Road Capacity and Allocation of TimeRoad Capacity and Allocation of Time

David M LevinsonUniversity of Minnesota

Seshasai KanchiICF Consulting

81st Transportation Research Board Meeting

16th January 2002Levinson, David and Seshasai Kanchi (2002) Road Capacity and Allocation of Time, Journal of Transportation and Statistics 5(1) pp 25-46.http://nexus.umn.edu/Papers/RoadCapacity.pdf

OutlineOutline

Introduction Data Travel Time and Activity Duration Analysis Theory of Daily Time Budgets Methodology Results Conclusions

IntroductionIntroduction

Travel and Activity are Two Sides of the Same Coin

Time in Travel = f (Time at Activity, # Trips) Primary Activities Considered: Home, Work,

Shop and Other Daily Activity Budget (24 hrs)

DataData

1990/91 and 1995/96 Nationwide Personal Transportation Survey

1990 and 1995 Federal Highway Administration Highway Statistics

Individuals whose total activities did not add to 1440 minutes (24 hrs), excluded

This study looks only at adults, 18-65 years of age Excluded travelers with a daily shopping time

greater than 420 minutes

Activity Duration CalculationActivity Duration Calculation

PERSON ID ORIGIN DEST TT START TIME END TIME* TIME SPENT*1 H O 15 8:30 8:45 301 O W 15 9:15 9:30 3601 W O 15 15:30 15:45 1051 O O 10 17:30 17:40 201 O H 10 18:00 18:20 8502 H W 20 8:00 8:20 3402 W O 15 14:00 14:15

Travel time and Activity durationTravel time and Activity durationComparison of 1990 and 1995 NPTSComparison of 1990 and 1995 NPTS

* Denotes Significance at 95% level

SEX_YEAR WORK STATUS TRAVEL CASESHOME WORK SHOP OTHER

Female1995 Non Worker 1172* 0 42* 166* 60 28121990 1220 0 35 127 58 10201995 Worker 944 313* 25* 93* 65 80141990 928 284 30 132 65 2535Male1995 Non Worker 1171* 0 30 177* 62 12401990 1222 0 29 130 59 3611995 Worker 900 365* 15* 90* 70 77161990 903 338 20 110 69 2275

TIME SPENT AT

Comparison of Travel and Activity Comparison of Travel and Activity Patterns of 1990 and 1995 NPTSPatterns of 1990 and 1995 NPTS

Time Spent at Home Decreased for Non Workers and Female Workers

Time at Home in 1990 Substituted for Work in 1995, especially for Female Workers

Time Spent at Other Declined for Workers but Increased for Non-Workers

Overall Travel Times have either remained Stable or Increased, but not significant

Induced TravelInduced Travel

S1: Supply before

S2: Supply after

Price of Travel

Quantity of Travel (VMT)

Q1 Q2

P1

P2

Demand

How does Highway Expansion affect How does Highway Expansion affect Travel and Activity PatternsTravel and Activity Patterns

Makes network becomes faster, higher attainable speeds lead to time savings in travel

Increases accessibilityBroadens commuter’s travel choicesMore non-travel activitiesIndividuals maximize their utility

Time Spent at Activities Decreases

S1 S2

U

+ve

AS1 AS2

TS1

TS2

Time Spent

Traveling Decreases

Travel Time (T)

Activity Duration (A)

1440 minutes

1440 minutes

Utility Increases with Expansion

Fixed Daily Time Budget

“Demand” Line

Travel and Activity Duration Production FunctionTravel and Activity Duration Production Function

Model EstimationModel Estimation

Seemingly Unrelated Regression Estimation (SURE) is used

They use asymptotically efficient, feasible generalized least squares estimation

It overrules the assumption of OLS that error residuals are not interrelated

Description of VariablesDescription of Variables

T90i Time spent at activity "i" in 1990

i Index of activities (travel to and duration at home, work, shop and other)

A Age

D Local population Density

G Gender

H Household Income levels

L Family lifecycle characteristics

M Month of year interview was conducted

S State specific variables

W Day of week interview was conducted

MethodologyMethodology

• Since the NPTS was not conducted as a panel survey, we first estimate a model of 1990 individuals, and then apply that model to 1995 individuals in the form

T90i = f(A,D,G,H,L,M,S,W)• Estimated Travel Behavior for 1990 individuals can be determined in the form

^

T90i = f(A,D,G,H,L,M,S,W)95

• Then we estimated a difference model of change in travel behavior between the 1995 individuals their best estimate of 1990 behavior

ΔΧ

Difference in lane miles for all roadway types between 1995 and 1990

C90 Sum of lane miles for all roadway type in 1990

Δ

Φ Δ ι φ φ ε ρ ε ν χ ε ι ν σ τ α τ ε α ϖ ε ρ α γ ε Φ υ ε λ π ρ ι χ ε σ β ε τ ω ε ε ν 1 9 9 5 α ν δ 1 9 9 0

F90 State average Fuel price 1990

Δ

Ι Δ ι φ φ ε ρ ε ν χ ε ι ν σ τ α τ ε α ϖ ε ρ α γ ε π ε ρ χ α π ι τ α Ι ν χ ο μ ε β ε τ ω ε ε ν 1 9 9 5 α ν δ 1 9 9 0

I90 State level per capita Income in 1990

Δ

Π Δ ι φ φ ε ρ ε ν χ ε ι ν σ τ α τ ε Π ο π υ λ α τ ι ο ν β ε τ ω ε ε ν 1 9 9 5 α ν δ 1 9 9 0

P90 State Population in 1990

D95 Local population Density estimates in 1995

G95 Individual’s sex in 1995

L95 Family Lifecycle characteristics in 1995

Δ iT = 95iT −^

T90i change in time at activity "i" between 1995 and 1990 (estimated)

Δ iT = f(ΔC /C90,ΔF /F90,ΔI / I90,ΔP /P90,D95,G95,L95)

s.t. Δ iTi=1

8

∑ =0

Difference ModelDifference Model

Hypothesis for WorkersHypothesis for Workers

TRAVEL TO TIME SPENT AT

H W S O

- - - -

TimeSavings

From Travel

FasterNetwork

TimeSavings

From Travel

H W SO

- -+ +

ReducedPeak

Spreading

Big BoxStores

PleasureOriented

TimeSavings

From Travel

Hypothesis for Non WorkersHypothesis for Non Workers

TRAVEL TO TIME SPENT AT

H S O

+ + -

More # of Home-Shop

Trips

FasterNetwork

H SO

++ -

LessDiscretionary

PleasureOriented

Elasticity of Time with respect to CapacityElasticity of Time with respect to Capacity

* Indicates significance at 95% confidence level

Dependent variableChange in Minutes MinutesTravel time to

Home -3.17E-04 -1.08E-02 1.48E-02 * 5.28E-01Work -7.06E-03 * -1.23E-01 - -Shop -4.71E-02 * -1.90E-01 3.39E-02 * 2.35E-01Other -9.80E-03 * -1.60E-01 -2.91E-02 * -6.06E-01Time Spent atHome 7.27E-03 * 6.56E+00 2.19E-03 * 2.60E+00Work -1.80E-02 * -5.66E+00 - -Shop -3.44E-02 * -7.67E-01 2.54E-02 * 1.19E+00Other 2.72E-03 3.49E-01 -2.83E-02 * -3.95E+00

Non-workers Workers

%elasticities %elasticities

ResultsResults

Due to Highway Capacity Expansion Workers spend more Time at Home and

Other, Less Time at Work and Shop Non-Workers spend more Time at Home

and Other, Less Time at Other Non-workers take more Home to Shop trips

ConclusionsConclusions

Overall Travel Times have remained stable while Activity Durations changed significantly

Increase in highway capacity has a small but significant impact on individual’s activity and travel patterns

Effect on Workers and Non-Workers are different

Questions?

Time Budgets and Induced Demand: How Time Budgets and Induced Demand: How Access Affects ActivityAccess Affects Activity

By

David M Levinson

Seshasai Kanchi

University of Minnesota

CTS Research Conference

25th May 2000

Time Budgets and Induced Demand: How Time Budgets and Induced Demand: How Access Affects ActivityAccess Affects Activity

By

David M Levinson

Seshasai Kanchi

University of Minnesota

Symposium on Induced Traffic Research

University of California,Berkeley

June 8-9, 2000

Time Budgets and Induced Demand: How Time Budgets and Induced Demand: How Access Affects ActivityAccess Affects Activity

By

David M Levinson

Seshasai Kanchi

University of Minnesota

9th International Association For

Travel Behavior Research Conference

Gold Coast, Queensland, Australia

July 2-7, 2000

Whence Induced Demand: How Access Whence Induced Demand: How Access Affects ActivityAffects Activity

By

David M Levinson

Seshasai Kanchi

University of Minnesota

Western Regional Science Association

Palm Springs California

February 2001