Complete Streets State Statutes and Provisions, January 1972 - July 2018

Informed by established legal mapping methods, a comprehensive survey and systematic qualitative analysis was conducted of Complete Streets legislative statutes made effective between January 1972 and July 2018. LexisNexis Academic and Fastcase legal research databases were used to search for and obtain full statute texts in all 50 states and the District of Columbia. The search terms that were used included: “complete streets,” “pedestrian facilities,” “pedestrian accommodation,” “pedestrian and bicycle,” “pedestrian or bicycle,” “road construction,” and “routine accommodation.” To validate and inform the search results, comparisons were made between the laws obtained through the two databases and a publicly available list of known state Complete Streets statutes that have been documented by the AARP et al. and the National Complete Streets Coalition.In cases where only a citation or act number was available in either database, websites of state legislatures were visited to obtain the full statute texts. After reviewing the text of each statute, a list of variables were developed and defined to systematically code the statute texts. Variables were informed by the text of the laws themselves, as well as policy elements developed by the National Complete Streets Coalition to describe a comprehensive Complete Streets policy. To ensure familiarity with legal terminology, two law students coded each of the statutes. An initial list of variables and definitions was reviewed and refined by the coders and supervising researcher through an iterative process. Laws were coded for a total of 35 variables within 16 discrete categories, including road user types, roadway development and preservation activities referenced, and provisions related to design standards, exceptions, and funding allocations. Each student coded the laws for all variables separately, blinded to the other’s results. To calculate interrater reliability, half of the coded statutes were randomly selected using the random number generator in Excel. Rates of divergence were recorded for all quantitative variables. Interrater agreement was calculated to be very high at 0.92, indicating strong interrater reliability. Divergences were reviewed by the supervising researcher and both coders; through several discussions, the divergences were ultimately resolved within the data set. Upon completing the coding process, the codebook was finalized and included variable names, definitions, values, labels, and notes.