Last winter, VCU’s Center for Urban and Regional Analysis put out a report about Richmond’s bus network redesign (PDF) that claimed, among other things, that after the redesign 22% fewer lower-income dwelling units were within 1/4 mile of a bus stop.

Today, the folks who redesigned the bus network, Jarrett Walker + Associates, released a memo detailing errors in methodology and with calculations in that original report (PDF). In fact, following CURA’s methodology and using their data, JWA found that only 2% fewer lower-income dwelling units were within 1/4 mile of a bus stop—and, of course, public transportation is about more than proximity to a bus stop. In fact, JWA notes that, with the new bus network, the average resident can access 6% more jobs in 45 minutes while the average lower-income resident can access 10% more jobs in that same time period.

From the Jarrett Walker + Associates release:

The CURA report also contains numerous calculation errors and key mistakes in its methods. For example:

• CURA used a roadway network that excluded many potential walking paths. As a result, they calculated many walks as being longer than they actually are.
• CURA excluded large parcels on the edge of their miscalculated walking areas entirely, even if only a small portion of the parcel was outside the walk area. For example, in their analysis of the old network, Creighton Court is fully within the ¼ mile area and all 356 dwelling units are counted in their numbers. For the new network, because less than 1% of Creighton Court is not covered by their walk area, none of the 356 dwelling units are counted.
• CURA made manual adjustments to the results for dwelling units in low income areas but only for the old bus system and not for the new one, thus skewing the results in favor of the old network. This manual adjustment, and the other issues with their methods, lead to their erroneous result purporting to show a 22% decline in dwelling units in low income areas within ¼ mile of a bus stop. Strictly following their flawed methods would have shown only a 2% decline, without their skewed manual adjustments

And from a GRTC release:

JWA chose to review GRTC’s network changes after a December 2018 report erroneously and falsely concluded a significantly negative impact on disadvantaged populations. The Center for Urban and Regional Analysis (CURA) at Virginia Commonwealth University released the report entitled “Impact of the GRTC 2018 Reroute on Richmond’s Disadvantaged Population” based on analysis mistakes, manual manipulations and incorrect data calculations, as JWA concludes.

GRTC Interim Chief Executive Officer Sheryl Adams says, “We cannot permit false assumptions to perpetuate in this community when GRTC, in partnership with local jurisdictions and quasi-governments, is committed to responding to the needs of the communities it serves. GRTC strongly endorses the conclusions of Jarrett Walker + Associates.”