Second look: the flaw in the traffic/parking analysis was to focus on Nets games rather than smaller events drawing more vehicles

As I recently wrote, regarding the confounding plan for a single entrance to a 758-space parking garage on Dean Street, there seems to be a fundamental flaw in the state's analysis of traffic and parking, one that appeared not only in the studies of the project before the opening of the Barclays Center, but also after the arena had been operating.

While a Nets game, given the likelihood of 15,000-plus attendees, was assessed as drawing the largest crowds, experience shows it does not necessarily provoke the most congestion: basketball fans are more attuned to public transportation, while concerts (like Barbra Streisand) that attract older audiences and family shows (like Disney On Ice) that draw children are more likely to bring vehicles, whether for parking or for drop-off.

Even the post-operating study of the Barclays Center, focusing on Nets games, looked only briefly at concerts, and while it noted that family shows and some concerts drew a greater number of cars, it didn't associate them with neighborhood gridlock.

It was not implausible in the initial analysis to think Nets fans from New Jersey would drive, but the long goodbye to New Jersey meant few fans followed the team.

Let's look at the documents.

2006 Final Environmental Impact Statement

From the 2006 Final Environmental Impact Statement (Final EIS, or FEIS), Chapter 12 (Traffic and Parking), at bottom:

The analyses of the off-street parking conditions during the weekday evening and Saturday midday periods each incorporate the anticipated peak parking demand generated by a Nets basketball game at the proposed arena.
As discussed in Chapter 13, “Transit and Pedestrians,” the project site is accessible via seven subway stations served by a total of a dozen subway routes, and is adjacent to the Long Island Rail Road’s Brooklyn terminus. In addition, a total of 11 local bus routes operate within ¼ mile of the project site. Locating high density, high quality development in proximity to transit nodes is recognized as an important planning strategy for reducing the numbers of auto trips that would otherwise be generated by such development. The proximity of good transit service, especially rail transit service, has also become a consideration in the site selection for major event venues such as the proposed Atlantic Yards arena. A recent example is the 1997 Verizon Center in Washington, D.C., home to the NBA Washington Wizards, which was constructed immediately above the Gallery Place–Chinatown station complex, a junction of three Metro lines. The Atlantic Yards arena and other components of the proposed project would similarly benefit from direct access to the ten subway lines serving the Atlantic Avenue/Pacific Street subway station complex via a major new on-site subway entrance that would be located adjacent to the arena. Internal circulation improvements to the subway station complex are also planned. The project site’s excellent accessibility by subway, commuter rail and bus is reflected in the relatively high mode shares for transit shown in Table 12-10—a combined 75 percent subway/bus share on weekdays for residential uses, for example. Transit facilities serving the project site are discussed in more detail in Chapter 13, “Transit and Pedestrians.”

(All emphases added)

The document explains why a Nets game was selected as a RWCS, or reasonable worst-case scenario. Note that, at the time, the review assessed an arena with 850,000 square feet, not 675,000 square feet, as later described after the arena was downsized:

The proposed project’s 850,000 gsf arena would accommodate 18,000 to 19,925 seats, depending on the event. The capacity for a basketball game, for example, would be 18,000 seats, whereas for the largest concert or other events, additional space for seating could be available on the arena floor. As a RWCS for the EIS transportation analyses, the weekday and Saturday travel demand forecasts examine the demand that would be generated by a Nets basketball game at the arena. A Nets basketball game was selected as an RWCS based on both the frequency of home games and the relatively high level of travel demand that such games are expected to generate compared with most other uses. Using the 2005-2006 season as a guide, approximately 41 games would occur at the arena during a typical basketball season from early November to late April (not including playoff games which could continue through June). Approximately 26 of these games would occur on a weekday, four on a weekend afternoon (Saturday or Sunday) and 11 on a weekend evening. (Although weekend evening games would occur somewhat more frequently than weekend afternoon games, the afternoon pre- and post-game peak hours would coincide with peak retail activity and typically experience higher ambient traffic levels. These periods were therefore selected for analysis.) 

What about non-NBA games?

Here's the crucial passage regarding other events:

Non-basketball events, such as concerts, ethnic shows, general fixed fee rentals (graduations, receptions, job fairs, etc.), and religious/motivational shows, other sporting events, family shows and community events, are each expected to occur with less frequency, would often attract fewer spectators, and would likely generate a lower level of travel demand than a Nets basketball game. As noted above, for the largest concerts and other events, additional space for seating would be available on the arena floor. However, when such factors as technical production requirements (stage size and placement, backdrop pieces, camera platforms, lighting, etc.), sightline restraints, and space requirements for wheelchair seating are accounted for, the actual capacity for most events at the arena would be less than the 18,000-seat capacity for a Nets basketball game. For example, for the most typical concert configuration–setting a stage at one end of the floor with approximately 270 degrees of potential seating–the capacity of the arena would be approximately 15,000 persons. Overall, conditions identified in the analysis for the pre-game and post-game peak hours are expected to occur at only limited instances throughout the year. Conditions in the large majority of days would experience fewer trips, including fewer vehicular trips.

Indeed, as I reported in 2017, over the arena's first four years of operation, as reported in the 2016 bond prospectus, concerts averaged between 9,711 and 11,944 attendees a year.

The key issue, we've learned, is not the number of spectators, but the number likely to drive or take vehicles, as opposed to public transportation.

Sellout ≠ full house

The analysis did wisely observe that a "sold-out" game did not mean a full house:

The travel demand forecast for the arena assumes a sold-out game for all 18,000 seats, and a daily trip generation rate of two trips per seat. It should be noted, however, that the actual number of spectators at a game is typically fewer than the number of tickets distributed, and that even a sold-out game typically has about 90 percent attendance. The daily trip generation rate of two trips per seat for all 18,000 seats therefore also accounts for trips by employees, players, coaches, team staff, and other such non-spectator demand.

Indeed, as I wrote in August 2018, the Nets averaged approximately 14,900 fans per game over the first three years, which I calculated as about 13.2% less than announced attendance. In the fourth season, "average attendance fell by 22% over the previous three-year average." That's a drop of 3,278, which works out to 11,622 attendees.

The 2014 Final SEIS

From Chapter 4D, Operational Transportation (also at bottom), from the 2014 Final Supplemental Environmental Impact Statement (Final SEIS), similarly addressed Nets games:

To be conservative, the traffic analysis for the Saturday pregame peak hour assesses conditions resulting from Phase II with an afternoon Nets game at the Arena, even though other types of events with lower attendance than a Nets game are typically scheduled on a Saturday afternoon and Nets games rarely occur at that time. All of these peak hours are consistent with those analyzed in the 2006 FEIS. The weekday and Saturday post-game peak hours for Arena demand that were analyzed in the 2006 FEIS are not included, as Project demand during these periods is primarily Arena-related and they are not typically considered peak travel periods for the residential, retail and public school uses that comprise Phase II of the Project.

The experience of basketball games, though not Saturday afternoon ones, was analyzed on the ground:

Weekday traffic and pedestrian counts for the SEIS transportation analyses were conducted on days when evening Nets games were scheduled at the Arena, and therefore the Existing conditions baseline traffic and pedestrian networks reflect game-day Arena demand in the weekday peak hours. However, as no Saturday afternoon Nets games or other major events were scheduled during the count program, the Existing conditions Saturday pregame peak hour traffic and pedestrian networks do not reflect Arena demand. Travel demand from a Saturday afternoon Nets game at the Arena was therefore forecasted for use in developing the Future Without Phase II background condition.

The chapter noted that the Barclays Center has a maximum capacity of 17,700 seats for basketball (actually 17,732) versus the 18,000 seats assumed in the Final EIS, but actual attendance during the 2012-2013 season (including Playoffs) averaged approximately 14,974, and no games were at capacity:

To be conservative, as well as to account for the potential for a Saturday afternoon game to attract a higher than average level of attendance, the travel demand forecast for Saturday Arena demand assumes a Nets game with 16,900 spectators. This number of spectators is consistent with the highest attended sold-out game during the 2012-2013 season. The Saturday pregame peak hour travel demand forecast for the Arena assumes a daily trip generation rate of two trips per seat for spectators. The temporal distribution, modal splits and vehicle occupancies of Arena trips are based on data from surveys of Arena patrons attending three weekend evening Nets games in January and February 2013. More than 600 attendees, age 16 and over, were interviewed at each game to provide a statistically significant sample, and the survey results were weighted to account for the actual distribution of attendees by broad seating sections based on ticket scan data and to account for children, who were not eligible for the survey. As arena employees typically arrive before the spectators, most if not all employee trips occur outside of the analyzed Saturday pregame peak hour for spectator demand. Arena employee trips are therefore not reflected in the travel demand forecast for the Saturday pregame peak hour.

The chapter noted that fewer people took vehicles and more walked, with more people inside vehicles and taxis than predict:

Compared to the 2006 FEIS, the Arena travel demand factors for the Saturday pregame peak hour in this SEIS reflect a lower temporal distribution for the Saturday pregame peak hour (27.2 percent versus 37.5 percent), a lower auto/taxi mode share (38.5 percent versus 43 percent), a lower subway mode share (39.3 percent versus 44 percent), a substantially higher walk mode share 9.5 percent versus 3 percent), and higher auto and taxi vehicle occupancies (3.22 and 2.82, respectively, versus 2.35 for both autos and taxis in the FEIS). As discussed above, the factors used in this SEIS are based on extensive surveys of spectators attending Nets games at the Arena.

Appendix C notes constraints on crowds:

While there is the potential for additional seating capacity for non game events, Americans with Disabilities Act (ADA) accessibility, production equipment, and line of sight, operational and staging requirements typically limits attendance at non basketball events to under 17,700.

The post-opening study

As I wrote in June 2013, arena consultant Sam Schwartz, during his public presentation, danced around the issue of drivers inundating the neighborhood for free parking, because his charge was getting fewer people to drive, rather than caring what happens when those come to the neighborhood. From the memo:

The overall goal of the TDM program described in the FEIS was to reduce peak hour auto traffic within 1/2 mile of the arena projected in the FEIS [Final Environmental Impact Statement] 2010 Build Condition by approximately 30%, by reducing the overall auto mode share and diverting 250 total autos to remote parking. The volume of peak hour autos generated by the arena were found to meet these goals, with approximately 40% fewer autos in the pre-game peak hour and approximately 20% fewer in the post-game peak hour than projected in the FEIS 2010 Mitigated Condition.
Although surveys found that remote parking facilities are minimally used, the combination of lower than projected auto mode share, lower overall attendance than assumed in the FEIS (which conservatively assumed an attendance of 18,000), higher vehicle occupancy, and lower peak hour percentages results in surpassing the auto trip reduction goals described in the FEIS.

As shown in the table below, a large percentage of Nets fans took public transit on both weekdays and weekends.

Outside of basketball

Events outside of basketball were surveyed, but without any action items in conclusion, given that the goal of the environmental review was met: to reduce auto share to NBA basketball games.

A large percentage of concertgoers for two youth-oriented acts, Mumford & Sons and Swedish House Mafia--took public transit on weekdays.

For weekend concerts, Swedish House Mafia still drew on transit, but Marc Anthony, who attracts a broader demographic in terms of age and location, had more than half the attendees arrive by car, with nearly 11% coming by taxi. Even so, that hardly approximates the impact of a Barbra Streisand or a Billy Joel concert.

For Disney on Ice, more than half the patrons drove, and a reasonable number took taxis. But the analysis didn't assess how driving and drop-offs affected neighbors.

Atlantic Yards FEIS Chapter 12 Traffic Parking by Norman Oder on Scribd

Atlantic Yards Final SEIS Chapter 4D_Operational_Transportation by Norman Oder on Scribd

Barclays Center TDM Memo 6_... by Norman Oder on Scribd