Reinventing the Urban Interstate: A New Paradigm for Multimodal Corridors (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

79 This appendix presents evidence suggesting that transit and freeways can coexist and thrive in the same corridor. These findings also provide evidence to support the concepts and tools of the new paradigm, including separated corridor travel markets that can be achieved through complementary multi- modal coordination, transit-oriented land uses and station access, constrained freeway capacity and, where appropriate, high transit operating speeds. Total Corridor Performance: How Well Do Transit and Freeways Work Together? Although some may believe that freeways and transit do not mix, analysis of existing multimodal corridors sug- gests this is not always true. There are examples of transit lines that thrive in the same corridors as freeways. These corridors have varying combinations of characteristics that can help the transit line compete effectively for patronage. They are • Multimodal corridor coordination • Transit-oriented corridor urban form • Transit-oriented station access • High transit operating speeds (where appropriate) • Constrained freeway capacity Although there are several ways to evaluate the patronage performance of multimodal corridors, the total patronage of both the transit and freeway facilities gives an indication of how well these two modes are working together as a multimodal system to facilitate travel along the corridor. However, focusing on total throughput can mask cases where one mode dominates the other—specifically, when freeways capture most of the corridor travel market. Therefore, while our discussion of multimodal corridor performance begins by looking at total patronage, we follow this by looking at transit ridership in each corridor to evaluate how well each transit line competes with its freeway neighbor. Transit’s share of total corridor patronage is a useful metric to see how well transit competes with the freeway. Table B-1 shows the estimated total, freeway-only, transit- only, and transit mode share of patronage in each multimodal corridor studied. These data are used to evaluate how well transit lines perform in multimodal corridors, whether and how transit and freeways can work together, and what corridor conditions help foster success for all modes. Multimodal Corridor Coordination As discussed in previous chapters, coordination between the various transportation facilities in a corridor can be achieved by complementary or supplementary coordination. In complementary coordination, the transit and freeway facilities are designed and operated to serve different travel markets, activity patterns, and land uses within the same corridor. A corridor with supplementary coordination has roughly equal station and interchange spacings. These corridors put their freeway and transit components in direct competition with each other for the same travel markets. Two complementary coordination configurations were also proposed in previous chapters: • Transit-oriented complementary coordination has long interchange spacings on its freeway component and relatively short station spacings on its transit line. • Automobile-oriented complementary coordination has long station spacings on its transit facility and relatively short interchange spacings on its freeway component. There are few real-world examples of transit-oriented complementary multimodal corridors, but there are cases where sections of corridors have transit-oriented characteristics. A P P E N D I X B Evidence on the Patronage Impacts of Multimodal Corridors

80 The Benefits of Complementarity Complementary corridors have several distinct advantages over supplementary corridors: • Separated Corridor Travel Markets for Transit and Free- way: The combination of long interchange and short sta- tion spacings (transit-oriented complementary) encour- ages short- and medium-distance corridor travelers to use transit while long-distance travelers are encouraged to use the freeway. The combination of short interchange and long station spacings (automobile-oriented complementary) en- courages long-distance corridor travelers to use transit while short- and medium-distance travelers are encouraged to use the freeway. In both configurations, direct competition between the transit and freeway facilities is minimized. • Competitive Transit Operating Speeds: When there are fewer transit access points (stations) along congested cor- ridors, transit can operate at higher average speeds and compete favorably with automobile trips in travel time and travel-time reliability. • Increased Local Access for Transit and Increased Free- way Speeds: Providing fewer freeway access points allows for increased freeway speeds, higher flow rates, and higher volumes. This can be supported by providing tran- sit alternatives for local trips, especially in or near more densely developed areas. • Fewer Station/Interchange Conflicts: By offsetting transit stations from freeway interchanges it is possible to increase and diversify the “customer” base for travel along a given corridor. • Enhanced Potential for Transit-Oriented Development: When interchanges and stations are separated, the auto- mobile traffic associated with interchanges is removed from transit station walking environments, allowing clustered, high-density, pedestrian-oriented development patterns to take root. The Effects of Multimodal Coordination on Corridor Patronage The performance of a corridor can be understood in many ways. For the purposes of this analysis it is not important to establish results in terms of return on investment or the relative performance of transit versus auto. Rather, a simple aggregate measure of person-trip throughput provides an adequate indicator of corridor performance. Based on a review of existing multimodal facilities in the United States, corridors with complementary coordination tend to carry more total patrons. Most corridors that carry more passengers either have a combination of long station spacings and short interchange spacings, although in one case it is the opposite. Multimodal Corridor Freeway Transit* Total %Transit 1 Atlanta North-South Line/Route 400 HR T 1 0 3 26,000 22,000 348,00 0 6 % 2 Chicago Blue Line/Eisenhower Expwy. HR T 8 255,00 0 2 4,000 2 79,000 9% 3 Chicago Blue Line/Kennedy Expwy. (I-90) HR T 6 400,00 0 5 9,000 4 59,000 13 % 4 Chicago Red Line/Dan Ryan Expwy. HR T 6 312,00 0 4 2,000 3 54,000 12 % 5 Denver Central/I-25 LRT 6 2 70,000 18,000 288,00 0 6 % 6 Denver TREX/I-25 LR T 6 270,00 0 2 3,000 2 93,000 8% 7 Houston Northwest/U.S. 290 BR T 6 316,00 0 6 ,000 322,00 0 2 % 8 Los Angeles El Monte Transitway/I-10 BR T 8 287,00 0 7 ,000 294,00 0 2 % 9 Los Angeles Gold Line/I-210 LRT 6 2 42,000 24,000 266,00 0 9 % 10 Los Angeles Green Line/Century Freeway LR T 1 0 3 11,000 42,000 353,00 0 1 2% 11 Los Angeles Harbor Freeway (I-110)/Harbor Transitway BR T 6 387,00 0 4 ,000 391,00 0 1 % 12 New Haven Line/I-95 CR 6 1 63,000 87,000 250,00 0 3 5% 13 Portland MAX Airport/I-84 Red Line LR T 6 195,00 0 7 ,000 202,00 0 3 % 14 Sacramento North Line/S.R. 160 & I-80 LR T 6 70,000 6,00 0 7 6,00 0 8 % 15 San Francisco Daly City Line/I-280 HR T 8 254,00 0 5 1,000 3 05,000 17 % 16 San Francisco (BART) Dublin Line/I-580 HR T 8 257,00 0 2 0,000 2 77,000 7% 17 San Francisco (BART) Pittsburgh/Bay Point Line/S.R. 24 HR T 8 204,00 0 5 7,000 2 61,000 22 % 18 San Jose Guadalupe/San Jose S.R. 87 & 85 LR T 6 182,00 0 7 ,000 189,00 0 4 % 19 Washington D.C. Orange Line/I-66 HR T 6 127,00 0 1 39,000 266,00 0 5 2% Average 7 2 61,400 33,750 295,15 0 1 2% Source: TCRP H-36 Interim Report March 2009 * - Transit daily patronage estimated using daily boardings. Note: Transit patronage figures were typically available for entire lines and have been adjusted to represent travel within the study corridors (which are often portions of larger lines). Estimated Daily Patronage Transit Mode Freeway Lanes Corridor ID Table B-1. List of study multimodal corridors and key performance measures.

81 Measuring Multimodal Coordination The following formula was used to construct a measure of multimodal corridor coordination for the study corridors: The higher the calculated value for a corridor; the more complementary the freeway and transit services in the corridor, while the lower the value, the more supplementary the corridor. By taking the absolute value of this calculation, this measure does not distinguish between complementary corridors where transit provides area coverage and the freeway emphasizes operating speeds, and complementary corridors with the reverse configuration. Figure B-1 provides a graph of multimodal coordination and total corridor patronage (daily freeway patrons plus daily transit boardings) for each of our study corridors. A linear regression line drawn on this graph indicates that if there is a statistically valid relationship between these variables, it is not linear. Multimodal Corridor Coordination Median Int = erchange Spacing Median Station Spacing− Additional exploratory analysis of these data showed that the relationship between multimodal coordination and corridor patronage is not linear. A log-log model was fitted and graphed in Figure B-2. By graphing the relationship between multimodal coordi- nation and total corridor patronage (Figure B-2) a positive relationship is suggested (though not statistically proven due to an insufficient sample size) where complementary corridor coordination is associated with more total corridor patronage. More detailed multivariate linear regression results are pre- sented in Table B-2. The coefficient for multimodal coordi- nation score in predicting throughput was significant at the p = 0.05 level. To further test this relationship, a series of additional regressions were performed to determine to what extent the re- lationship is driven by either sensitivity to interchange spacing or sensitivity to transit station spacing irrespective of comple- mentary multimodal access. For example, no statistically signif- icant correlation was identified for either the influence of inter- change spacing on freeway throughput without transit or the relative influence of transit station spacing on transit ridership. Note: Commuter rail cases (i.e.; the New Haven Line/I-95 corridor) have been excluded since they tend to attract automobile and bus access riders from further distances from their stations than other transit modes. Transit-Optimized/Freeway Constrained cases (i.e.; Chicago Blue Line/Kennedy Expwy. (I-90), Washington D.C. Orange Line/I-66, and San Francisco East Bay (BART) Pittsburgh/Bay Point Line/S.R. 24) were also excluded since their freeway capacity constraints give their transit lines an operational advantage that masks the benefits of complementary coordination. Sacramento’s North Line/S.R. 160 & I-80 was also excluded since the freeway sample txt point was along S.R. 160 where volumes are low. Figure B-1. Multimodal coordination and total corridor patronage— linear regression line.

82 Note: Commuter rail cases (i.e.; the New Haven Line/I-95 corridor) have been excluded since they tend to attract automobile and bus access riders from further distances from their stations than other transit modes. Transit-Optimized/Freeway Constrained cases (i.e.; Chicago Blue Line/Kennedy Expwy. (I-90), Washington D.C. Orange Line/I-66, and San Francisco East Bay (BART) Pittsburgh/Bay Point Line/S.R. 24) were also excluded since their freeway capacity constraints give their transit lines an operational advantage that masks the benefits of complementary coordination. Sacramento’s North Line/S.R. 160 & I-80 was also excluded since the freeway sample txt point was along S.R. 160 where volumes are low. Figure B-2. Complementary multimodal coordination is associated with improved corridor performance—log-log transformation. B Std. Error (Constant) 10.046 1.12E+00 8.94 *** Natural Log of Multimodal Coordination 0.152 4.81E-02 3.16 ** Park-&-Ride Spaces per Station 0.000 8.89E-05 -2.63 * Average Ramps Touching Down w/in 1/4-Mile of Stations 0.102 6.30E-02 1.62 Total Freeway Lanes 0.031 2.70E-02 1.15 Heavy Rail Dummy (0=No, 1=Yes) -0.008 9.51E-02 -0.08 Housing Unit Density w/in 1/2-Mile of Stations 0.000 4.62E-05 -0.93 Natural Log of CBD Size (Sq. Ft. Office) 0.136 5.98E-02 2.28 * Notes: R-Square = 0.56 F-Sig. = 0.03 N = 16 *** = p < 0.01 ** = p < 0.05 * = p < 0.10 Coefficients t-stat. Sig. Table B-2. Log-linear regression model results predicting total corridor patronage (freeway & transit).

83 While the planning and design of multimodal facilities is more complicated than the planning or design of either transit or automobile facilities in isolation, the potential benefits of doing so suggest that both can and should be planned and designed in a coordinated and mutually beneficial fashion. This analysis suggests that multimodal coordination may be an important factor in planning successful new paradigm corridors. However, the lack of data and consequent inability to perform a statistically valid analysis means that this concept requires further study. Can Transit Thrive in Multimodal Corridors? While it seems obvious that transit and freeways tend to conflict with each other’s operations, there is no evidence that they have to. The success of one does not mean the other must suffer. Figure B-3 shows the estimated daily patronage for each multimodal corridor studied, for both the freeway and tran- sit facility components. The cases in this figure are sorted with decreasing freeway patronage estimates from left to right. If transit patronage success always came at the expense of freeway patronage, then we would expect to see increasing transit patronage as freeway patronage decreases. But while we see cases with large transit ridership values—cases such as the Washington DC Orange Line, the Chicago Blue Line/ Kennedy Expressway, and San Francisco’s Pittsburg/Bay Point Line corridors—these cases do not have consistently lower freeway patronage levels. If transit ridership always suppressed freeway patronage, we would expect that corridors with low transit ridership would have consistently high levels of freeway patronage. This too, is not the case, since the San Jose Guadalupe Line, the Portland MAX Red Line, and the Sacramento North Line all have very low transit ridership and low-to-moderate freeway patronage. These findings suggest that the performance of transit and freeways in multimodal corridors is not a zero-sum game, where only one mode thrives, not both. Other dynamics might also be at work, other ways that transit and freeways might be affecting each other when sharing a corridor. In most of the United States, the automobile is the dominant mode of travel. This could mean that while transit does not take patrons from freeways, freeways may prevent nearby transit lines from thriving. Figure B-4, where corridors are sorted by freeway patronage descending from left to right, suggests this is not the case. If it were impossible for transit to successfully attract riders in a multimodal corridor, we would expect to see the lowest transit ridership cases on the left (where freeway patronage is the highest) and the highest transit ridership cases clustered on the right of Figure B-4. Since cases with high freeway patronage appear on the left, we would expect to see low transit patronage cases on the left as well, with high transit pa- tronage cases clustered to the right of the graph where the cases have low freeway patronage. This is not the case. Instead, high transit ridership corridors appear to be spread evenly through- out the graph, without reference to the patronage of their adjacent freeway facilities. 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 O ra n ge Li n e N ew H av en Li n e K en . B lu e Li ne B A R T Pi tts . L in e B A R T D al y Ci ty Ch ic . R ed L in e LA G re en Li ne Ei s. B lu e Li ne LA G ol d Li ne D en v er T - R EX A tl. N . - S. L in e B A R T D u b. L in e D en . Ce n . Li n e El M o n te Tr a n si tw a y Po rt . R ed L in e S. J.G ua da lu pe H ou . N W Sa c. N o rt h Li n e H ar bo r Tr a n si tw a y 19 12 3 17 15 10 694 2 161 5 13 188 7 14 11 Multimodal Corridor T o ta l D ai ly C o rr id o r Pa tr on a ge (T ra n sit + Fr ee w a y) Transit Freeway Figure B-3. Transit success does not always mean low freeway patronage in multimodal corridor.

84 More important, there are several cases where transit is high—both in absolute terms and in comparison to the neighboring freeway facilities. Figure B-5 shows the estimated daily transit ridership for the study multimodal corridors. Three of the four corridors with the highest transit rider- ship share some key characteristics. Washington DC’s Orange Line/I-66 corridor has the best-performing transit line (in terms of ridership) of any multimodal corridor evaluated for this study. The second, third, and fourth best- performing transit lines are the New Haven (commuter rail) line, Chicago’s Blue Line (Kennedy), and San Fran- cisco’s BART Pittsburg/Bay Point Line. These findings confirm expectations that high-capacity and high-speed transit lines attract more patronage, even in multimodal corridors. Clearly, freeways do not always make a corridor inhospitable to transit. Other factors that determine the success of each facility at attracting patrons must be at work. 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 K en . B lu e Li ne H ar bo r Tr a n si tw a y A tl. N . - S. L in e H ou . N W Ch ic . R ed L in e LA G re en L in e El M o n te Tr an si tw a y D en . Ce n. Li n e D en v er T- R EX B A R T D u b. Li ne Ei s. B lu e Li n e B A R T D al y Ci ty LA G ol d Li ne B A R T Pi tts . L in e Po rt . R ed L in e S. J.G u a da lu pe N ew H av en Li n e O ra ng e Li n e Sa c. N o rt h Li n e 3 1 411 7 10 58 6 216 15 9 17 13 18 12 19 14 Multimodal Corridor To ta l D a ily C or ri do r Pa tr on ag e (T ra ns it + Fr ee w a y) Transit Freeway Figure B-4. Freeway success does not always mean low transit patronage. 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 O ra ng e L in e N ew H av en Li n e K en . B lu e Li ne B A R T Pi tts . L in e B A R T D al y Ci ty Ch ic . R ed Li n e LA G re en Li n e Ei s. B lu e Li n e LA G o ld Li n e D en v er T- R EX A tl. N . - S. L in e B A R T D ub . Li n e D en . Ce n . Li n e El M o n te Tr a n si tw a y Po rt. R ed L in e S. J.G u a da lu pe H ou . N W Sa c. N o rth Li ne H ar bo r T ra n si tw a y 19 12 3 17 15 4 10 692 16 81 5 13 18 147 11 Multimodal Corridor D a ily Tr a n si t P at ro n a ge Figure B-5. Transit line ridership in multimodal corridors.

85 The Effects of Transit Mode on Transit Ridership The operating characteristics of the transit line can play an important role in determining transit ridership. For the sake of brevity and ease of analysis, the type of transit mode in each study corridor was used as a proxy to suggest their operating characteristics. Therefore, in general it was assumed (as dis- cussed in Chapter 5) that heavy rail has the highest carrying capacities and operating speeds, followed by commuter rail, light rail, and BRT. Figure B-6 confirms this point. The best-performing cases in terms of transit ridership are heavy rail transit (HRT), while the lowest-ridership cases are bus rapid transit (BRT) and light rail transit (LRT). Five of the top six transit ridership cases are HRT, while two of the bottom five are LRT and the other three are BRT. These differences are partially due to the operating character- istics of the various transit modes (see discussion in Chapter 5 for further details). LRT vehicles run singly or in short trains on tracks in various right-of-way environments, including mixed-flow surface streets, dedicated lanes with grade cross- ings, and fully grade-separated dedicated facilities.1 Therefore, depending on the design of the right-of-way (grade-separated or mixed-flow), fare collection systems, station platforms, and station spacings, light rail systems can approach heavy rail performance in terms of capacity and operating speeds. The flexible performance parameters of LRT can be seen in several cases, where light rail lines attract riders at simi- lar levels to heavy rail. Three cases stand out in this regard. The Los Angeles Green Line, Denver’s T-REX, and the Los Angeles Gold Line all attract between 23,000 and 42,000 week- day boardings within the multimodal corridor sections of each line. BRT is often seen as a low-cost alternative to more capital- intensive fixed-rail alternatives. One of the most important feature of BRT (unlike regular bus service) is that it runs on a dedicated, exclusive lane of travel, giving it a high level of service reliability (since it does not compete for right-of-way with other modes) and speed. When running in mixed-flow traffic, bus priority technologies (such as signal prioritiza- tion) are often used to improve travel times and provide a competitive edge to BRT vis-à-vis other modes in the corri- dor. Off-bus fare collections as well as platform boarding and alighting are frequently used to reduce dwell times at stops.1 In addition to operational improvements, the cost of a BRT system can be about one-third that of a light rail system.2 This makes BRT cost-feasible for somewhat less dense and smaller central business district corridors than more capital-intensive rail systems. Consequently, BRT systems are often used in the United States as an alternative to more expensive fixed-rail options and are typically deployed in corridors where these other op- tions are infeasible. Therefore, although BRT has proven ca- pable of performing at levels equal to fixed-rail in other coun- tries, the locations where it has been implemented in the United States have tended to limit its success at attracting rid- ers at levels equal to fixed-rail alternatives. 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 O ra n ge L in e N ew H av en Li n e K en . B lu e Li ne B A R T Pi tts . Li n e B A R T D al y Ci ty Ch ic . R e d Li n e LA G re en Li ne LA G ol d Li ne Ei s. B lu e Li ne D en v er T- R EX A tl. N .-S . Li n e B A R T D u b. Li ne D en . Ce n. Li n e Po rt. R e d Li n e El M o n te Tr a n si tw a y S. J.G u a da lu pe H ou . N W Sa c. N or th Li n e H ar bo r Tr a n si tw a y HRT HRT HRT HRT HRT HRT HRTLRT HRT HRT BRT BRT BRTLRT LRT LRT LRT LRTCR Multimodal Corridor D a ily T ra n si t P at ro n a ge HRT = Heavy Rail Transit LRT = Light Rail Transit CR = Commuter Rail BRT = Bus Rapid Transit Figure B-6. Transit patronage and transit mode. 1Pushkarev, B. and J. Zupan, 1971. Public Transportation and Land Use Policy. Don Mills, Ontario: Indiana University Press. 2Leal, Monica T. & Robert L. Bertini, Bus Rapid Transit: An Alternative For Developing Countries, http://web.pdx.edu/∼bertini/brt.pdf

86 These limitations are manifest in the patronage rankings of BRT multimodal corridors shown in Figure B-6. Three of the four-lowest ridership cases are BRT systems—Los Angeles’s El Monte Transitway and Harbor Transitway, and Houston’s Northwest/U.S. 290 corridor. In these three cases, BRT does not run in its own right-of-way, but shares HOV lanes with automobiles. Therefore, when traffic congestion slows traffic in the HOV lane, BRT suffers as well and cannot offer a travel time premium compared to the freeway. It should also be mentioned that the operational character- istics of each transit mode are not the only factors that deter- mine performance in a multimodal corridor. For example, HRT not only offers speed and capacity advantages (and thus time competitiveness with the automobile-freeway system), but in most cases studied here, HRT corridors tie into larger regional transit networks that provide comparatively high levels of regional rail accessibility. The HRT multimodal cor- ridor transit lines in San Francisco; Washington, DC; and Chicago feed into many destinations in each region’s central city, providing the transit rider with wider spatial coverage and higher regional connectivity/accessibility. These higher levels of accessibility and connectivity give the transit lines that run in multimodal corridors additional performance advantages. Corridor Orientation and Transit Ridership The performance of a multimodal corridor’s transit line also depends on its relationship to its surrounding environ- ment. We refer to this transit-environment relationship as corridor orientation, comprised of two components: corridor urban form and corridor station access. Each component is described in greater detail and analyzed in terms of its effects on corridor performance below. Corridor orientation is described in Chapter 4 as a contin- uum with two poles: transit- and automobile-orientation (see Fig. 4-1). Automobile-oriented corridors are planned to optimize automobile mobility over nonautomobile access. Transit-oriented corridors, on the other hand, are designed to maximize nonautomobile access to land uses and transit stations. Land uses are generally high-density with minimal parking. Multimodal corridors are, by definition, neither purely automobile- nor transit-oriented, but lie between the extremes of the corridor continuum, as shown in Fig. 4-1. Each point along the multimodal corridor continuum has a different com- bination of the critical facility design and surrounding land use factors that serve to optimize (or degrade) the capabilities of the corridor to function as a balanced, multimodal system. As discussed previously, factors that support a multimodal transit-oriented corridor are those that maximize access to transit stations by all modes of travel, but particularly by pedestrians. As a freeway facility will be running near it, a key challenge to creating an effective multimodal transit-oriented corridor is to minimize the negative externalities of the vehic- ular traffic traveling to and from the freeway. Factors that support a multimodal automobile-oriented cor- ridor are similar to those typically used to describe a purely automobile-oriented corridor (see Fig. 4-1), and like the multi- modal transit-oriented corridor, its differences are mainly those of emphasis. Transit stations or stops are designed to maximize automobile access and parking. Park-and-ride lots dominate the immediate station environments, and high-capacity road connections between station areas and the freeway encourage peak-period commuters to reduce freeway congestion by park- ing their cars and transferring to transit. The Effects of Corridor Urban Form Corridor urban form plays an important role in determining mode choice for corridor residents, visitors, and employees. The critical factors that describe urban form are discussed in Chapter 4. To measure the urban form orientation of each study corridor, several variables (see Table B-3) were chosen to rep- resent each of the four “D” factors. From these variables, Theoretical Component Component Measure Density Housing Units per Square Mile Diversity Entropy Index (Jobs-Housing Balance) Design 4-Leg Intersections per Square Mile Corridor Clustered Destinations Sq. Ft. Office Space in CBD Note: Entropy (Diversity) index calculated as mixed-use entropy (within 1/2 mile of each station) = –1*{[Σi (pi) (ln pi)]/ln k}, where p = proportion of total land uses; k = category of land use (single-family housing units, multifamily housing units, retail/service employment, office employment, manufacturing/trade/other employment); ln = natural logarithm. Table B-3. Urban form corridor orientation index components.

87 factor analysis was performed, and a single urban form factor score variable was created. The relationship between multimodal corridor urban form and the percentage of station area commuters using transit (see Figure B-7) suggests a positive relationship. Consistent with theory and the discussions above, the more transit-oriented the corridor urban form, the more riders the transit line attracts from its station neighborhoods (that is, within a half-mile of each station). However, as discussed earlier, caution should be used when interpreting these graphs, since the low sample size prevented more robust and statistically reliable testing. Figure B-7 also suggests the following five cases are the top performers, both in terms of running through corridors with predominantly multimodal transit-oriented urban form and attracting riders within a half-mile of their stations: • #2 Chicago Blue Line/Eisenhower Expressway • #3 Chicago Blue Line/Kennedy Expressway • #4 Chicago Red Line/Dan Ryan Expressway • #15 San Francisco Daly City Line/I-280 • #19 Washington D.C. Orange Line/I-66 San Francisco’s Daly City Line/I-280 offers a good example of a multimodal transit-oriented urban form corridor. A combination of residential density, mixed uses, pedestrian- oriented design, and a large CBD make this one of the most transit-oriented multimodal corridors in the United States. Table B-4 compares the urban form measures values for the Daly City Line corridor and the median values of the study corridors. The Daly City values are all above the study median, with the CBD size substantially higher, suggesting that size of Note: Commuter rail cases (i.e.; the New Haven Line/I-95 corridor) have been excluded since they tend to attract automobile- and bus-access riders from further distances from their stations than other transit modes. Figure B-7. Transit-oriented urban form increases corridor transit patronage. Component Measure Study Median Value S.F. Daly City Corridor Value Density (DUs/Ac.) 4.9 5.3 Diversity (Entropy Ind.) 0.85 0.86 Design (4-Leg Int./Ac.) 0.12 0.16 Destination (CBD Size) 42 mil. s.f. 110 mil. s.f. Table B-4. Urban form characteristics of the San Francisco Daly City/I-280 corridor.

88 a corridor’s anchor plays a critical role in determining transit ridership performance. Review of other cases suggests that while CBD size is important in determining transit line ridership in multimodal corridors, it does not guarantee it. Figure B-8 shows that while none of the top five transit ridership cases have CBDs smaller than 60 million square feet of office floor space, several cases with moderate or low transit ridership have CBDs of equivalent or greater sizes. Chicago’s multimodal corridors illustrate this point. Down- town Chicago has one of the largest concentrations of office floor space in the United States. This provides a large trip attractor at the end of each study corridor that encourages commuters to use both the transit and freeway facilities. Analysis of Chicago’s three multimodal corridors suggests that CBD size does not guarantee transit line ridership. Chicago’s dominant CBD helps the Blue Line/Kennedy Expressway corridor to attract the second-largest number of transit riders of any multimodal corridor transit line studied, but does not help the Blue Line/Eisenhower Expressway corridor place in the top five. When these corridors all serve the same, large CBD, what is different about the Blue Line/Eisenhower corridor that keeps it from attracting the same ridership as the Blue Line/Kennedy and the Red Line corridors? Table B-5 compares the transit ridership, the corridor’s commuter mode share, and the 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 Or an ge L in e K en . Bl u e Li n e BA RT P itt s. Li ne BA RT D al y C ity C hi c. R ed Li n e LA G re en Li ne LA G ol d Li n e Ei s. Bl u e Li n e D en v er T- RE X A tl. N .-S . L in e BA RT D ub . Li ne D en . C e n . Li n e Po rt . R ed L in e El M o n te Tr an sit w ay S. J.G u ad al u pe H o u . N W Sa c. N o rt h Li n e H ar bo r Tr an sit w ay HRT HRT HRT HRT HRT HRTLRT LRT LRT LRT LRT LRT LRTHRT HRT BRT BRT BRT 19 3 17 15 4 2 110 9 6 16 13 85 18 7 14 11 Multimodal Corridor D a ily Co rr id or Pa tr on a ge (T ra ns it + Fr ee w a y) 0 20,000,000 40,000,000 60,000,000 80,000,000 100,000,000 120,000,000 O ff ic e Fl oo rs pa ce (S qF t) Transit CBD Size HRT = Heavy Rail Transit LRT = Light Rail Transit CR = Commuter Rail BRT = Bus Rapid Transit Figure B-8. Transit patronage and central business district (CBD) size. Ridership/Component Measure Eisenhower Kennedy Dan Ryan Transit Line Ridership (Daily Boardings) 20,070 59,390 42,460 Corridor Transit Commuter Mode Share 24% 28% 31% Density (DUs/Ac.) 4.7 7.4 5.3 Diversity (Entropy Index.) 0.61 0.91 0.86 Design (4-Leg Int./Ac.) 0.18 0.15 0.15 Table B-5. Comparison of transit patronage and corridor urban form for Chicago’s multimodal corridors.

89 non-CBD urban form characteristics of each Chicago multi- modal corridor. In terms of ridership performance, the Eisenhower corridor underperforms its neighboring Chicago corridors, with less than half of the daily boardings of the Kennedy and Red Line transit lines, and a lower commuter transit mode share for residents living within a half-mile of its stations. The lower commuter transit mode share for the Eisenhower line suggests that the urban form in this corridor is more automobile-oriented than its neighboring corridors. For the most part, this appears to be the case, with housing densities and land use diversity (the amount of mixed use) substantially lower in the Eisenhower than in the two other corridors. While the urban design (as measured by the number of four-legged intersections per acre within a half-mile of the corridor’s stations) of the Eisenhower corridor appears to be somewhat more transit-oriented than either of its neighbors, all three Chicago cases have intersection densities above the study median of 0.11 four-legged intersection per acre. Therefore, while the Eisenhower corridor has development densities in its surroundings that are higher than many sub- urban corridors in this study, they are lower compared to its neighboring Chicago multimodal corridors. So while its densities are not adequate to provide high levels of walk-on patronage, park-and-ride is not practical because many stations are too close to the city center and pedestrian security can be a problem. It is reasonable to conclude that while a large CBD can help create a successful, well-patronized transit line, the line will benefit from a transit-oriented urban form along the rest of the corridor as well. Although it would be best to build all multi- modal facilities in corridors with transit-oriented urban form characteristics, most freeway corridors in the United States— where the lion’s share of multimodal corridor opportunity sites exist—have decidedly automobile-oriented land uses and urban design qualities. Therefore, as discussed in Chapter 3, we suggest that the new paradigm offers a two-step process of multimodal corridor planning, design, and construction, wherein transit facilities are designed and built in freeway corridors with the perfor- mance characteristics that allow them to compete with the free- way facility on a travel time basis using automobile-oriented multimodal coordination as the first step. Later, as conditions and resources permit, more transit-oriented land uses and operational characteristics can be introduced that will help the transit line reach its full potential as part of a larger new paradigm corridor. Therefore, our conception of the new paradigm does not discriminate against corridors with automobile-oriented urban form; rather, we see them as opportunities to build cost-effective, automobile-oriented transit lines that can be slowly transformed into transit-oriented lines. There are several examples of automobile-oriented multimodal corridors that have successfully taken this crucial first step. Urban form in Denver’s T-REX/I-25 corridor suggests a decidedly automobile-oriented pattern, but its early success at capturing transit riders suggests this as a prime example of “step one” in the new paradigm evolution toward a transit- oriented corridor (see Table B-6). Ridership for this new light rail line is excellent considering the fact that it is just over the study median (which includes many heavy rail lines that tend to attract higher ridership numbers) and is 26 percent higher than the median for study light rail lines. That T-REX’s station areas have a very low (6 percent) transit commute mode share compared to the study median of 15 percent suggests it draws most of its riders from beyond the half-mile walking distance buffer—a distance at which most patrons are likely to use buses or automobiles to park-and-ride. These ridership patterns are consistent with an automobile- oriented urban form pattern. The urban form metrics confirm this conclusion. Housing densities were among the lowest found in the study group, with less than one unit per acre (gross), substantially less than the study median of roughly five. The corridor is also decidedly residential in character; with a diversity score less than one-quarter that of the study median. Ridership/ Component Measure Study Median Value Denver T-REX Value Transit Line Ridership (Daily Boardings) 20,070 23,000 Corridor Transit Commuter Mode Share 15% 6% Density (DUs/Ac.) 4.9 0.18 Diversity (Entropy Ind.) 0.85 0.19 Design (4-Leg Int./Ac.) 0.12 0.05 Destination (CBD Size in Office Space) 42 mil. s.f. 23 mil. s.f. Table B-6. Transit patronage and urban form characteristics of the Denver T-REX/I-25 corridor.

90 In terms of urban design characteristics, the T-REX corri- dor is also decidedly automobile-oriented, with an average density of 0.05 intersection per acre compared to the study median of 0.12. The size of Denver’s CBD (roughly 23 million square feet) and the fact that the line serves the Denver Tech Center—an office park concentration south of the CBD— appears to make up for some of the automobile-oriented characteristics of the T-REX corridor, providing a relatively strong anchor on which to build the light rail line’s ridership. The line is also in a corridor that is growing in population. Given T-REX’s surprisingly low residential densities and high ridership (with the Denver Tech Center as a major trip generator), it seems likely that the design feature that matters more than anything in a park-and-ride access corridor is the number of park-and-ride spaces it provides at its stations. If Denver’s T-REX corridor is an example of a nascent park-and-ride access multimodal corridor with potential to evolve into a transit-oriented one, San Francisco’s Pittsburg- Bay Point line/SR 24 offers an example of a more mature and successful automobile-oriented corridor already undergoing some of the transformations into a more transit-oriented one. Table B-7 shows the relevant transit ridership and urban form metrics. The corridor serves a combined 51 million square feet of office space in the heart of the Bay Area—an important factor determining the corridor’s high transit commuter mode share (25 percent), and suggesting that a substantial number of its 57,000 daily boardings are coming from within a half-mile walking distance of its stations. As the corridor has developed over the last 36 years, the urban form of its stations’ areas has become steadily more transit-oriented. Currently, housing densities and mixed- use are roughly equal to the study medians, suggesting the corridor is neither automobile- nor transit-oriented in terms of urban form. However, its urban design qualities (as suggested by the density of four-legged intersections) are still somewhat automobile-oriented, with an average density of 0.09 four- legged intersection per acre compared to the study median of 0.12. As this corridor continues to evolve, planning policies that encourage TOD, the construction of infill stations along the corridor, and station access measures that encourage non- automobile access could lead to this case reaching its full potential as a multimodal transit-oriented corridor. Anecdotal evidence suggests these changes are already underway at sev- eral corridor stations.3 Determining successful transit line performance depends on which ridership performance measure is used. Transit line ridership counts—obtained from the transit agencies them- selves and adjusted to estimate the ridership along each study corridor segment—provide a measure that takes into account riders no matter how far they traveled to reach the transit line, or by what mode they arrived there. The “Transit Commuter Mode Share” value offers a dif- ferent take on transit ridership success. This measure suggests how well the transit line competes with other modes in cap- turing commuter trips in the corridor—in essence, the tran- sit orientation—specifically within reasonable walking dis- tance of the corridor’s stations (0.5 mile). As a result, it can (and does) happen that a particular transit line may attract high transit ridership numbers, while attracting a low share of the transit commuters within a half- mile of its stations. A comparison of Los Angeles’s Green Line/ I-105 and Harbor Transitway/I-110 corridors illustrates this point (see Table B-8). The Green Line (LRT) serves roughly 42,000 daily boardings, while the Harbor Transitway BRT line only serves roughly 4,000. However, the transit mode share within a half-mile of each line’s stations tells a different story. While the Green Line’s station areas have roughly 10 percent commuter mode share Ridership/ Component Measure Study Median Value Pittsburg-Bay Point Value Transit Line Ridership (Daily Boardings) 20,070 57,110 Corridor Transit Commuter Mode Share 15% 25% Density (DUs/Ac.) 4.9 4.5 Diversity (Entropy Ind.) 0.85 0.85 Design (4-Leg Int./Ac.) 0.12 0.09 Destination (CBD Size in Office Space) 42 mil. s.f. 51 mil. s.f. Table B-7. Transit patronage and urban form characteristics of the San Francisco Pittsburg-Bay Point Line/S.R. 24 corridor. 3Cervero, R., et al. TCRP Report 102: Transit Oriented Development in America: Experiences, Challenges, and Prospects.

91 among its residents, the Harbor Transitway’s station area residents have a 16 percent mode share. These seemingly contradictory results suggest that the Green Line is more successful at attracting riders from beyond its one-half mile station area radius, while the Harbor Transitway is successful at helping to encourage transit mode share within a half-mile of its stations, but does not attract riders from beyond. Part of the reason why the Harbor Transitway may be more successful in its immediate neighborhoods is the relatively higher transit-orientation of its corridor’s urban form. The Harbor Transitway corridor is substantially different from the Green Line corridor’s urban form in only one urban form characteristic—residential density, where the Harbor Transitway station areas are more than three times as dense as the Green Line’s. So while encouraging transit-oriented station area urban form can be an effective tool for encouraging station area ridership, it may not be sufficient to ensure high transit line ridership. The Effects of Corridor Station Access Similar to urban form, corridor station access reflects the design and operational elements within and near stations that encourage either automobile access (automobile-oriented) or pedestrian- and other non-automobile access (transit-oriented) modes. A high number of freeway ramps that touch down near transit stations can impede pedestrian station access. Similarly, the negative externalities of the freeway itself (for example, noise and air pollution) near transit stations can discourage pedestrian activities. Finally, although park-and-ride lots encourage automobile access to transit stations, they tend to impede pedestrian access. Our analysis used the variables shown in Table B-9 to represent the four components of corridor station access. Analysis of each variable individually, collectively as part of factor-analysis-generated index scores, and as part of multi- variate linear regression models found that the most important station access variable affecting multimodal corridor transit ridership was the number of freeway ramps that touch down within a quarter-mile of a station. Figure B-9 provides a graph of the average number of freeway ramps that touch down within 1⁄4-mile of stations per corridor station and the estimated transit line patronage (daily transit boardings) for each of our study corridors. A linear regression line drawn on this graph indicates that if there is a statistically valid relationship between these variables, it is not linear. Further exploratory analysis of these data suggests that the relationship between corridor patronage and multimodal coordination may be non-linear. Figure B-10 illustrates this relationship, where the more freeway ramps there are near Ridership/ Component Measure Green Line Harbor Transitway Line Ridership (Daily Boardings) 42,000 4,000 Transit Commuter Mode Share 10% 16% Density (DUs/Ac.) 7.2 24.7 Diversity (Entropy Ind.) 0.94 0.79 Design (4-Leg Int./Ac.) 0.11 0.10 Destination (CBD Size in Office Space) 42 mil. s.f. 42 mil. s.f. Table B-8. Comparison of transit line patronage and corridor urban form for the Los Angeles Green Line and Harbor Transitway corridors. Theoretical Component Component Variable Freeway Ramps Impede Pedestrian Station Access but Enhance Automobile Access Number of Freeway Ramps that Touch Down within ¼-Mile of Stations per Corridor Station Freeway Facility Negative Externalities Average Distance from Corridor Stations to Freeway Facility Park-&-Ride Lots Impede Pedestrian Station Access but Enhance Automobile Access Average Number of Park-&-Ride Spaces per Corridor Station Bus Access to Stations Average Number of Bus Lines Serving Stations per Corridor Station Table B-9. Corridor station access index components.

92 Figure B-9. Average ramps that touch-down within 1⁄4-mile of corridor stations and the estimated transit line patronage—linear regression line. Figure B-10. Transit ridership is higher when there are fewer freeway ramps near stations.

93 corridor transit stations, the lower the patronage for the transit line as a whole. There are several case studies that illustrate the importance of station access. Table B-10 compares the ridership and station access characteristics of the Eisenhower corridor with the median values of the study’s cases. While the urban form of the Eisenhower corridor is automobile-oriented, its station access characteristics tend to be more transit-oriented. This mismatch may be partially responsible for this transit line’s lower patronage levels than other Chicago area heavy rail lines. The corridor’s stations have the lowest number of park-and-ride spaces of any study case. Since park-and-ride spaces encourage automobile access to stations and discourage pedestrian, bicycle, and bus access, this implies that the transit line is designed to primarily serve corridor trips for people living near the corridor’s stations, rather than attracting automobile-to-transit transfers that often originate further away. That the corridor’s stations have a lower-than-median number of bus lines per station (3.2 versus 6.2 per station for all study corridors) reinforces the impression that the Blue Line’s stations in the Eisenhower corridor are designed to serve walk-access residents in its directly adjacent neighborhoods. However, its catchment area is limited because there are parallel rapid transit lines less than a mile to the north and about 1.5 miles to the south. The placement of the Blue Line’s stations in relation to the freeway facility discourages non-automobile access as well. The average distance from the corridor’s stations to the free- way is roughly 0.02 miles—essentially directly adjacent to the freeway and significantly lower than the median distance for the rest of the study corridors of 0.09 miles. This relatively short separation distance serves to increase the negative impacts of the freeway on the transit line. It is useful to contrast station access at the stations along the Eisenhower corridor to those along the Kennedy. Table B-11 compares the patronage and station access characteristics of the Kennedy and Eisenhower corridors in reference to the study median values. The success of the Kennedy corridor branch of the Blue Line at attracting transit patrons, both from within a half-mile walk- ing distance of its stations and beyond, is partially due to the reinforcing and complementary effects of the corridor’s transit- orientation, both in terms of urban form and station access. Ridership/ Component Measure Study Median Value Eisenhower Value Transit Line Patronage (Daily Boardings) 23,500 24,000 Corridor Transit Commuter Mode Share 15% 24% Average Number of Ramps per Station 2.8 2.8 Station to Freeway Dist. 0.09 0.02 Park-&-Ride Spaces/Station 4420 81 Bus Lines/Station 6.2 3.3 Table B-10. Transit patronage and station access characteristics of Chicago’s Blue Line/Eisenhower Expressway corridor. Ridership/ Component Measure Study Median Value Eisenhower Value Kennedy Value Transit Line Patronage (Daily Boardings) 23,500 24,000 59,000 Corridor Transit Commuter Mode Share 15% 24% 28% Average Number of Ramps per Station 2.8 2.8 2.6 Station to Freeway Dist. 0.09 0.02 0.20 Park-&-Ride Spaces/Station 420 81 166 Bus Lines/Station 6.2 3.3 5.1 Table B-11. Transit patronage and station access characteristics of Chicago’s Blue Line/Kennedy Expressway corridor.

94 While the median number of ramps that touch down within a quarter-mile of the corridor’s stations is only slightly lower than that seen in the Eisenhower corridor and the study cases as a whole, its stations are 10 times as far from its freeway neighbor as in the Eisenhower corridor, and more than double the distance seen in the study as a whole. While the number of park-and-ride spaces per station in the Kennedy corridor is roughly double the number found at the typical Eisenhower corridor station, Kennedy’s number is less than half that typically seen in the study cases, suggesting this corridor’s stations are designed to favor nonautomobile access. Furthermore, compared to the Eisenhower corridor, Kennedy corridor stations have been designed to encourage bus access. While the number of bus lines serving Kennedy stations is slightly lower than the typical study station, it is substantially higher than that seen in the Eisenhower corridor, suggesting these stations have been designed to encourage bus access. Seen as a whole, station access design in the Kennedy corridor’s stations are transit-oriented, thus reflecting and reinforcing the transit-orientation of the corridor’s land uses. This impression is consistent with the Blue Line’s his- tory in this corridor, where the elevated line was built in the late 1890s and the subway portions were built in the 1950s and 1970s. Thus, these areas were designed for an era where the primary modes of station access were non-automotive. These characteristics help explain the disparities in patron- age performance between the Eisenhower and Kennedy corridors. Consistent with the automobile-orientation of its corridor land uses and its multimodal coordination (that is, its station spacings are longer than its interchange spacings), access to the T-REX line’s stations are decidedly automobile-oriented as well (see Table B-12). On average, there are roughly three freeway ramps touch- ing down within a quarter-mile of each station (slightly higher than the 2.8 study median), suggesting that the T-REX light rail line was designed to offload traffic from the freeway onto transit. The average distance between stations and the freeway is roughly 0.05 mile, well below the study average of 0.9. While the number of park-and-ride spaces per station in this corridor (513) is below average compared to the study group (420), it is well above the median for study corridors that have light rail transit (261), suggesting that for a light rail line, this corridor’s stations are highly automobile-oriented. The automobile- orientation of this corridor’s stations complements and enhances the automobile-orientation of its corridor land uses, helping to make this new light rail line a ridership success. The Pittsburg-Bay Point/S.R. 24 corridor’s stations offer a useful example of automobile-oriented stations within an increasingly transit-oriented urban form context (see Table B-13). Prominent in this assessment is the fact that the average number of park-and-ride spaces per station in this corridor is roughly 1,600—more than double the study average of 420. The corridor’s stations are also close to the freeway (roughly 0.05 mile on average, compared to the study median of roughly 0.09), providing an attractive op- tion to freeway drivers to exit, quickly park, and complete their trips via BART. While an automobile-oriented station access profile is consistent with the corridor’s history of automobile-oriented urban form patterns, the transit line would benefit from measures to enhance the transit-orientation of its stations to match its transit-oriented urban form. The number of ramps per station is just below average and the number of bus lines per station is better than average, suggesting that the station access orientation can be made to favor pedestrians and transit relatively easily by consolidating or removing park- and-ride spaces. Ridership/Component Measure Study Median Value T-REX Value Transit Line Patronage (Daily Boardings) 23,500 23,000 Corridor Transit Commuter Mode Share 15% 6% Average Number of Ramps per Station 2.8 3.1 Station to Freeway Dist. 0.09 0.05 Park-&-Ride Spaces/Station 420 513 Bus Lines/Station 6.2 3.9 Table B-12. Transit patronage and station access characteristics of Denver’s T-REX/I-25 corridor.

95 The Effects of Constrained Freeway Capacity Of the common threads found among the case studies, constrained freeway capacity may be one of the most decisive factors in enabling transit to compete with the adjacent freeway. A constrained-capacity freeway has a substantial capacity bottleneck that creates congestion and causes delay. The bottlenecks found in this project are either caused by lane drops where the number of freeway lanes is reduced or where the capacity of the freeway was designed and built intentionally to be lower than forecast demand. As discussed previously, Washington D.C.’s Orange Line/ I-66 Corridor is an excellent example of a corridor where the freeway was purposely built as a capacity-restricted facility. As part of the financing package from Congress to fund the construction of the Orange Line, the Interstate was restricted to six lanes.4 This case sets an example of how freeway capac- ity restriction can substantially boost parallel transit line ridership and may also restrict total corridor throughput. As a result, this corridor is the only case studied for this project where the estimated transit mode share exceeds the estimated freeway mode share. The success of Chicago’s Kennedy corridor stems from several interlocking and mutually supporting factors. First, it has a heavy rail line, which provides fast, high-capacity transit service directly to downtown Chicago. This transit advantage is complemented by the freeway’s design, which has a relatively modest six lanes in its western portion, giving the rail line an advantage during peak congestion hours on the freeway. This capacity constraint allows the transit line to effectively compete with the freeway, garnering roughly 59,000 daily passenger boardings in the corridor. For San Francisco’s Pittsburg/Bay Point corridor, as in the case with Washington DC Orange Line/I-66, the restriction of the freeway’s capacity plays an important role in the story of the adjacent transit line’s success. Where Highway 24 and the BART line bore through the Oakland/Berkeley hills to reach the core Bay Area, the Caldecott Tunnel shrinks the freeway’s capacity from eight to six lanes. The center bore of the tunnel is reversible, so during commuting hours, the peak direction of flow always has four lanes of travel. However, the non-peak direction is reduced to two lanes, and as a result, there is always congestion and delay in both directions of travel during the A.M. and P.M. peak commute hours at the tunnel. While this nonpeak direction capacity constriction does not directly encourage peak direction use of the BART line, it does restrict nonpeak direction flow, thus providing a direct incentive for nonpeak direction BART ridership and indirectly promoting the general perception that BART is the more hassle-free corridor alternative. Summary The analysis of case studies of multimodal corridors in the United States for TCRP Project H-36 suggests that the following factors contribute to the capability of transit lines to effectively compete with and survive in a corridor with a freeway facility: a large CBD with limited and expensive park- ing, constrained freeway capacity, urban form, station access, multimodal coordination, and transit operating speeds. Based on our review and analysis of the case studies, the re- search team has identified the following desirable attributes for multimodal corridors: • Complementary multimodal coordination between tran- sit and freeway facilities • Transit-oriented land development around key stations that is readily accessible from station platforms • At least one large activity center or anchor, usually a CBD with high levels of employment Ridership/ Component Measure Study Median Value Pittsburg-Bay Point Value Transit Line Patronage (Daily Boardings) 23,500 57,000 Corridor Transit Commuter Mode Share 15% 25% Average Number of Ramps per Station 2.8 2.5 Station to Freeway Dist. 0.09 0.05 Park-&-Ride Spaces/Station 420 1,600 Bus Lines/Station 6.2 6.8 Table B-13. Transit patronage and station access characteristics of San Francisco’s Pittsburg-Bay Point Line/S.R. 24 corridor. 4Wikipedia, http://en.wikipedia.org/wiki/Interstate_66, accessed March 1, 2009.

96 • Limited and costly parking in the CBD • Effective transit distribution in the CBD, preferably off- street • Constrained freeway capacity such as lane drops, route convergence, and travel barriers • Good access to stations on foot, by car, and/or by public transport. This includes a minimum number of freeway interchange ramps within walking distance of transit stations The multimodal corridors examined in this study are generally successful in terms of transit riders carried and performance (that is, transit speeds). They are perhaps less successful in enhancing pedestrian access to stations and in achieving transit-oriented development. While it appears that a multimodal corridor need not possess the best qualities and quantities of each of these factors to perform well, it seems that there are optimal combinations of these qualities that lead to superior performance. It is intriguing to consider an optimal multimodal corridor system that combines, for example, a capacity constrained freeway, a large CBD, transit-oriented corridor urban form and station access, and high transit operating speeds.