13.10.2013 Views

The effects of road pricing - Feweb - Vrije Universiteit Amsterdam

The effects of road pricing - Feweb - Vrije Universiteit Amsterdam

The effects of road pricing - Feweb - Vrije Universiteit Amsterdam

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

3.3.5 Vehicle occupancy<br />

Whereas congestion <strong>pricing</strong> and the derivation <strong>of</strong> optimal tolls have received ample attention,<br />

the issue <strong>of</strong> carpooling has largely been neglected in the literature. For instance, Daganzo<br />

(1982) made a first attempt to incorporate carpooling in equilibrium traffic assignment in<br />

general networks, but the demand for carpools is assumed to be fixed. This means that the<br />

model fails to include the mode choice behaviour <strong>of</strong> commuters. Yang and Huang (1999),<br />

however, use a deterministic equilibrium mode choice model with overall elastic demand.<br />

<strong>The</strong>ir analysis concerns the carpooling and congestion <strong>pricing</strong> problem in a multi-lane<br />

highway with HOV (high-occupancy vehicle) lanes. Numerical examples show the efficiency<br />

<strong>of</strong> HOV lanes, first- and second-best <strong>pricing</strong> and their combinations. Despite the fact that the<br />

model is highly simplified, the findings may be relevant for the design and operations <strong>of</strong><br />

combined congestion <strong>pricing</strong> and HOV lanes in practice, such as in California. <strong>The</strong>y found<br />

that marginal cost <strong>pricing</strong> provides the most efficient policy option under either elastic or<br />

inelastic demand. Reserving one or more entire lanes for carpooling vehicles rarely reduces<br />

total costs and nor does it improve social welfare, given marginal cost prices. Nevertheless,<br />

conventional carpool lanes alone can be justified at reasonably high congestion levels.<br />

Differentiation <strong>of</strong> tolls across commuting modes should be cautiously implemented. <strong>The</strong><br />

optimal uniform second-best toll is a weighted sum <strong>of</strong> the marginal external congestion costs<br />

between non-carpooling and carpooling commuters.<br />

3.3.6 Location choices<br />

<strong>The</strong> <strong>effects</strong> <strong>of</strong> <strong>road</strong> <strong>pricing</strong> on transport and location patterns have been much discussed.<br />

Simulation models for a typical city can be used to examine the impacts <strong>of</strong> <strong>road</strong> <strong>pricing</strong> on<br />

transport and location patterns. <strong>The</strong> model city should then be calibrated to replicate typical<br />

observed location and travel patterns. Model parameters, elasticities, functional forms, and so<br />

on can be taken from land use and transport models estimated on observed data. This should<br />

ensure that the conclusions are also applicable in reality. <strong>The</strong> first studies on this issue date<br />

back to the 1970s when Solow and Vickrey (1971) look at traffic patterns, congestion costs,<br />

and land use. <strong>The</strong>y show that, without congestion tolls, the market value <strong>of</strong> land may be a<br />

poor guide to land-use decisions, and lead to an overallocation <strong>of</strong> land to transportation.<br />

Solow (1973) confirms this result by using a simulation model <strong>of</strong> urban activity including the<br />

congestion costs <strong>of</strong> transportation. He shows that rents nearer the city are too low to properly<br />

reflect the social worth <strong>of</strong> transportation costs in the absence <strong>of</strong> congestion tolls. This will<br />

result in more land to be taken for use as <strong>road</strong>s than is socially desirable. It remains unknown<br />

what the consequences are for location patterns. Another topical contribution from this period<br />

compares a model <strong>of</strong> equilibrium land use patterns in the residential ring <strong>of</strong> an urban area with<br />

an optimal one (Oron et al., 1973). <strong>The</strong>y show that scarce resources can be saved by<br />

households moving from the periphery to more central locations in a distorted situation with<br />

suboptimal congestion tolls. Proper <strong>pricing</strong> <strong>of</strong> transportation will most probably lead to less<br />

suburbanisation.<br />

More recently, Eliasson and Mattsson (2001) have developed a simulation model <strong>of</strong> a generic<br />

city to study the <strong>effects</strong> <strong>of</strong> general congestion <strong>pricing</strong> and more specific toll ring <strong>pricing</strong>. <strong>The</strong><br />

city, consisting <strong>of</strong> homogenous zones, contains four types <strong>of</strong> actors: households, employers,<br />

shops, and service establishments. <strong>The</strong> location <strong>of</strong> households is determined by their value <strong>of</strong><br />

plot size and accessibility to workplace, shops and service establishments. Firms locate only<br />

according to accessibility to households. Shops and service establishments locate according to<br />

the accessibility to households with this difference – that they evaluate the expected number<br />

<strong>of</strong> customers in each zone. <strong>The</strong>y also take into account the proximity to working places and<br />

the size <strong>of</strong> the plot. It would go too far to discuss all other assumptions <strong>of</strong> the model at this<br />

point; rather we focus on the outcomes. It appears that the location <strong>effects</strong> <strong>of</strong> congestion<br />

38

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!