for certain types of businesses.

“How Do Bike Lanes Affect the Sales of Nearby Businesses?” (Qian, March 2014)
Exerpt — “Some studies, especially recently, have attempted to articulate the effects of bike lanes on businesses using surveys or by comparing retail sales trends on streets before and after the installation of a bike lane. However, none thus far have used robust statistical analysis to evaluate whether the bike lane had any significant effect on business health that can be directly associated with the bike lane, rather than with larger economic forces in the city, variation over time or unmeasured business characteristics. This study combines spatial and statistical analysis to construct a business‐level panel dataset that eliminates these potential confounding effects.”

Download the Report – PDF 2MB

‘San Francisco: Change on the Streets” (Ben Weinryb Grohsgal, March 2014)
Exerpt — “As the result of the recent push for open and digitalized data, city governments like San Francisco have more data available to them and their citizens than ever before. But in the realm of urban planning, specifically impact evaluation of streetscape interventions, that data is rarely put to best use. Instead, traditional intercept-survey-based methods and in-person people-counting still dominate the area of impact evaluation despite being time-consuming, expensive, and often methodologically statistically poor. This PAE presents an alternative framework for impact evaluation that: (1) Relies on existing datasets; (2) Attempts to distill the impacts of an intervention using sound statistical methodology; and (3) Takes into account citywide and neighborhood-wide trends that may be interacting with the effects of the intervention.”

Download the Report – PDF 12MB