RESIDENTIAL PARKING WORKING GROUP MEETING TEN READ- AHEAD MATERIALS

RESIDENTIAL PARKING WORKING GROUP MEETING TEN READ- AHEAD MATERIALS In preparation for Residential Parking Working Group Meeting Ten, staff has assembled answers to a variety of questions that we received at the last Working Group meeting on February 6 th, 2017. Question or Request Short Answer Does Denver allow car sharing spaces No; only off-street car sharing spaces may be counted. within the 1,500 ft. limit to be located on street, or is it only off street? How much does 10 bike parking spaces Staff does not have cost estimates at this time. We ask cost or what is a unit cost for bike parking that any members of the Working Group with knowledge that we can compare to vehicle parking of these costs to bring information. cost? What are Arlington s bike parking location We have included the basic standards for secure, longterm bike parking. standards? Do we have an idea of the cost of a No; we ask that representatives enquire about this with developer-paid revenue-guarantee for car colleagues. However, regulations in Vancouver, Canada sharing? may provide guidance on how to price a car sharing service guarantee. Please provide map of Capital Bikeshare Staff have provided a map. However, note that the County expansion plans in Arlington is willing to accept the installation of a station at any location provided that the developer: Locates it on private property with public access. Pays for the capital cost. Pays for operating expenses. What would staff propose as a vehicleparking reduction in exchange for station with a sponsorship agreement, 3 for a 15-dock A reduction of 2 vehicle parking spaces for one 11-dock investment in Capital Bikeshare stations? station, 4 for 19-dock station. What were the trip-performance Developer agreed to monitor grocery store trips for 2 standards in the Mazda Site [SP440] years. If the share of trips is above 50%, then mitigation approval? measures will be required. Please provide unweighted averages of Instead of simply providing the unweighted figures, we parking demand by transit areas. have provided a description of the entire data-analysis process used to arrive at the numbers shown at the February 6 th meeting. Complete answers can be found on the following pages. We have also prepared a schedule for the remainder of the Working Group process (dates subject to change). Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 1 of 29

Schedule for the remainder of the Working Group process (subject to change) Month Day Event/Deadline February 22 Working Group Meeting Ten 28 Final Working Group Meeting March 2 Presentation to CMO on draft Working Group recommendation prior to public engagement 6-31 Public Engagement: (Goals: discover any missed considerations; record issues with the Working Group s proposal and the magnitude of feeling about the issue for consideration as staff develops proposal). a. Publicize through newsletters (Arlington County Government and ATP) b. Online comment c. Receive e-mails to general e-mail address d. Present to Arlington Ridge, Aurora Highlands, and Ashton Heights Civic Associations (if requested) e. Engage smart growth advocates f. Online webinars to reach out to BIDs, NAIOP, and affordable housing developers, as well as others 10 (ideally) Working Group finishes report April 3 Info item at Transportation Commission 4 Presentation to Civic Federation (James Schroll unavailable in March) 11 Info item at Economic Development Commission 13 Info item at Housing Commission 17 (week of) Complete DES, CPHD, and AED review and approve staff recommendation (starts in February) 24 (week of) Brief Mark Schwartz and make final modifications to the staff recommendation May June 1 Action item at Transportation Commission 5 or 7 at Planning Commission 8 at Housing Commission 12 (week of) Board Briefings 13 Action item at Economic Development Commission 17 or 20 Board Action Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 2 of 29

Does Denver allow car sharing spaces within the 1,500 ft. limit to be located on street, or is it only off street? No; only off-street car sharing spaces may be counted. The following conditions are part of Article 10, Section 4.5.3 B (page 10.4-11): The car sharing program shall be: (1) Located on a zone lot and not within public right-of-way; (2) Located within a walking distance no greater than 1,500 of the zone lot containing the subject primary use; (3) Either in existence or being developed concurrently with the proposed development; and (4) Has the capacity to meet a portion of the vehicle parking needs of the subject primary use. Also, note that while on-site car sharing can be provided at a ratio of 5 required vehicle parking spaces reduced for each 1 car share space provided, that reduction is [d]etermined by the Zoning Administrator if the car sharing service is provided off site (p. 10.4-11). How much do 10 bike parking spaces cost or what is a unit cost for bike parking that we can compare to vehicle parking cost? Staff does not have good data on the cost of building secure bike parking. We ask that anyone on the Working Group who has estimates to bring them for the Working Group s consideration. Keep in mind that for underground bike parking, the total cost is cost of constructing an underground space and then installing a secure bike cage or room as well as bike racks; the total cost of constructing an above ground bike room is simply the cost of constructing any room with a locked door and then installing bike racks. What are Arlington s bike parking location standards? Arlington County has standards for two kinds of bike parking: short-term visitor parking (class III) and long-term secure parking (class I). As the Working Group has been discussing long-term secure parking, staff have summarized the standards for that parking here. Class I parking must provide protection from weather; security of locked room or cage; ability to lock bicycle to a rack within the room or cage. Location Best: Fully enclosed ground floor room with sidewalk access Second Best: Fully enclosed room in garage Third Best: Cage in a garage In Special Circumstances: Bike lockers Room Design Requirements Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 3 of 29

At least 30% of bicycle parking must be horizontal and at ground level Cage walls must be made of industrial grade expanded metal or welded wire mesh Doors must be hollow metal Doors must use a heavy-duty cipher lock or electronic lock Cages only: door handles and locks supplemented by sheet metal plating, 3 feet in each direction, as needed to prevent tampering Cage walls must reach all the way to ceiling For more information, see www.commuterpage.com/bikeparkingstandards. Do we have an idea of the cost of a developer-paid revenueguarantee for car sharing? Staff does not have data to answer this question and believe that the number will vary not only by region but also by location within the region. This question could be answered by Working Group members or observers who are in the real-estate development community and who have worked with car sharing providers to offer service on site at residential projects. One point of reference comes from the City of Vancouver, Canada. There, developers are required to agree to a letter of credit at a rate of $50,000 per year for 3 years. If the building owner does not ensure that car sharing service is provided, the City may call on that letter of credit from the developer and use the funds to make sure the service is provided. Any funds in excess of what the City needs to ensure that service is provided can be refunded to the building owner. Provide map of Capital Bikeshare expansion plans in Arlington The cost of a Capital Bikeshare station has two elements: the one-time cost of installation and on-going operation costs to cover maintenance and other services. Arlington County has an existing sponsorship program by which private entities can pay for the cost of station installation and operating expenses. The private entity must also provide space for the station on private property, as some locations do not have space in the public right of way. Currently, the cost to install a Capital Bikeshare station follows a schedule based on the number of docks. Approximate Capital Bikeshare Station Sponsorship Costs Station Size Equipment and Installation Cost (approx.) Operating Costs* Monthly (yearly) Station with bikes and 11 docks $40,000 $1,200 ($14,400) Station with bikes and 15 docks $50,000 $1,600 ($19,200) Station with bikes and 19 docks $60,000 $2,100 ($25,200) *Operating costs subject to increases up to 5% each year. Currently there is no minimum time period over which a sponsor must pay the operating costs, though staff would likely consider a minimum time period as part of a parking-policy recommendation. Staff Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 4 of 29

does not support arrangements that would pay for capital costs alone, as this would create an operating cost for the County that would not have been budgeted. Staff would not support the sponsorship of a Capital Bikeshare station in lieu of secure, on-site bike parking. Note, however, that the County will accept sponsorship agreements for any location as long as space for the station is provided, as well as capital and operating costs. This map depicts current expansion plans for the Capital Bikeshare system in Arlington. Arlington Capital Bikeshare Expansion Map Locations with green dots are potential future sites Stations planned for installation in FY2017, FY2018, and FY2019 are marked with their year of planned expansion Source: http://www.bikearlington.com/pages/bikesharing/planned-arlington-stations/ What would staff propose as a vehicle-parking reduction in exchange for investment in Capital Bikeshare stations? For consistency with the off-street, secure bike-parking proposal presented at the February 6 th meeting, staff propose that sponsorship for one 11-dock station would be accepted for a reduction of 2 parking spaces; a 15-dock station would be accepted for a reduction of 3 parking spaces, and a 19-dock station would be accepted for a 4-space reduction. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 5 of 29

As a point of reference, the City of Denver also allows reductions of one space for every five bike share docks. What were the trip-performance standards in the Mazda Site [SP440] approval? The following text is from page 21 and 22 of the staff report for the project. Transportation Demand Management (TDM): The applicant has agreed to implement a standard Transportation Management Plan (TMP) to provide a program oriented towards decreasing single-occupancy vehicle (SOV) trips to and from the site. The TMP program includes the elements summarized below: Designation of a transportation coordinator to oversee the TMP program. Contribute to Arlington County Commuter Services (ACCS) to support TMP activities. Provision of a $65 SmarTrip card, or a one-year Capital Bikeshare membership, or a one-year carshare membership to each new residential lessee and on-site management employee, at initial occupancy. Display and distribute transit-related information. Reimburse the County for and participate in periodic transportation performance monitoring studies, to determine average vehicle occupancy and mode choice. Submission of an annual report to the County regarding TDM activities on the site. In addition, the applicant has agreed to enhance its TDM program to attempt to further reduce the number of SOV trips at the site, and to offer transportation alternatives to residents, employees, and shoppers using the site. The enhanced TDM measures include: hiring a TDM coordinator specifically for the grocery store; providing information and promotional signage for the availability of carshare, and rideshare or taxi designated pick-up areas; providing real-time transportation informational displays visible from within and outside the grocery store, at building entrances on N. Glebe Road and N. Tazewell Street; providing carshare parking spaces within the garage; providing additional incentives to grocery store employees to use transit or alternative transportation modes; and providing a limited number of free promotional wire pullcarts for grocery store patrons. Additionally, the applicant agrees to monitor grocery store trips for two years, to ensure that the vehicular driving rate remains below 50% of trips. Mitigation measures, including additional services, programs, and other initiatives to reduce drive-alone trips will be put in place should the driving mode share exceed the 50% rate. [emphasis added] This 50% threshold is related to assumptions made for a traffic impact analysis submitted by the developer in April 2015. From the same staff report (page 15): Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 6 of 29

Trip Generation: Gorove/Slade Associates, Inc. prepared a traffic impact analysis (TIA) dated May 26, 2015 1, assuming a development plan for the site that includes 483 residential units, a 43,000 square-foot grocery store, and an additional 24,000 square feet of ground floor retail use. Upon site build-out (projected for 2018) the project is estimated to generate 215 morning, 361 evening, and 402 Saturday peak-hour vehicle trips. The trip generation assumptions for this site were agreed to at the original scoping for the TIA in April 2015. One of those assumptions was that the drive-alone mode share for trips at this site would be reduced by 50% from the Institute of Transportation Engineers (ITE) standard. That assumption is based upon the urban location of the site, its high walk and bike scores, its proximity to high-speed and high-frequency transit, the large number of residents living within a short distance from the site, both current and projected, the large number of employees working within a short distance of the site, and the observed mode share of similar types of mixed-use grocery store developments in Arlington and Washington. An analysis 2 of five mixeduse development grocery stores in close proximity to Metro (< 0.5 mile) with high walk and transit scores and proximity to major arterials showed that the number of vehicular trips is 15% to 35% of the number of trips that would be expected using the ITE standard. One mixed-use grocery store development with high transit and high walk scores, but further (1.7 miles) from Metro, had a vehicular trip generation rate of 39% (am peak) to 67% (pm peak) of the ITE standard. For this site, Gorove/Slade estimates that vehicular traffic generated by this site would be 44% to 45% of ITE s estimates. Thus, the 50% mode share reduction has been assumed and incorporated into the subsequent analyses. Provide unweighted averages of parking demand by transit areas At Working Group Meeting Nine, staff presented an overview of the analysis of recent parking-demand data that they undertook in order to inform proposals for new parking minimums related to transit access. Recall that recent parking demand is only one way to set parking supply requirements for the future. For analysis, staff consulted four sources of data that describe parking demand in the recent past in Arlington: 1. Counts of vehicles at garages in residential Site Plan buildings as gathered through periodic monitoring studies that are required of many Site Plan developments. 2. Vehicle-registration data from the Commissioner of Revenue expressed as the number of vehicles per unit of housing registered to addresses at multi-family buildings in Arlington. 3. Parking-space leases at rental apartments in the two Metro Corridors as collected by Neal Kumar of Rushmark Properties and the Arlington Economic Development Commission. 4. A 2015 survey of Arlington residents fielded by Arlington County Commuter Services. 1 Updated September 11, 2015, with a supplemental technical memorandum issued on January 21, 2016. 2 Observed Trip Generation in Mixed-Use Developments with Grocer: Washington, DC and Arlington, VA, Gorove/Slade, March 4, 2016 Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 7 of 29

The following table numerically describes these data sets and also provides comment on how each data set might be biased toward greater numbers (marked with a + sign) or biased toward fewer numbers ( - sign) simply because of the way each data set was collected. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 8 of 29

Characteristics of Various Parking Demand Data Sources for Multi-Family Buildings in the Rosslyn-Ballston and Jefferson Davis Corridors Data from Neal Kumar Commissioner of Revenue Data Resident Study 2015 Performance Monitoring Studies Unit of Measure Parking Parking Registered Vehicles/ Vehicles/Unit Lease/Unit Lease/Occupied Unit Vehicles/Unit Household Highest Quartile 1.51 1.59 1.24 1 1.3 75 Percentile 1.07 1.13 0.95 1 1.07 Mean 0.98 1.03 0.84 1.02 0.93 Median 0.93 0.98 0.81 1 0.89 25 Percentile 0.78 0.82 0.68 1 0.8 Lowest Quartile 0.49 0.52 0.33 1 0.64 Characteristics of Data Sample Size 49 bldgs 49 bldgs 97 bldgs 976 individuals 22 bldgs Includes On and Off-Street? No (-) No (-) Yes (+) Yes (+) No (-) Includes Visitor Parking? No (-) No (-) No (-) No (-) Yes (+) Captures Un-Registered Yes (+) Yes (+) No (-) Yes (+) Yes (+) Vehicles? Occupied Units? No (-) Yes (+) No (-) Yes (+) No (-) Subject to Seasonal Parking No No No No Yes Demand? Other Notes Potential sample bias The box and whisker plot on the next page depicts the same data sets graphically. In this plot, the box represents the range between the 25th and 75th percentile values. The x marks represent the mean, while the line inside of the box is the median (or 50th percentile value). The thin lines or whiskers extend to the 0th percentile and 100th percentile; dots represent outliers. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 9 of 29

Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 10 of 29

Vehicles/Unit Staff then looked at the relationships between the distance to Metro and the maximum observed parking demand (in the case of the Site Plan studies) or the number of vehicles registered per unit (in the Commissioner of Revenue data). Maximum Parking Demand by Distance from Metro Stations, Site Plan Buildings 1.4 1.2 y = 0.7254x + 0.7401 R² = 0.2495 1 0.8 0.6 0.4 0.2 0 0 0.5 1 1.5 2 2.5 Distance from Metro (miles) Less than One Mile, Unbundled More than One Mile from Metro, Unbundled Linear (Less than One Mile, Unbundled) Less than One Mile, Bundled More than One Mile from Metro, Bundled Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 11 of 29

As seen in the chart above, analysis of the Site Plan data set revealed that a statistically significant relationship between parking demand and distance to Metro is only present for buildings closer than 1 mile to a Metro station (in this case, as measured by walking distance, not straightline or crow-flies distance). The relationship was stronger at properties that unbundle parking charges. The following linear regression output describes the relationship between the independent variable (distance to Metro) and the dependent variable (maximum observed parking demand) for those 17 buildings with unbundled parking in the data set. REGRESSION SUMMARY OUTPUT FOR PROPERTIES LESS THAN 1 MILE FROM METRO, WITH UNBUNDLED PARKING Regression Statistics Multiple R 0.499513505 R Square 0.249513742 Adjusted R Square 0.199481325 Standard Error 0.173552359 Observations 17 ANOVA df SS MS F Significance F Regression 1 0.150211791 0.150211791 4.987041517 0.041197329 Residual 15 0.451806318 0.030120421 Total 16 0.602018109 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 0.740142404 0.093003454 7.958224926 9.1839E-07 0.541910235 0.938374573 Distance to Station (miles) 0.7254494 0.324852067 2.233168493 0.041197329 0.03304361 1.417855191 Note: Three buildings were not included due to a lack of parking-demand data. The Significance F value is significant at the 95% confidence level and the confidence interval (0.33 to 1.42 ) are both positive, which indicates that the relationship between distance to Metro and parking demand is significant, and not merely due to chance. Staff found a statistically significant relationship between distance to Metrorail station (again, defined as the walking distance, not a straightline distance) and the number of vehicles per unit registered to multi-family buildings across the entire data set and for the buildings within one mile of a Metro station. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 12 of 29

Vehicles Registered per Unit This scatter plot describes the relationship between distance to Metro and vehicle registrations per unit of housing for entire Commissioner of Revenue data set. 2 Vehicles Registered per Unit by Distance to Metro 1.8 1.6 1.4 R² = 0.3129 1.2 1 0.8 0.6 0.4 0.2 0 0 0.5 1 1.5 2 2.5 3 Distance to Metro (miles) Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 13 of 29

This regression output describes the same data set numerically. REGRESSION SUMMARY OUTPUT FOR ALL PROPERTIES IN THE COMMISSIONER OF REVENUE DATA SET Regression Statistics Multiple R 0.559418 R Square 0.312949 Adjusted R Square 0.307223 Standard Error 0.253942 Observations 122 ANOVA df SS MS F Significance F Regression 1 3.524805 3.524805 54.6595 2.12E-11 Residual 120 7.738391 0.064487 Total 121 11.2632 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 0.751184 0.033386 22.49984 7.13E-45 0.685082 0.817287 Distance to Metro 0.216791 0.029323 7.393207 2.12E-11 0.158733 0.274848 The Significance F value is significant at the 99% confidence level and the confidence interval (0.16 to 0.27 ) are both positive, which indicates that the relationship between distance to Metro and parking demand is significant, and not merely due to chance. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 14 of 29

The regression output for those properties only within one mile of a Metro (below) also describe a statistically significant relationship. SUMMARY OUTPUT FOR PROPERTIES LESS THAN ONE MILE FROM A METRO STATION 1 MILE USING COMMISSIONER OF REVENUE DATA Regression Statistics Multiple R 0.274239 R Square 0.075207 Adjusted R Square 0.065045 Standard Error 0.248437 Observations 93 ANOVA df SS MS F Significance F Regression 1 0.456761 0.456761 7.400431 0.007812 Residual 91 5.616596 0.061721 Total 92 6.073357 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 0.701771 0.058172 12.06379 1.41E-20 0.58622 0.817322 Distance to Metro 0.345024 0.12683 2.720373 0.007812 0.093093 0.596956 Though the R square value is low for the buildings closer than 1 mile to Metro (0.075 ), the Significance F and p-value for the Distance to Metro variable are significant at the 99% confidence level and the confidence interval (0.093 to 0.59 ) are both positive, which indicates that the relationship between distance to Metro and parking demand is significant, and not merely due to chance. The data set provided by Neal Kumar did not have measurements for the distance to Metro for each property which meant that a similar analysis was not possible with that data. With a relationship found between parking demand and distance to a Metro station, staff then divided the available data sets into ¼-mile ranges (0 to ¼ mile from a Metro station, ¼ mile to ½ mile from a Metro station, etc.) which are described graphically in the following chart. Staff chose to divide the data into ¼-mile ranges because these are easily understood and commonly used thresholds in transportation planning. The following chart describes these data sets divided into ¼-mile subsets. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 15 of 29

Note that there were no buildings in the Site Plan Performance Monitoring data set between 0.5 and 1.0 miles from a Metro station. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 16 of 29

The following table describes the same data numerically with data from the Resident Study also added. Note that while the entire Resident Study data set is heavily influenced by the number of one-car households, the mean value of reported vehicles owned per household is smaller for households closer to Metro than those far away. Parking Demand/Vehicle Registration/Vehicle Ownership based on Distance to Transit from Various Data Sources for Multi-Family Buildings in the Rosslyn-Ballston and Jefferson Davis Corridors Under 0.25 Miles 0.25 to 0.5 Miles 0.5 to 0.75 Miles 0.75 to 1.0 Miles SP COR Resident Study SP COR Resident Study SP COR Resident Study SP COR Resident Study Highest Quartile 1.21 1.11 1.3 1.24 1.18 1.71 75 Percentile 1.06 0.89 1.04 0.93 1 1.51 Mean 0.93 0.83 0.97 0.9 0.79 1.04 0.8 1.23 1.33 1.21 Median 0.9 0.81 0.82 0.79 0.79 1.28 25 Percentile 0.81 0.74 0.78 0.65 0.58 1.05 Lowest Quartile 0.67 0.52 0.64 0.33 0.38 0.72 Number of Buildings 13 25 7 39 21 8 In order to summarize this data as a reference for the Working Group s discussions on minimum parking requirements, staff decided to take the average of the mean, median, 25 th percentile, and lowest quartile values from the Site Plan Monitoring Study and Commissioner of Revenue datasets. However, given that the two datasets had different numbers of buildings, staff decided to take the weighted average of these two values based on the number of buildings in each (in other words, the Commissioner of Revenue data, as a larger data set, had more weight in the calculations). Of course, the number of vehicles reported in the Commissioner of Revenue data is smaller than in the Site Plan studies; this is expected in part because some households do not register their vehicles with Arlington County, and because the Site Plan studies would have included any visitor-parking demand in the garage while the Commissioner of Revenue data would not have captured visitor-parking demand. In order to compensate for this difference, staff compared the two data sets, and found that the difference in the median and mean values was approximately 0.1; staff decided to adjust the Commissioner of Revenue values by 0.1 vehicles per unit in order to make the two datasets more comparable. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 17 of 29

The following table describes the weighted averages for the two datasets after adjusting the Commissioner of Revenue data by 0.1 vehicles per unit. Weighted Average Parking-Demand Values based on Distance to Metro for Multi-Family Buildings in the Rosslyn-Ballston and Jefferson Davis Corridors Under 0.25 Miles 0.25 to 0.5 Miles 0.5 to 0.75 Miles 0.75 Miles to 1 Mile Highest Quartile 1.21 1.33 1.28 1.81 75 Percentile 1.01 1.03 1.10 1.61 Mean 0.93 0.89 0.90 1.43 Median 0.91 0.88 0.89 1.38 25 Percentile 0.83 0.75 0.68 1.15 Lowest Quartile 0.64 0.46 0.48 0.82 These data were the basis for the potential minimum ratios (by distance to Metro) presented at the February 6 th Working Group meeting. The Working Group was also interested in measures of transit access that would capture the value of all modes of transit, not just Metrorail. Staff decided to refer to a data set from the University of Minnesota s Accessibility Observatory that describes transit access to jobs for all Census Blocks in the United States. Staff used the latest data set available, which the University released in late 2016. 3 In this case, transit access is defined as the number of jobs that can be reached by transit within 30 minutes during a weekday, morning peak period of transit service. 3 A full methodology report is available at http://ao.umn.edu/research/america/transit/2015/. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 18 of 29

Maximum Number of Vehicles Parked per Unit The following scatter plot describes how parking demand varies with job access at the Site Plan buildings closer than 1 mile to Metro (no significant relationship was found when buildings more than 1 mile away from a Metro station were included). 1.40 Parking Demand at Arlington Site Plan Buildings by Job Accessibility on Transit, Less than 1 Mile from Metro, Bundled and Unbundled Parking Charges 1.20 1.00 0.80 R² = 0.2202 0.60 0.40 0.20 0.00 0 100000 200000 300000 400000 500000 600000 700000 800000 Number of Jobs Accessible within 30 Minutes Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 19 of 29

This regression output describes the same data numerically. ACCESSBIILITY BY TRANSIT; NUMBER OF JOBS REACHABLE IN UNDER 30 MIN; SUMMARY OUTPUT FOR PROPERTIES LESS THAN 1 MILE FROM METRO, BOTH BUNDLED AND UNBUNDLED PARKING Regression Statistics Multiple R 0.469270652 R Square 0.220214945 Adjusted R Square 0.179173627 Standard Error 0.16152811 Observations 21 ANOVA df SS MS F Significance F Regression 1 0.139998 0.139998 5.365689 0.031862 Residual 19 0.495735 0.026091 Total 20 0.635733 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 1.222253434 0.130632 9.356478 1.52E-08 0.948838 1.495669 Jobs Accessible within 30 minutes by transit -5.46384E-07 2.36E-07-2.3164 0.031862-1E-06-5.3E-08 Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 20 of 29

Maximum Number of Vehicles Parked per Unit When only properties closer than a 1-mile walk to a Metro station with unbundled parking were included, the relationship was somewhat weaker. 1.40 Parking Demand at Arlington Site Plan Buildings by Job Accessibility on Transit, Unbundled Parking and Less than 1 Mile from Metro 1.20 1.00 0.80 R² = 0.2092 0.60 0.40 0.20 0.00 0 100000 200000 300000 400000 500000 600000 700000 800000 Number of Jobs Accessible within 30 Minutes Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 21 of 29

ACCESSBIILITY BY TRANSIT; NUMBER OF JOBS REACHABLE IN UNDER 30 MIN; SUMMARY OUTPUT FOR PROPERTIES LESS THAN 1 MILE FROM METRO WITH UNBUNDLED PARKING Regression Statistics Multiple R 0.45735571 R Square 0.209174245 Adjusted R Square 0.156452528 Standard Error 0.17815563 Observations 17 ANOVA df SS MS F Significance F Regression 1 0.125926683 0.125926683 3.967515799 0.064915254 Residual 15 0.476091426 0.031739428 Total 16 0.602018109 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 1.206745376 0.147734901 8.168316162 6.65699E-07 0.891855889 1.521634862 Jobs Accessible within 30 minutes by transit -5.34745E-07 2.68465E-07-1.991862394 0.064915254-1.10696E-06 3.74742E-08 The R square is slightly lower and the significance F is no longer significant at the 95% confidence level (though it is still significant at the 90% confidence level), which is why the confidence interval at 95% crosses 0. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 22 of 29

Vehicles Registered Per Unit The following scatter plot describes how vehicles registered per unit from the Commissioner of Revenue changes as job accessibility on transit changes. 2 Vehicles Registered per Unit by Number of Jobs Accessible by Transit in 30 Minutes; All Locations 1.8 1.6 1.4 1.2 1 0.8 0.6 R² = 0.2684 0.4 0.2 0 0 100000 200000 300000 400000 500000 600000 700000 800000 Number of Jobs Accessible within 30 Minutes on Transit Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 23 of 29

The following regression output describes the same data numerically. VEHICLES REGISTERED PER UNIT BY NUMBER OF JOBS ACCESSIBILE BY TRANSIT IN 30 MINUTES; ALL LOCATIONS Regression Statistics Multiple R 0.518036 R Square 0.268362 Adjusted R Square 0.262861 Standard Error 0.266112 Observations 135 ANOVA df SS MS F Significance F Regression 1 3.454647 3.454647 48.78382 1.24E-10 Residual 133 9.418452 0.070815 Total 134 12.8731 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 1.26418 0.051748 24.42967 5.19E-51 1.161825 1.366535 Jobs Accessible within 30 minutes by transit -8.7E-07 1.24E-07-6.98454 1.24E-10-1.1E-06-6.2E-07 The Significance F value is significant at the 99% confidence level and the confidence interval are both negative, which indicates that the relationship between distance to Metro and parking demand is significant, and not merely due to chance. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 24 of 29

Vehicles Registered per Unit When looking at only the buildings in the Metro Corridors, vehicle registration data remains correlated to job accessibility. 1.8 Vehicles Registered per Unit by Number of Jobs Accessible by Transit in 30 Minutes; Within Metro Corridors Only 1.6 1.4 1.2 1 0.8 0.6 R² = 0.0784 0.4 0.2 0 0 100000 200000 300000 400000 500000 600000 700000 800000 Number of Jobs Accessible within 30 Minutes by Transit Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 25 of 29

The correlation is weaker for the buildings within the Metro Corridors alone, but this is likely, in part, simply because there are fewer data points. The following regression output describes the same data numerically. VEHICLES REGISTERED PER UNIT BY NUMBER OF JOBS ACCESSIBILE BY TRANSIT IN 30 MINUTES; METRO CORRIDORS ONLY Regression Statistics Multiple R 0.279929 R Square 0.07836 Adjusted R Square 0.068659 Standard Error 0.243413 Observations 97 ANOVA df SS MS F Significance F Regression 1 0.478569 0.478569 8.077158 0.00548649 Residual 95 5.628722 0.05925 Total 96 6.107291 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 1.045943 0.076459 13.67976 3.5E-24 0.894152787 1.19773412 Jobs Accessible within 30 minutes by transit -4.6E-07 1.62E-07-2.84203 0.005486-7.8029E-07-1.38493E-07 Given the difficulty of using job accessibility by transit as a measurement for parking-minimum policy, staff decided to consider recent parking demand data by planning area. The planning areas are commonly understood and well defined, and they also reflect differences in transit service around the County. The following box-and-whisker plot describes the Commissioner of Revenue data, Site Plan data, and space-rental data from Neal Kumar divided up into station areas. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 26 of 29

Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 27 of 29

The following table describes these data numerically. Parking Demand/Vehicle Registration/Vehicle Ownership based on Metro Corridor Planning Area Ballston Virginia Square Clarendon Courthouse Rosslyn Pentagon City Crystal City COR NK SP COR NK* SP COR NK SP COR NK SP COR NK SP COR NK SP COR NK SP Highest 1.52 1.63 1.1 1.18 1.08 0.82 1.04 1.1 1.18 1.01 1.21 1.23 0.91 1.1 1.48 1.41 1.3 0.96 1.48 1 Quartile 75th 1.1 1.32 1.1 1.09 1.08 0.82 1.01 1.02 0.98 0.91 1.21 0.92 0.91 1.16 1.08 1.16 0.82 1.07 1 Percentile 5 Mean 0.89 1.19 0.98 0.95 0.83 0.71 0.9 0.84 0.89 0.79 0.97 0.79 0.78 0.85 0.93 0.97 0.77 0.95 0.89 Median 0.83 1.07 0.92 0.89 0.88 0.73 0.95 0.77 0.81 0.79 0.86 0.78 0.76 0.68 0.83 0.96 0.77 0.92 0.89 25th 0.72 0.94 0.9 0.89 0.54 0.6 0.8 0.72 0.74 0.65 0.84 0.65 0.59 0.61 0.73 0.8 0.68 0.73 0.77 Percentile Lowest 0.33 0.49 0.9 0.89 0.54 0.48 0.63 0.67 0.6 0.58 0.84 0.33 0.59 0.84 0.51 0.65 0.64 0.65 0.73 0.77 Quartile N 23 16 5 9 3 5 6 6 15 8 3 21 3 3 9 9 8 15 7 2 Highest 1.62 1.63 1.1 1.28 1.08 0.92 1.04 1.1 1.28 1.01 1.21 1.33 0.91 1.1 1.58 1.41 1.3 1.06 1.48 1 Quartile 75th 1.2 1.32 1.1 1.19 1.08 0.92 1.01 1.02 1.08 0.91 1.21 1.02 0.91 0 1.26 1.08 1.16 0.92 1.07 1 Percentile 5 Mean 0.99 1.19 0.98 1.05 0.83 0.81 0.9 0.84 0.99 0.79 0.97 0.89 0.78 0 0.95 0.93 0.97 0.87 0.95 0.89 Median 0.93 1.07 0.92 0.99 0.88 0.83 0.95 0.77 0.91 0.79 0.86 0.88 0.76 0 0.78 0.83 0.96 0.87 0.92 0.89 25th 0.82 0.94 0.9 0.99 0.54 0.7 0.8 0.72 0.84 0.65 0.84 0.75 0.59 0 0.71 0.73 0.8 0.78 0.73 0.77 Percentile Lowest Quartile 0.43 0.49 0.9 0.99 0.54 0.58 0.63 0.67 0.7 0.58 0.84 0.43 0.59 0.84 0.61 0.65 0.64 0.75 0.73 0.77 COR = Commissioner of Revenue vehicle registration data. NK = Data provided by Neal Kumar. *Note that in this data set, Ballston and Virginia Square were grouped together. SP = Site Plan Performance Monitoring Studies In order to summarize this data as a reference for the Working Group s discussions on minimum parking requirements, staff again took the average of the mean, median, 25 th percentile, and lowest quartile values from the Site Plan Monitoring Study, Commissioner of Revenue, and space-rental datasets. Once again, staff adjusted the Commissioner of Revenue data up by 0.1 vehicles per unit. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 28 of 29

Weighted Average Values of Parking Demand/Vehicle Registration/Vehicle Ownership based on Metro Corridor Planning Area Ballston Virginia Square Clarendon Courthouse Rosslyn Pentagon City Crystal City Highest Quartile 1.56 1.23 1.03 1.19 1.26 1.44 1.18 75th Percentile 1.23 1.17 0.99 1.04 0.89 1.17 0.97 Mean 1.06 1.00 0.85 0.93 0.78 0.95 0.90 Median 0.98 0.96 0.85 0.87 0.77 0.85 0.89 25th Percentile 0.87 0.88 0.74 0.78 0.65 0.74 0.75 Lowest Quartile 0.51 0.88 0.63 0.68 0.49 0.63 0.75 The values between the mean and lowest quartile served as the basis for the proposed range of new parking minimums shown at the February 6 th Working Group meeting. Residential Parking Working Group Meeting Ten Read-Ahead Materials Page 29 of 29