How much can vehicle travel be reduced in California through ...HH, indicator variables for income class, urban area, and region. Instrumental variables (IVs): %of housing units built

How much can vehicle travel be reduced inCalifornia through land use policies?

David R. Heres

Postdoctoral Researcher

2010 Behavior, Energy & Climate Change ConferenceSacramento, November 17

Transportation GHG mitigation strategies towardsCalifornia’s 2020 target

Regional transportation GHG targets are ranked third among themitigation strategies within the transportation sector.

1. Vehicle GHG standards (18.8%)

2. Low carbon fuel standard (9.5%)

3. Regional transportation GHG targets (2%)

Measures being considered by MPOs to achieve theregional reduction targets

I Extend bicycle networksI Increase high occupancy lanesI Improve the extension and service of transitI Incentivize telecommutingI Promote the use of alternative modes of transportation and

travel reductionI Promote compact development - high density with mix of

land uses

Contrasting types of housing development

Average residential density in the city of San Francisco is almosttwice that from Los Angeles and 6 times that from a typicalCalifornian suburb.

Mean daily VMT per HH for different residential density,business-housing ratio, and transit density classes

15

17

19

21

23

25

27

29

31

33

35

1 2 3 4 5 6 7 8 9 10

MeanHHDailyVMT

Class

Residen2alDensity Mix "TransitDensity"

Previous estimates for VMT reductions responding toincreases in residential density in California

I CEC(2007): 20% to 40% VMT reduction

I Rodier (2008): 4% reduction in VMT in a 10-year horizon.Based on studies assessing the impacts of land-use changesand transit improvements

I Fang (2008): 3% VMT reduction in response to a doubling inresidential density

I Brownstone and Golob (2009): 12% VMT reduction inresponse to a doubling in residential density

Previous estimates for VMT reductions responding toincreases in residential density in California

I CEC(2007): 20% to 40% VMT reductionI Rodier (2008): 4% reduction in VMT in a 10-year horizon.

Based on studies assessing the impacts of land-use changesand transit improvements

I Fang (2008): 3% VMT reduction in response to a doubling inresidential density

I Brownstone and Golob (2009): 12% VMT reduction inresponse to a doubling in residential density

Previous estimates for VMT reductions responding toincreases in residential density in California

I CEC(2007): 20% to 40% VMT reductionI Rodier (2008): 4% reduction in VMT in a 10-year horizon.

Based on studies assessing the impacts of land-use changesand transit improvements

I Fang (2008): 3% VMT reduction in response to a doubling inresidential density

I Brownstone and Golob (2009): 12% VMT reduction inresponse to a doubling in residential density

Previous estimates for VMT reductions responding toincreases in residential density in California

I CEC(2007): 20% to 40% VMT reductionI Rodier (2008): 4% reduction in VMT in a 10-year horizon.

Based on studies assessing the impacts of land-use changesand transit improvements

I Fang (2008): 3% VMT reduction in response to a doubling inresidential density

I Brownstone and Golob (2009): 12% VMT reduction inresponse to a doubling in residential density

Non-linearity

Frequency distribution of HH daily VMT0

001000

1000

1000200020

0020003000

3000

3000400040

004000Frequency

Freq

uenc

yFrequency0

0

0100

100

100200

200

200300

300

300400

400

400500

500

500VMT

VMT

VMT

Residential self-selection

Residential self-selection

Elasticities for different specifications of the Modified TwoPart Model with Instrumental Variables

Dependent variable: Vehicle miles traveled(1) (2) (3) (4) (5) (6) (7)

residential density -0.1898* - -0.1934* -0.1399* - -0.1622* -

business-housing ratio - -0.4442* 0.0524 - -0.2118 0.1486 -

transit density - - - -0.08611 -0.1476* -0.0814 -0.1569*

Are IVs’ relevant? Yes Yes Yes Yes Yes Yes -Are IVs’ valid? Yes No Yes Yes No Yes -

Other control variables: age of older member in the HH, % of HH members that work , HH size, number of vehicles in theHH, indicator variables for income class, urban area, and region.Instrumental variables (IVs): % of housing units built before 1940, % of HHs where head’s race is other than white, % offamily HHs (all measured for the census tract where the HH is located)

Required increases on two alternative policies to achievea reduction of 4% in VMT

Median LO UPResidential density increase 24.6% 20.7% 28.6%Gasoline price increase 20.1% 13.6% 26.7%

Trajectory of the price of gasoline

100

150

200

250

300

350

400

450

500

Jan01

,200

7

Mar01,200

7

May01,200

7

Jul01,200

7

Sep01

,200

7

Nov01,200

7

Jan01

,200

8

Mar01,200

8

May01,200

8

Jul01,200

8

Sep01

,200

8

Nov01,200

8

Jan01

,200

9

Mar01,200

9

May01,200

9

centspe

rgallo

nofgasolineinCalifo

rnia

64%increase

62%decrease

73%increase

Conclusions

I Results from this study imply that, everything else equal,doubling residential density would reduce VMT by roughly20%

I The estimates indicate that the 4% reduction in VMT toachieve the GHG reductions from land-use policies inCalifornia would require increasing residential density byalmost 25%.

I On the other hand, estimates from studies looking at the effectof price variations on travel demand imply that a 20% increaseon top of the price of gasoline would suffice to reduce travelby those same amounts

I A combination of pricing and land use policies may deliver thebest results. However, political opposition and technicalconstraints, among other obstacles, will likely hinder timelyimplementation of either policy

Conclusions

I Results from this study imply that, everything else equal,doubling residential density would reduce VMT by roughly20%

I The estimates indicate that the 4% reduction in VMT toachieve the GHG reductions from land-use policies inCalifornia would require increasing residential density byalmost 25%.

I On the other hand, estimates from studies looking at the effectof price variations on travel demand imply that a 20% increaseon top of the price of gasoline would suffice to reduce travelby those same amounts

I A combination of pricing and land use policies may deliver thebest results. However, political opposition and technicalconstraints, among other obstacles, will likely hinder timelyimplementation of either policy

Conclusions

I Results from this study imply that, everything else equal,doubling residential density would reduce VMT by roughly20%

I The estimates indicate that the 4% reduction in VMT toachieve the GHG reductions from land-use policies inCalifornia would require increasing residential density byalmost 25%.

I On the other hand, estimates from studies looking at the effectof price variations on travel demand imply that a 20% increaseon top of the price of gasoline would suffice to reduce travelby those same amounts

I A combination of pricing and land use policies may deliver thebest results. However, political opposition and technicalconstraints, among other obstacles, will likely hinder timelyimplementation of either policy

Conclusions

I Results from this study imply that, everything else equal,doubling residential density would reduce VMT by roughly20%

I The estimates indicate that the 4% reduction in VMT toachieve the GHG reductions from land-use policies inCalifornia would require increasing residential density byalmost 25%.

I On the other hand, estimates from studies looking at the effectof price variations on travel demand imply that a 20% increaseon top of the price of gasoline would suffice to reduce travelby those same amounts

I A combination of pricing and land use policies may deliver thebest results. However, political opposition and technicalconstraints, among other obstacles, will likely hinder timelyimplementation of either policy

Thank you

David R. HeresBasque Centre for Climate ChangeBilbao, Bizkaia, [email protected]

Further reading: Heres & Niemeier (2011) in TransportationResearch Part B, 45(1): 150-161