Monday, June 13, 2016

Does public transit reduce car travel externalities? Quasi-natural experiments' evidence from transit strikes

One of the unanswered questions in the field of urban economics is to which extent subsidies to public transit are justified. We examine one of the main benefits of public transit, a reduction in car congestion externalities, the so-called congestion relief benefit, using quasi-natural experimental data on citywide public transit strikes for Rotterdam, a city with mild congestion levels. On weekdays, a strike induces travel times to increase only marginally on the highway ring road (0.017 min/km) but substantially on inner city roads (0.224 min/km). During rush hour, the strike effect is much more pronounced. The congestion relief benefit of public transit is substantial, equivalent to about 80% of the public transit subsidy. We demonstrate that during weekends, travel time does not change noticeably due to strikes. Furthermore, we show that public transit strikes induce similar increases in number of cyclists as number of car travelers suggesting that bicycling-promoting policies to reduce car congestion externalities might be attractive in combination with first-best congestion pricing.
In Table 9, column 1, we provide the assumptions which allow us to calculate the external cost of congestion during strikes. In Rotterdam, cars trips have an average length of 15 km of which 62% are driven on inner city roads and the remaining 38% on highways.  Given the estimated travel time increase of 0.224 min/km on inner city roads (the weighted-average effect) and 0.017 min/km on highways, a strike induces an additional external cost of congestion of €607,052. The majority of this cost, €580,071 is on inner city roads, and an additional €26,981 on highways.
Based on the external costs of strikes, we aim to calculate the (long-term) congestion relief benefit of public transit. We will start from the idea that the beneficial effect of public transit provision is likely larger in the long-run than in those estimated using the short-run strike data. Hence, by equating the congestion benefit to the external congestion losses of strikes, and annualizing the congestion benefit (assuming 252 working days), we will provide underestimates of the long-run congestion relief benefit. Based on this calculation, the annual benefit is €153 million (see Table 9), about €128 per inhabitant. This excludes any benefits of public transit provision on weekends that we assume to be negligible. Given 721 million public transit passenger kilometers (OVPRO, 2014), the congestion reduction benefit per public transit kilometer is €0.21. This benefit is substantial given that the cost per public transit kilometer is €0.46.

The costs of providing public transit in Rotterdam are partially covered by subsidies, about €200 million per year, i.e. €0.28 per public transit kilometer.46 The congestion relief benefit is then about 77% of subsidies.47 It is useful to examine this result under different assumptions. For example, if we assume that the specification of complete strikes (reported in Table 5, column 4) is more indicative of the congestion relief benefit, then the benefit even is almost identical with current subsidies (see column 3). In contrast, if we make very conservative assumptions by assuming a trip length of 10 km, an equal split in distance traveled on highway and inner city roads, and that only inhabitants in the city of Rotterdam (and not the whole metropolitan area) are affected by the strike, then the congestion relief benefit is still 21% of the subsidy (column 2).48 These estimates indicate that the congestion relief benefit alone is substantial but possibly insufficient to justify the current supply of public transit in Rotterdam. Additional gains of public transit provision, such as economies of scale in public transit provision and productivity increases due to decreased car congestion might support current levels of subsidies (Graham, 2007). To do an overall welfare analysis of public transit provision is however beyond the scope of this paper.49
It is important to emphasize that there are reasons to believe that we have either overestimated or underestimated the public transit congestion relief benefit because we have equated the long-run effect of strikes to their short-run effect. First, we underestimate the long-term congestion relief benefit because during strikes about 20% of trips are canceled (see also PbIVVS, 1984 and van Exel and Rietveld, 2001). For longer periods without public transit, particularly for commuting, it is unlikely that so many trips are canceled and these trips will contribute to additional car congestion. Second, we underestimate the benefit, because for longer periods without public transit, current public transit travelers will increase car ownership and their use, so increasing car congestion. This effect is likely limited and has a clear upper bound. For example, in Dutch rural areas, where public transit is virtually absent, car ownership per household is only 30% higher than in urban areas (CBS, 2014). Note that car ownership in rural areas is also higher because of lower population and employment densities. Hence, a 20% increase in trips seems a more reasonable estimate. This is in line with studies which show that long-term demand elasticities of public transit are often substantially higher than in the short-term (Goodwin, 1992).  Third, we may overestimate the benefit by ignoring residential and workplace location decisions that are based on travel times (Kantor et al., 2014, Kok et al., 2014 and Johnson, 2014). Without public transit and higher levels of car congestion, households and firms would re-evaluate their location decision and may move closer to each other, hence reducing car travel. The size of this effect is unknown but seems to be small compared to the effects of canceled trips and increased car ownership because the location decision is a second-order effect conditional on decisions in the transport market.50 Hence, arguably, we most likely underestimate the long-term benefits of public transit.

by Martin W. Adler and Jos N. van Ommeren both of VU University Amsterdam, Spatial Economics, De Boelelaan 1105, 1081 HV, Amsterdam, The Netherlands
Journal of Urban Economics via Elsevier Science Direct
Volume 92, March 2016, Pages 106–119
Keywords: Transit subsidies; Public transit; Traffic congestion; Congestion relief benefit; Strike

No comments:

Post a Comment