Residential Parking Supply Has a Stronger Influence on Household Travel Choices Relative to a Neighbourhood’s Walkability and Access to Transit

Abstract: https://journals.sagepub.com/doi/full/10.1177/0042098021995139#abstract

Issue

A  large  volume  of  research  suggests  that  neighbourhood  attributes   such   as   public   transportation   access,   residential density, and walkability can have important impacts on   people’s   travel,   carbon   footprint,   and   employment   opportunities. More often than not, however, the research conducted on  this  topic  is  subject  to  “self-selection  bias.”  For example, people who prefer to walk and ride transit are likely to move to neighbourhoods that support these choices, while those who prefer to drive instead move to places with nearby highway access and ample parking. To overcome these biases, a research team at the University of  California,  Santa  Cruz  surveyed  successful  applicants  to San Francisco’s affordable housing lotteries about their transportation  choices  and  employment.  In  San  Francisco,  nearly  all  new  housing  developments  with  ten  or  more  residential  units  must  provide  affordable  housing  units at  below-market  rates.  The  units  are  normally  allocated  to  qualifying  households  through  city-run  lotteries.  The  chances  of  winning  a  lottery  are  small  —  less  than  two  percent  —  and  so,  not  surprisingly,  households  are  not  selective about which lotteries they enter. Analysis of lottery application  data  shows  that  applicants  do  not  appear  to  enter any particular lottery based on where they would end up living, or if the residence has ample parking, or based on the neighbourhood’s walkability or transit accessibility.

Key Research Findings

Residential parking supply is a major factor affecting car   ownership.   In   recent   years,   San   Francisco   made   significant   changes   to   its   parking   policy.   Instead   of requiring  new  buildings  to  provide  a  minimum  amount  of  parking  —  typically  one  parking  space  per  residential  unit  — San Francisco now caps the amount of parking in transit-friendly  neighbourhoods,  often  at  one  space  for  every  two  or four units. In buildings that did not have on-site parking, the survey  found  that  only  38  percent  of  lottery-allocated  households own a car. In buildings with at least one parking space  per  unit,  more  than  81  percent  of  households  own  automobiles  (see  Figure  1).  This finding  reveals  that households in buildings with ample parking are more likely to choose to own a vehicle. Transit  accessibility  affects  transit  use,  but  parking supply  has  a  stronger  influence.  Households  living  in  a  building  with  less  parking  and  in  a  neighbourhood  with  good  transit  accessibility  tend  to  use  public  transit  more.  The effect of parking on transit use is approximately three times as large as living in a neighbourhood with good transit access.   By   contrast,   increased   on-site   parking   reduces   transit  use  and  also  increases  the  amount  of  driving  by  a  similar  amount.  More  parking  also  discourages  walking  by  a  smaller,  but  still  statistically  significant,  amount.  In summary, cities  that walking should consider reducing the quantity of residential parking as well as plan for alternatives to the private car.

Non-work trips  are  more  sensitive  to  parking  supply  and   transit   accessibility   than   work   trips.   Households   with  abundant  residential  parking  tend  to  commute  more  by  private  car  (driving  alone  or  carpooling)  and  are  less  likely  to  commute  by  transit.  Greater transit  accessibility  has  the  opposite  effects.  However,  non-work  trips  are more  affected  by  these  dynamics  —  possibly  because commute  trips  are  more  constrained  by  the  availability  of  workplace  parking,  transit  proximity  to  the  workplace,  or  distances that preclude walking or biking. In contrast, non-commuting trips, such as going to the grocery store, entail more choices  of  potential  destinations,  and  so  residential  parking and neighbourhood walkability have greater effects on household travel behaviour. 

Better        public        transit        improves        employment        opportunities, but  parking  has  no  effect.  One  natural  concern about reducing residential parking is that it might limit  access  to  jobs,  but  no  evidence  was  found  that  this  trade-off  exists.  More  on-site  residential  parking  has  no detectable   impact   on   commute   length   or   employment   mobility  (a  measure  of  worker  flexibility  to  change  jobs). Indeed,   parking   spaces   intended   for   below-market-rate   units  often  go  unclaimed,  even  in  buildings  with  less  than  one  space  per  unit.  While winning  the  affordable  housing lottery is highly prized, households seem to care less about winning  the  “parking  lottery.”  Greater transit  accessibility,  in contrast, has a moderate positive effect on labour market outcomes,   leading   to   shorter   commutes   and   slightly   more  job  turnover.  However,  neither  transit  accessibility  nor  parking  ratios  affect  the  probability  of  a  household’s members  being  employed  full-time,  possibly  because  of  the  strong  economy  and  minimal  unemployment  in  San  Francisco at the time of the survey in 2019.

Neighbourhood design also affects walking and cycling. Increasing  a  neighbourhood’s  accessibility  by  walking  or  bicycling,  as  measured  by  its  Walk  Score  and  Bike  Score,  increases   the   frequency   that   household   members   use   the  corresponding  travel  mode,  even  after  controlling  for  household income and demographics and for the building’s parking ratio. 

Figure 1.  Residential parking  supply  affects  household  car  ownership decisions

Read the accompanying article on Urban Studies OnlineFirst here.