Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Is Public Transit's 'Green' Reputation Deserved?

119 views

Published on

Justin Beaudoin, University of Washington Tacoma

Published in: Education
  • Be the first to comment

  • Be the first to like this

Is Public Transit's 'Green' Reputation Deserved?

  1. 1. Is Public Transit’s “Green” Reputation Deserved? Evaluating the Effects of Transit Supply on Air Quality Justin Beaudoin (with Cynthia Lin Lawell) University of Washington Tacoma Portland State University: Friday Transportation Seminar Beaudoin (UWT) Transit and Air Quality November 2, 2018 / 51
  2. 2. Motivation Transit advocated as a “sustainable” alternative to the car Reducing congestion Improving air quality Is there evidence to support these claims? Beaudoin (UWT) Transit and Air Quality November 2, 2018 1 / 51
  3. 3. Improving Air Quality 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Ambient Pollution (Mean Daily Maximum, 1991 = 1.00) ܱܵ2 ܱܰ2 ܲ‫01ܯ‬ ܱ3 ‫ܱܥ‬ Beaudoin (UWT) Transit and Air Quality November 2, 2018 2 / 51
  4. 4. Recent Transit Trends: Prima Facie Evidence? 0.8 0.9 1 1.1 1.2 1.3 1.4 1991 1995 1999 2003 2007 2011 Year Travel Time Index Transit Passenger-Miles Transit Service Frequency (vehicle revenue-miles per directional route-mile) Transit Network Size (directional route-miles) Beaudoin (UWT) Transit and Air Quality November 2, 2018 3 / 51
  5. 5. Motivation Sound Transit (ST3) in WA: initiative passed in Nov 2016 $54 billion in capital expenditures Plus additional operating subsidies ≈ $170 per capita increase in annual taxes Claim: ST3 will. . . ↓ auto VMT by 200-300m Help mitigate climate change Beaudoin (UWT) Transit and Air Quality November 2, 2018 4 / 51
  6. 6. “They also recognized the important role public transportation plays in addressing population growth, economic development, increased traffic congestion, and reducing pollution.” Beaudoin (UWT) Transit and Air Quality November 2, 2018 5 / 51
  7. 7. Research Questions Should public transit investment be increased as a means to address traffic congestion and air pollution? How effective have past public transit investments been in reducing congestion and improving air quality? Implications How we evaluate future transit investments (≈ $18 billion per year in U.S.) Beaudoin (UWT) Transit and Air Quality November 2, 2018 6 / 51
  8. 8. Recent Transit Trends: Prima Facie Evidence? -1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 -1 -0.5 0 0.5 1 1.5 2 %changepollution(1991-2011) % change transit supply (1991-2011) CO Beaudoin (UWT) Transit and Air Quality November 2, 2018 7 / 51
  9. 9. Recent Transit Trends: Prima Facie Evidence? -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 -1 -0.5 0 0.5 1 1.5 2 2.5 %changepollution(1991-2011) % change transit supply (1991-2011) NO2 Beaudoin (UWT) Transit and Air Quality November 2, 2018 8 / 51
  10. 10. Recent Transit Trends: Prima Facie Evidence? -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 -1 -0.5 0 0.5 1 1.5 2 2.5 %changepollution(1991-2011) % change transit supply (1991-2011) O3 Beaudoin (UWT) Transit and Air Quality November 2, 2018 9 / 51
  11. 11. Recent Transit Trends: Prima Facie Evidence? -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 -1 -0.5 0 0.5 1 1.5 2 2.5 %changepollution(1991-2011) % change transit supply (1991-2011) PM10 Beaudoin (UWT) Transit and Air Quality November 2, 2018 10 / 51
  12. 12. Recent Transit Trends: Prima Facie Evidence? -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 %changepollution(1999-2011) % change transit supply (1999-2011) PM2.5 Beaudoin (UWT) Transit and Air Quality November 2, 2018 11 / 51
  13. 13. Recent Transit Trends: Prima Facie Evidence? -1.5 -1 -0.5 0 0.5 1 1.5 -1 -0.5 0 0.5 1 1.5 2 %changepollution(1991-2011) % change transit supply (1991-2011) SO2 Beaudoin (UWT) Transit and Air Quality November 2, 2018 12 / 51
  14. 14. Literature Overview: Transit & Air Quality Many studies linking auto travel and pollution Interest in adverse health effects Uptick of recent studies linking public transit and pollution Chen and Whalley (2012) Bauernschuster, Hener and Rainer (2017) Rivers, Saberian, Schaufele (2017) No clear empirical consensus Beaudoin (UWT) Transit and Air Quality November 2, 2018 13 / 51
  15. 15. Transit Supply & Air Quality Link between transit supply and air quality depends on: 1 Modal distribution of vehicle-miles traveled (VMT) Cross-elasticity of auto and transit demand wrt transit supply ⇒ ≈ 4x greater than fare elasticity 2 Emission rates per VMT by mode 3 Spatial and temporal distribution of trips by mode Beaudoin (UWT) Transit and Air Quality November 2, 2018 14 / 51
  16. 16. Auto Externalities: Our Second-Best World AV $ *, congestion AV u AV AKAV * cτ AD AMPC congestion AMSC congestion emissions AMSC + *, congestion emissions AV + congestion externality * c eτ + emission externality Beaudoin (UWT) Transit and Air Quality November 2, 2018 15 / 51
  17. 17. Empirical Model Setup For pollutant p ∈ {CO, NO2, O3, PM10, PM2.5, SO2} in region r and year t : Air qualityprt = β1·Transit Capacityrt + β2·Freeway Capacityrt + β3·Arterial Road Capacityrt + β4·Fuel Costrt + β5·Transit Farert + β6·Trucking activityrt + β7·Employmentrt + β8·Incomert + β9·Populationrt + β10−11·Pollution Point Sourcesrt + β12−15·Weather Controlsrt + β16−17·NAAQS Standard Dummies + UZA and Census-Division Fixed Effects + εprt Travel volumes not included on RHS to allow for induced demand effect Beaudoin (UWT) Transit and Air Quality November 2, 2018 16 / 51
  18. 18. Pairwise correlation between pollutant concentrations, 1991-2011 CO NO2 O3 PM2.5 PM10 SO2 CO 1.000 - - - - - NO2 0.553 1.000 - - - - O3 0.009 0.253 1.000 - - - PM2.5 0.049 0.446 0.502 1.000 - - PM10 0.341 0.498 0.268 0.379 1.000 - SO2 0.318 0.334 0.128 0.538 0.174 1.000 Notes: CO and O3 are in units of parts per million (ppm). NO2 and SO2 are in units of parts per billion (ppb). PM2.5 and PM10 are in units of micrograms per cubic meter (µg/m 3 ). Beaudoin (UWT) Transit and Air Quality November 2, 2018 17 / 51
  19. 19. Identification and Potential Endogeneity Focus is on variation in air quality & transit supply within urban areas Using urban area fixed effects to control for time-invariant regional heterogeneity Potential endogeneity of transit investment 1 As policy measure to address existing congestion or environmental concerns 2 Component of growth/development strategy Beaudoin (UWT) Transit and Air Quality November 2, 2018 18 / 51
  20. 20. Instrument for Transit Investment Require: variable(s) correlated with transit capacity but uncorrelated with unobserved factors affecting congestion & air quality Instrument: Federal transit funding for capital expenses Excludes State and Local funds ( ≈ 67% of capital funding) Supported by 2009 GAO report Beaudoin (UWT) Transit and Air Quality November 2, 2018 19 / 51
  21. 21. Data Overview 96 ‘Urban Areas’ (UZAs) across the U.S. 44 states; 351 counties 1996 UZA-year observations (1991-2011) More of a regional focus than existing studies Considering intensive margin (more policy-relevant) Beaudoin (UWT) Transit and Air Quality November 2, 2018 20 / 51
  22. 22. UZAs Included Roadway Congestion, 2011 Beaudoin (UWT) Transit and Air Quality November 2, 2018 21 / 51
  23. 23. EPA Monitors: Portland-Vancouver-Hillsboro, OR-WA Urbanized Area Beaudoin (UWT) Transit and Air Quality November 2, 2018 22 / 51
  24. 24. EPA Air Quality Monitors CO NO2 O3 PM2.5 PM10 S02 Mean 2.76 3.29 6.97 5.99 4.10 2.83 Median 2 2 5 4 3 2 Minimum 1 1 1 1 1 1 Maximum 19 18 30 35 32 12 # of UZAs with ≥ 1 monitor for ≥ 2 years 91 82 96 96 94 88 Units of Measurement ppm ppb ppm µg/m 3 µg/m 3 ppb Notes: Each monitor also records the AQI for each pollutant. ppm: parts per million, daily maximum. ppb: parts per billion, daily maximum. µg/m 3 : micrograms per cubic meter, daily maximum. Beaudoin (UWT) Transit and Air Quality November 2, 2018 23 / 51
  25. 25. Data Sources Auto: congestion, capacity, travel, fuel Texas Transportation Institute Urban Mobility Report Federal Highway Administration (FHWA): Highway Statistics Transit: investment, ridership, fares/funding Federal Transit Administration: National Transit Database (NTD) Air Quality: ambient pollution levels Environmental Protection Agency (EPA) Weather: precipitation, temperature National Oceanic and Atmospheric Administration (NOAA) Socioeconomic: population, employment, income Bureau of Economic Analysis (BEA) Beaudoin (UWT) Transit and Air Quality November 2, 2018 24 / 51
  26. 26. Recent Study Beaudoin, Justin and C.-Y. Cynthia Lin Lawell (2018).“The effects of public transit supply on the demand for automobile travel,” Journal of Environmental Economics and Management, 88: 447-467. Beaudoin (UWT) Transit and Air Quality November 2, 2018 25 / 51
  27. 27. Spatial Heterogeneity: UZA Characteristics 0 .2 .4 .6 .8 Density 0 5 10 15 20 Population (millions) Population 0 .1 .2 .3 .4 .5 Density 0 2 4 6 8 10 Population (000s) per sq. mile Population Density 0 10 20 30 Density 0 .2 .4 .6 .8 % of transit VRM by FG modes % of FG Transit 0 .05 .1 .15 .2 .25 Density 0 10 20 30 40 Transit DRM per sq. mile Transit Accessibility 0 20 40 60 80 100 Density 0 .01 .02 .03 .04 .05 Transit VRM per capita Transit Capacity 0 10 20 30 40 50 Density 0 .1 .2 .3 Transit PMT per auto VMT Transit Use Beaudoin (UWT) Transit and Air Quality November 2, 2018 26 / 51
  28. 28. Spatial Heterogeneity: Population Density −.4 −.3 −.2 −.1 0 1 2 3 4 5 Population Density (’000 per sq. mile) Elasticity 95% Confidence Interval Beaudoin (UWT) Transit and Air Quality November 2, 2018 27 / 51
  29. 29. Cross-Elasticity: Induced Demand −.1 −.05 0 .05 .1 Elasticityofautotravel withrespecttotransitcapacity t t−1 t−2 t−3 t−4 t−5 t−6 t−7 t−8 t−9 t−10 Population, Income & Employment: Lagged 0 to 10 years Beaudoin (UWT) Transit and Air Quality November 2, 2018 28 / 51
  30. 30. Overview of Results: Congestion Empirically, transit investment does help alleviate congestion On average, 10% ↑ transit capacity ⇒ 0.8% ↓ congestion However, congestion-reduction effect dependent upon: Population size and density of region Characteristics and technology of public transit network The timing of the change and role of induced/latent demand Elasticity range: -0.02 to -0.3 Beaudoin (UWT) Transit and Air Quality November 2, 2018 29 / 51
  31. 31. Empirical Model Setup For pollutant p ∈ {CO, NO2, O3, PM10, PM2.5, SO2} in region r and year t : Air qualityprt = β1·Transit Capacityrt + β2·Freeway Capacityrt + β3·Arterial Road Capacityrt + β4·Fuel Costrt + β5·Transit Farert + β6·Trucking activityrt + β7·Employmentrt + β8·Incomert + β9·Populationrt + β10−11·Pollution Point Sourcesrt + β12−15·Weather Controlsrt + β16−17·NAAQS Standard Dummies + UZA and Census-Division Fixed Effects + εprt Beaudoin (UWT) Transit and Air Quality November 2, 2018 30 / 51
  32. 32. CO Emission Share, On-Road Sources (2011) 33.9% Emissions, Million Tons (2011) 27.4 Short-run elasticity - (slightly insig.) Medium-run elasticity - (slightly insig.) Long-run elasticity - (slightly insig.) Some evidence that transit may modestly reduce CO Beaudoin (UWT) Transit and Air Quality November 2, 2018 31 / 51
  33. 33. NOx Emission Share, On-Road Sources (2011) 38.0% Emissions, Million Tons (2011) 5.9 Short-run elasticity + (slightly insig.) Medium-run elasticity + (slightly insig.) Long-run elasticity + Some evidence that transit may modestly increase NOx ; with CO result, consistent with some cross-modal substitution Beaudoin (UWT) Transit and Air Quality November 2, 2018 32 / 51
  34. 34. O3 Emission Share, On-Road Sources (2011) 4.5% Emissions, Million Tons (2011) 2.6 Short-run elasticity - (quite insig.) Medium-run elasticity + (quite insig.) Long-run elasticity + (quite insig.) Transit has no effect on O3 Beaudoin (UWT) Transit and Air Quality November 2, 2018 33 / 51
  35. 35. PM2.5 Emission Share, On-Road Sources (2011) 3.2% Emissions, Million Tons (2011) 0.2 Short-run elasticity + (slightly insig.) Medium-run elasticity + Long-run elasticity + Transit appears to increase PM2.5 Beaudoin (UWT) Transit and Air Quality November 2, 2018 34 / 51
  36. 36. PM10 Emission Share, On-Road Sources (2011) 1.8% Emissions, Million Tons (2011) 0.4 Short-run elasticity + (slightly insig.) Medium-run elasticity + Long-run elasticity + Transit appears to increase PM10 Beaudoin (UWT) Transit and Air Quality November 2, 2018 35 / 51
  37. 37. SO2 Emission Share, On-Road Sources (2011) 0.5% Emissions, Million Tons (2011) 0.03 Social cost per ton ? Short-run elasticity + (very insig.) Medium-run elasticity + (very insig.) Long-run elasticity + (very insig.) Transit has no effect on SO2 Beaudoin (UWT) Transit and Air Quality November 2, 2018 36 / 51
  38. 38. Interpreting Results Are the effects (statistically) zero? What is the economic significance? Appears to be masking heterogeneity: In areas with: More FG transit (particularly long-established rail networks), High existing transit accessibility, and High existing transit ridership, Additional transit supply: Decreases CO, and Lessens the increase in NOx and PM, relative to other regions. Beaudoin (UWT) Transit and Air Quality November 2, 2018 37 / 51
  39. 39. Transit Technology: Mixed Traffic Bus Very low cross-elasticity & higher marginal pollution per rider (?) Beaudoin (UWT) Transit and Air Quality November 2, 2018 38 / 51
  40. 40. Transit Technology: Fixed Guideway Commuter rail Light rail Heavy rail Higher cross-elasticity & lower marginal pollution per rider (?) Beaudoin (UWT) Transit and Air Quality November 2, 2018 39 / 51
  41. 41. Transit Supply Trends From 1991-2011, % of VRM by FG increased from 34.5% to 39.2%: - 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 14,000,000 16,000,000 18,000,000 20,000,000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Transit Supply: Vehicle revenue-miles (1991 - 2011) Commuter rail Light rail Heavy rail Bus 11.8% 60.7% 3.5% 24.0% Beaudoin (UWT) Transit and Air Quality November 2, 2018 40 / 51
  42. 42. Transit Ridership Trends From 1991-2011, % of PMT on FG increased from 52.1% to 61.8%: - 50,000,000 100,000,000 150,000,000 200,000,000 250,000,000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Transit Ridership: Passenger-miles traveled (1991 - 2011) Commuter rail Light rail Heavy rail Bus 34.7% 38.0% 4.6% 22.7% Beaudoin (UWT) Transit and Air Quality November 2, 2018 41 / 51
  43. 43. Changing Transit Technology & Fleet Composition? No direct effect found by: 1 Treating FG and MT transit capacity separately 2 Analyzing 1991-2001 and 2001-2011 in separate sub-samples Beaudoin (UWT) Transit and Air Quality November 2, 2018 42 / 51
  44. 44. Future/Ongoing Work Extend dataset from 2011 to 2014 Analyze the data at the monitor level Explore spatial heterogeneity in more detail Beaudoin (UWT) Transit and Air Quality November 2, 2018 43 / 51
  45. 45. Transit’s Effect on Accessibility Accessibility = Mobility x Proximity Transportation (congested travel) Location & land use (uncongested travel) Beaudoin (UWT) Transit and Air Quality November 2, 2018 44 / 51
  46. 46. Transit’s Effect on Accessibility Location Choices Amenities Infrastructure Congestion Urban Form Proximity Uncongested Mobility Accessibility Travel Choices Source: adapted from lecture by Gilles Duranton at the 2018 Canadian Economics Association annual meeting (6/2/2018, McGill University) Beaudoin (UWT) Transit and Air Quality November 2, 2018 45 / 51
  47. 47. Transit’s Effect on Accessibility $53 million BRT line (“The Vine”) 44,787 transactions in Clark County from 2012-2018 Beaudoin (UWT) Transit and Air Quality November 2, 2018 46 / 51
  48. 48. Transit’s Effect on Accessibility Beaudoin (UWT) Transit and Air Quality November 2, 2018 47 / 51
  49. 49. Transit’s Effect on Accessibility Beaudoin (UWT) Transit and Air Quality November 2, 2018 48 / 51
  50. 50. Transit’s Effect on Accessibility Beaudoin (UWT) Transit and Air Quality November 2, 2018 49 / 51
  51. 51. % change in property values due to Vine opening in Jan 2017 Walk Time Lower Bound Mean Upper Bound 0 - 10 minutes 8.5% 10.7% 12.9% 10 - 15 minutes 5.2% 7.1% 9.0% Driving Distance Lower Bound Mean Upper Bound 0 - 0.4 miles 3.0% 5.0% 7.1% 0.4 - 0.6 miles 8.7% 11.5% 14.4% 0.6 - 0.8 miles 7.0% 9.1% 11.2% Beaudoin (UWT) Transit and Air Quality November 2, 2018 50 / 51
  52. 52. Preliminary Conclusions Public transit has the potential to reduce congestion in some regions Less likely that public transit improves air quality (and may make it worse!), but there may be exceptions How does the story change if proper regulations are in place? Transit does lead to localized accessibility/livability benefits Adjust CBA and political debate accordingly Beaudoin (UWT) Transit and Air Quality November 2, 2018 51 / 51
  53. 53. Thank You Justin Beaudoin jbea@uw.edu Beaudoin (UWT) Transit and Air Quality November 2, 2018 51 / 51

×