Bridge Tolls Advocacy Project

Authors' Note

We are, respectively, an economist and a traffic engineer. We have advocated tolling New York City's East River bridges for three decades -- as policy analysts in the Lindsay Administration, and subsequently as private citizens.

In this report, we estimate the time savings drivers can expect to result from East River bridge tolls. To do so, we had to envision how bridge tolls will change long-standing travel patterns, and estimate the resulting changes in trip durations -- not just for bridge crossings but also for trips on streets and highways affected by bridge traffic.

Ordinarily, this kind of analysis relies on a sophisticated computer model, in which estimates of the value of time and other factors are used to "select" travelers' modes and routes; the resulting demand levels determine average travel speeds, and these in turn determine travel choices, with the process iterated until equilibrium is reached. And in fact state and city agencies actually do have such a "Best Practices Model," the product of years of labor and over $10 million in consultant costs. Yet they have refused to apply it to bridge tolls (see Section 9).

Norbert Weiner, the founding father of cybernetics, once mused that if computers had been around for the Manhattan Project, everyone would have insisted afterwards that the atomic bomb couldn't have been developed without them (we tip our hat to Neil Postman for this story). Encouraged by this thought, we decided to model the effects of bridge tolls without the $10 million software, instead employing maps, spreadsheets and our years-in-the-making knowledge of the complexities of NYC transportation.

The result is this report. Our primary finding is that the time savings will total some 16.3 million vehicle-hours -- or 37.5 million person-hours -- worth $650 million a year to drivers and passengers. We believe that this conclusion is well-founded. In fact, we've probably erred on the side of conservatism. For one thing, we've assumed that half of the trips that are "tolled off" the bridges are lured back by the improved traffic flow. We also assumed flat-rate tolling rather than a "value pricing" toll varying by time of day. (Modeling value pricing wasn't possible within the time available; but we believe the additional time savings would be substantial.)

To make our methodology transparent without burdening the text, we are making our computer spreadsheet available to other researchers (see p. 19); we urge readers to follow the text along with the spreadsheet. We welcome criticisms, comments and suggestions.

As we went to press, there were reports that Mayor Bloomberg was abandoning efforts to seek East River bridge tolls. We hope this report will help revive such efforts -- if not this year, then next. As we document here, the benefits of tolling include improved mobility as well as expanded municipal revenue and services -- benefits that our great but sorely pressed city can't afford to pass up.

-- Charles Komanoff
-- Brian Ketcham, P.E.
July, 2003

1. Summary of Findings

Tolling the East River bridges will induce drivers to shift some trips to less heavily-used MTA toll facilities and to forego some other auto trips altogether. This rearrangement and reduction in driving will improve traffic flow in some of the most congested parts of New York City, saving drivers considerable time with a significant monetary value.

Note: All findings in this report assume flat-rate tolling rather than tolls varying by time of day. "Value pricing" would add to the time savings from East River bridge tolls, perhaps significantly.

Key Finding #1: By reducing traffic volumes, bridge tolls will raise motor vehicle speeds significantly on the bridge spans -- by an estimated 24% on the Williamsburg Bridge, 29% on the Manhattan Bridge, 41% on the Brooklyn Bridge, and 47% on the Queensboro Bridge. (See Table 1.) Overall, bridge tolls will shorten typical travel times significantly for trips using the East River bridges -- by 6 minutes for round trips on the Queensboro Bridge (yielding a 6.5% improvement in the total door-to-door travel time), by 5½ minutes for round trips on the Williamsburg Bridge (a 5.2% improvement), and by 4½ minutes on the Manhattan and Brooklyn Bridges (around 5% improvements in both cases).

Key Finding #2: By increasing average vehicle speeds on connecting roads as well as on the bridge spans, East River bridge tolls will do away with more than 9% of the idle time that motorists, truckers and bus riders now lose in traffic tie-ups throughout New York City. The estimated 16,300,000 "vehicle-hours" saved annually will be realized as follows: 3.6 million vehicle-hours on Queens streets and highways (22% of the total); 5.9 million vehicle-hours in Brooklyn (36%); 3.4 million vehicle-hours in Manhattan (21%); and 3.3 million vehicle-hours on the bridge spans themselves (20%). (To convert to "person-hours," multiply vehicle-hours by 2.3; note that these figures take into account the anticipated lengthening of travel times on the three tolled MTA East River spans due to the "resettlement" of some East River bridge trips to those crossings.)

Key Finding #3: Half of the aggregate time savings from bridge tolls will be realized on trips that aren't even using an East River bridge, due to the "ripple effects" from reduced or redirected bridge traffic. Indeed, since fewer bridge-crossing trips will be made, and those that remain will be more evenly distributed, an indirect but powerful effect of bridge tolls will be to improve travel speeds across large areas of the city through which East River bridge traffic moves. Speeds will increase by an average of 1.5% through much of Queens and 2-3% across most of Brooklyn. In Manhattan, the tolls will speed up travel by an average of 0.5% in a 2-mile arc extending from the Queensboro Bridge, and by an average of 2.7% in a similar arc around the three East River bridges in lower Manhattan. (See precise estimates in Table 2.)

Key Finding #4: Based on an average value of $40 for each vehicle-hour (a weighted average of a range spanning $7.50 an hour for solo off-peak car trips to several hundred dollars an hour for 18-wheelers and commuter buses), the 16.3 million "saved vehicle-hours" equate to $650 million a year, enough to offset over 90% of the assumed $700 million out-of-pocket costs to drivers that the tolls will impose.

Key Finding #5: The 15% anticipated increase in use of the MTA crossings (the Brooklyn-Battery and Queens Midtown Tunnels, and the Manhattan leg of the Triborough Bridge), some 38,000 trips per day, should boost MTA toll revenues by $50 million a year, a windfall for the agency's transit and commuter-rail service.

These estimates reflect three important assumptions:

They assume that all four East River bridges -- the Brooklyn, Manhattan, Williamsburg and Queensboro Bridges -- are tolled at the same rates as the MTA's East River crossings (the Brooklyn-Battery Tunnel, the Queens Midtown Tunnel, and the Triborough Bridge). Lower toll rates would reduce the speed gains and time savings.
They assume flat-rate tolling rather than variable or "value" pricing in which higher rates are charged during peak-usage periods with lower rates at other times. It is virtually certain that variable pricing would compound the speed improvements and time savings to the point that the aggregate time savings to New York City drivers and car and bus passengers from East River tolls would significantly exceed the costs of the tolls.
They assume that one-half (50%) of the trip reductions that would be expected based on price-sensitivity factors will be eroded by new trips "attracted" by the easing of traffic; absent this "induced travel," the speed gains and time savings would be 80% greater than the above results.

Variations on these and other assumptions are discussed in Section 6.

A downside of East River bridge tolls is that speeds will decline on the three MTA crossings that will absorb some of the traffic now using the free bridges. Nevertheless, these speed reductions will cancel out only a fraction of the speed gains on the newly-tolled bridges. Traffic flow will improve overall, with an average 11.6% increase in driving speeds on all four spans linking Brooklyn with Manhattan, and a 5.5% increase on the three spans connecting Manhattan with Queens. Similarly, although trip times for the MTA crossings will lengthen by an average of 1.8 minutes per round trip, those crossings carry only half as much traffic as the East River bridges, on which the average round trip will be shortened by more than 5 minutes; the aggregate time gain on the East River bridges will therefore swamp the total loss on the MTA crossings by more than 3-to-1.

The other direct cost of tolling is the loss to drivers and passengers of the "utility" or value of the trips -- some 34,000 a day, by our estimates -- that will be "tolled off the road," i.e., that will switch to transit or ridesharing or will be foregone altogether. This aggregate cost is actually quite small, just $25 million a year, an estimate that derives from the observation that the net value of each such trip must be under $3.50 (else the addition of a $3.50 toll wouldn't dissuade the driver from continuing to make the trip). Using an average net value of $2 per trip yields $25 million annually for all eliminated trips, a relatively small sum compared to the overall toll revenues and time savings.

2. Introduction

This report, the second on East River tolls by the Bridge Tolls Advocacy Project (BTAP), examines the time savings that motorists and others can expect to reap from the changes in traffic flow that will result from tolling New York City's East River bridges.

We find that East River bridge tolls will save drivers each year an estimated 16,300,000 "vehicle-hours" (or 37.5 million "person-hours") they now lose in traffic tie-ups on the East River crossings and connecting highways and roads. In a single stroke, tolls will do away with more than 9% of the idle time that truckers, car drivers, car passengers and bus passengers now lose annually in traffic jams throughout New York City. If each vehicle-hour is valued at $40 (an estimate derived in Section 8), then the time savings from tolling the East River bridges amount to $650 million a year, offsetting more than 90% of the $700 million out-of-pocket cost to drivers of the tolls.

The Hours follows BTAP's March 2003 report on the "incidence" of bridge tolls on motorists, Who Will Really Pay That report established that daily East River bridge commuters constitute a very small fraction (2%) of adult New Yorkers and are predominantly higher-income workers who can better afford to pay. The other 98% of city residents age 18 to 80 will, on average, pay less than $50 a year each in East River tolls. Who Will Really Pay also placed bridge tolls in a broader fiscal context, and showed that their aggregate impact on residents of Brooklyn and Queens will be less than the costs already being borne by Manhattanites as a result of higher real estate taxes. (Who Will Really Pay may be downloaded from BTAP's Web site at http://www.bridgetolls.org /whowillpay/.)

With our findings here on the time savings, we can highlight the costs and benefits of tolling the Brooklyn, Manhattan, Williamsburg and Queensboro Bridges, as follows:

East River Bridge Tolls: Costs and Benefits to NYC

As the right-hand column shows, a key benefit of East River tolls is the almost $700 million in annual net revenue they will contribute to maintain the bridges and help finance transit and other municipal services. The sum of that benefit and the estimated $650 million in anticipated motorist time savings, is roughly double the direct cost to drivers of the tolls. Other benefits and costs, such as the improved quality of life, on the one hand, and the lost multiplier effect due to the trips that are foregone or shifted to other modes, are less tangible and are harder to estimate. Still, it seems highly likely that the positives from eliminating 9% of New York City's total traffic delays would far outweigh any negatives associated with eliminating or mode-shifting less than 1% of the city's total vehicular travel.

Tolling the East River bridges -- or, more accurately, re-tolling them, since all four crossings were built and operated as tolled facilities until 1911 -- can be expected to affect travel in two major ways:

Some toll opponents have raised the specter of congestion at East River bridge toll plazas. But this is a non-issue; high-speed electronic toll collection via in-car E-ZPass transponders read from overhead gantries has done away with the need for toll plazas. (Drivers without E-ZPass will purchase per-use E-Z "cards" from third parties, or re-route to an MTA crossing, all of which have cash toll booths.) Note too that much or all of the modest increases in times to cross the MTA facilities could be offset by converting them to high-speed collection in which vehicles are tolled at highway speeds.

More than half-a-million motor vehicles each day crossed the free East River bridges (the Brooklyn, Manhattan, Williamsburg and Queensboro Bridges) in 1999 and 2000, the last "normal" traffic years before September 2001. These trips accounted for 8% of all vehicle-miles traveled in New York City, although the mileage registered on the bridge spans themselves was less than one-fifth of that amount, or around 1.5% of the city's VMT. After meticulously analyzing travel patterns on these crossings, as well as those on the MTA's Triborough Bridge and Brooklyn-Battery and Queens Midtown Tunnels, we have reached the following conclusions:

Table 1: Bridge span vehicle volumes and speeds, before and after bridge tolls

East River crossings are shown in bold. Other crossings are MTA bridges and tunnels. Daily volumes are 1999-2000 averages. Speed data are annual averages (24/7) and are weighted by volumes to reflect a typical trip rather than the speed for an hour picked at random.

All speeds are annual averages (24/7) and are weighted by volumes on local roads, arterials and expressways to reflect a typical trip rather than the speed for an hour picked at random. Brooklyn crossings are the Williamsburg, Manhattan and Brooklyn Bridges and the Brooklyn-Battery Tunnel. Queens crossings are the Queensboro and Triborough Bridges (Manhattan Plaza) and Queens Midtown Tunnel. Near borough is Brooklyn or Queens area within 3 or 4 miles (respectively) of the East River bridges. Further borough is the 8-mile area in each borough, excluding the near zone. Manhattan is area in that borough within 2 miles of the bridges. See map, p. 14. Gain may not equal difference between before and after due to rounding.

See Table 1 for definitions of crossings and borough zones, and map on p. 14. Sums may not agree with totals, or percents with 100, due to rounding. "All borough" denotes the sum of the "near" and "further" borough zones, and encompasses 85% of all vehicle-miles traveled in Brooklyn, and 63% in Queens. To convert to person-hours, multiply vehicle-hours by 2.3.

4. Tolls and Trips

We have estimated that bridge tolls will have these effects on trips on the East River bridges:

"Elimination I and II" columns are, respectively, without and with "bounceback" -- the return of some old trips, or attraction of new ones, due to faster travel times resulting from reduced traffic volumes. Figures in the second column, shown in bold, were used to calculate the traffic, speed and time impacts in this report. Totals in last row are weighted averages reflecting varying volumes on the bridges.

The precise figures are shown in Table 4. These estimates underlie our calculations of speed improvements and time savings from bridge tolls.

Most of our bullet assumptions along with the numbers in Table 4 were derived through economic modeling, as follows:

Trip Resettlements: We estimated the volume of current trips on the (free) East River bridges that will resettle to the (tolled) MTA crossings by calculating the "excess" increase in usage of the free bridges that has built up over the past 30 years, as MTA tolls have been raised a half-dozen times while the East River bridges have remained untolled. From January 5, 1972, when tolls on the MTA crossings were doubled (after more than two decades with no increase), through 2000, daily crossings in either direction on all seven East River bridges and tunnels grew by an average of more than 5,500 vehicle trips each year. Based on the pre-1972 split in traffic between the MTA crossings and the free bridges, around 40% of the annual increase should have been realized on the MTA facilities; the actual average share has been only 20%, indicating an annual shortfall of 1,300 trips per day.

By 2000, or some 30 years later, the cumulative "diverted" traffic had reached 38,000 trips per day, a phenomenon we attribute to the cost differential between the two sets of crossings. We expect that when the East River bridges are tolled, this diverted traffic will resettle to its "natural" routes using an MTA bridge or tunnel. These 38,000 daily trips amount to 7.4% of current trips on the four East River bridges, and 5.0% of trips on all seven East River crossings including the three MTA facilities.

Trip Eliminations: In a more detailed analysis, we examined changes over the past 25 years (1975-2000) in East River vehicle volumes, New York City employment, and the out-of-pocket cost to drive into Manhattan. By sifting these data, we were able to infer the price elasticity of vehicle trips crossing the East River, to be around negative 0.4. That is, holding other factors constant, a 1% increase in the cost to drive across the East River into Manhattan should reduce the volume of such trips by around 0.4%. In the parlance of economists, a price-elasticity of (negative) 0.4 denotes a usage that is not highly elastic, i.e., it is relatively resistant to minor changes in price (as compared to elasticities of, say, one or greater, which correspond to products or services for which demand is very price-sensitive). This accords well with a common-sense assessment of drivers' unwillingness to forego their cars.

However, the toll will add significantly to the current cost to drive across a free bridge. Even with a relatively low price-elasticity, the effect on demand will be noticeable. We estimate that a typical crossing on a free bridge now costs the driver over $8, a figure that takes into account the wide range of parking costs from zero (for on-street or employer-paid parking) to expensive parking in midtown lots and garages. Based on the estimated negative 0.4 price elasticity, a 43% increase in this cost, via the addition of a $3.50 toll, would be expected to eliminate 13.2% of all such trips now being taken. (The 13.2% result is less than the simple product of 43% and 0.4 because an exponential rather than linear calculation is required to translate elasticities into changes in demand; our assumptions and reasoning, along with the equations themselves, are provided in the companion spreadsheet to this report.) This "gross" estimate was then converted to a net figure to reflect what we call "bounceback."

Bounceback: The one phenomenon for which we could not derive an estimate is the extent to which some trips that disappear -- as well as other trips that are not now made because drivers feel they would take too long -- will be lured back by the smoother traffic stream. (This is precisely the effect that computer-based regional traffic models are suited to estimate; indeed, it is one of the key rationales for the region's having invested $10 million in developing the Best Practices Model.) But "bounceback" is no less real for our inability to derive its magnitude; the expected time to accomplish a trip (or, conversely, the time the driver anticipates sitting in stop-and-go traffic) is a powerful determinant of the volume of automobile travel. In lieu of a procedure for estimating it, we have conservatively specified a bounceback rate of one-half, i.e., we assume that 50% of the trips that will be tolled off the road will be re-attracted by the improved traffic flow.

With that assumption, the gross 13.2% reduction in trips on the four East River bridges becomes a 6.6% net decrease, equating to approximately 38,000 "disappeared" trips a day. Taking the three MTA East River crossings into account, the share of all East River crossings (on the seven bridges and tunnels) predicted to be eliminated due to tolls on the East River bridges is 4.5%.

That result, which underlies our findings on speed improvements and travel-time savings, jibes with the New York City Department of Transportation's lone detailed analysis of bridge tolls, published a quarter-century ago (Traffic Impact of Tolls on the East & Harlem River Bridges, April 1977). A team directed by then-Deputy Commissioner Samuel I. Schwartz -- the transportation engineer and consultant popularly known as "Gridlock Sam" -- estimated that tolls on the city's East and Harlem River bridges would cause a 4.0% drop in daily traffic volumes across the East River (6.9% on the Harlem River bridges), a figure that agrees closely with our 4.5% estimate.

5. Traffic Volumes, Travel Speeds, and Travel Times

To estimate time savings from East River bridge tolls, we employed the standard sequence employed by travel-demand models such as NYMTC's Best Practices Model (see Section 9):

All calculations were disaggregated by three types of roadways: (i) limited-access highways or expressways, such as the BQE or FDR Drive; (ii) wide "arterial" streets or avenues such as Atlantic Avenue in Brooklyn, or most Manhattan avenues; and (iii) local streets.

Table 5 shows average speeds for each road type for the three boroughs that are the subject of our analysis.

Speeds were estimated by Brian Ketcham, based on NYMTC data provided at various times over the past decade in documents required for air quality conformity analyses. They are intended to reflect typical trips rather than speeds for an hour picked at random.

Since speeds vary within each borough, with slower traffic in neighborhoods closest to the bridges -- we subdivided Brooklyn and Queens travel into a "near bridge" component (within 3 miles of the Brooklyn crossings, and 4 miles of the Queensboro), and a "further" component extending up to 8 miles from the bridges (see map overleaf). Based on Census data for 2000, we estimated that 25% of westbound trips destined for the Brooklyn crossings originate in the "near" component, and the remaining 75% from the farther component; for the Queensboro Bridge, we used a split of 37% of westbound trips originating from the near zone, and 63