Background
Walking is important in cities for reducing automobile use, and promoting public health through increased physical activity. It is critical not only as a stand-alone mode, but also as the gateway mode to public transit. But retrofitting an expansive auto-oriented region into a multimodal one that facilitates walking travel is challenging. Trends in walking travel as it relates to transit supply and density at the metropolitan scale have been seldom studied and remain poorly understood.

Objectives
This research seeks to address previous gaps by investigating walking trends within the greater Los Angeles region over the past decade, and exploring the determinants of change including sociodemographic factors and transit availability. This study also uses a more nuanced definition of walking trips compared to previous studies by separating walking as a primary mode from walking trips to/from transit stops/stations.

Methods
We use household travel survey information from the 2001 regional travel survey conducted by the Southern California Association of Governments (SCAG) and the 2009 National Household Travel Survey (NHTS) to analyze trends in walking travel across 46 Regional Statistical Areas (RSAs) spread over five counties within the greater Los Angeles region. These travel survey data were merged with 2000 and 2010 U.S. Census data at the census tract level, corresponding with the years of the survey, in addition to transit data from SCAG. Dependent variables of interest include walking trip share and walking trip rate. Independent variables were selected based on findings from previous research, data availability and appropriateness at the subregional scale, and include the following variables aggregated at the RSA level: 1) population density; 2) employment density; 3) transit stop density; 4) proportion of households with children; 5) median household income.   Descriptive analysis of changes in aggregate walking trip shares and rates (with and without considering access/egress trips to transit as observations) were analyzed at the RSA level, and at the county level over the 2001-2009 period. To establish the direction of causality between independent variables and walking trips, we employed a fixed-effects regressions approach to investigate potential determinants of observed changes in our walking travel measures.

Results
The results show that walking has increased in the greater Los Angeles region from 2001 to 2009; statistically significant increases are observed in about 90 percent of the RSAs. Walking travel shares and rates were highest in the urbanized central and western Los Angeles basin. Additionally, increases in walking trips during this period generally correspond with increases in population, employment, and transit service, which were highest in Los Angeles County. Estimates from fixed-effects regression analysis generally suggest a positive association between population density and walking; RSAs that experienced greater increases in population density were associated with greater corresponding increases in non-transit related walking trips.    Households with children were associated with lower corresponding increases in walking trips, possibly due to increased errand trips by car. However, the transit-walking connection is unclear, except that transit investments have possibly been endogenously associated with ridership, and therefore areas with higher increases in transit stop density have experienced significantly higher increases in walk trips to and from transit stops and stations.

Conclusions
This study provides compelling evidence that residents of the greater Los Angeles region are taking a greater share of trips by walking relative to other modes; further, they are taking more walking trips per person on average in 2009 compared to 2001. In this traditionally auto-oriented region, the increase in walking trips during the past decade may represent a turning point.  While more detailed studies are needed to flesh out the precise, micro-scale factors that might most increase walking, the walking renaissance is real, and planners should continue to support and leverage pedestrian travel to promote public health and improve the environment.

Implications
These results have important policy implications for planning at the regional and metropolitan level. The Los Angeles region is undergoing a spatial transformation as they have embarked on a massive transit expansion program. These projects are often driven by planners and policy makers who embrace the notion that transit generates walking and creates walkable communities.  However, our evidence suggests that transit only has the potential to generate more walking trips through new access/egress trips in the short term; it can possibly promote overall walking in the longer term, if complemented with coordinated land use policies that improve the walking environment through increased urban densities, mixed land uses, and pedestrian-oriented design.